Abstract
Daily home–work travel is a habitual behavior that can be disrupted when the location of work, as one of the behavioral contexts, changes. It is then likely that individuals will reconsider their travel behavior more intentionally and choose alternative transport modes. To identify motivations and barriers to incorporating the use of sustainable modes into the individual’s daily travel, this article systematically reviews the literature on the impacts of involuntary workplace relocation on commuting behavior. Effective measures that incentivize sustainable commuting behavior are also discussed. This study on involuntary workplace relocation informs considerations of changes in travel behavior related to other contextual changes.
Keywords
Involuntary workplace relocation can happen when one or multiple job organization(s), at one or multiple location(s), move or merge (in)to another location. From the standpoint of the employees, who have very little part in initiating or controlling the move, this relocation is an exogenous life event that can influence their commuting behavior as well as their overall daily mobility. Whereas there is a small body of work about forced job relocations, other life events such as the birth of a child, residential relocation, job changes, and mobility tool ownership, that is, purchasing or disposing of a car, bike, or public transit ticket, have received a great deal of attention in the literature on mobility biographies.
During the past few decades, the concept of mobility biographies has garnered considerable attention in social-scientific research and practice, reflecting the increasing awareness of the transportation planning opportunities that lie behind behavioral changes triggered by key life events (Beige and Axhausen 2012; Clark, Chatterjee, and Melia 2016; De Groot et al. 2011; Diewald and Mayer 2009; Lanzendorf 2003, 2010; Müggenburg, Busch-Geertsema, and Lanzendorf 2015; Scheiner 2014; Scheiner and Holz-Rau 2013b; Schoenduwe et al. 2015). Mobility biographies focus on life events that can induce alterations in spatial context, lifestyle preferences, and travel behaviors (Clark, Lyons, and Chatterjee 2016; Scheiner 2007). It is suggested that when travel habits are disrupted, attentiveness to alternative modes increases, and thus, behavior is more likely to be deliberately reconsidered (Müggenburg, Busch-Geertsema, and Lanzendorf 2015; Stanbridge, Lyons, and Farthing 2004; Verplanken et al. 2008). From a planning perspective, these moments are great opportunities for introducing and encouraging the use of sustainable transportation alternatives and for promoting health and environmental concerns.
One notable life event that can induce contextual and travel behavior changes is relocation. Whether it is for personal or professional reasons, uprooting and moving to another location is a complex event from a social and psychological perspective. It exposes people to a novel situation in which their habitual coping skills may be ineffective (Hausman and Reed 1991; King, Dimond, and McCance 1987). Relocation causes alterations in various life dimensions, particularly daily mobility, which accounts for major daily life arbitration in a household (Schönfelder and Axhausen 2010). Among all types of daily displacement, commuting journeys, that is, daily home–work trips, account for a significant proportion of personal travel and are of particular importance in transport planning due to the following reasons: morning and afternoon rush hours derive the demand for capacity in road infrastructure and other transportation services, all of which dominate the urban network (Levinson and Krizek 2008). Additionally, commuting journeys play a frame-setting role for daily mobility by being much more regular and consistent than shopping and social trips (Manaugh, Miranda-Moreno, and El-Geneidy 2010; Shearmur 2006). Last but not the least, excessive dependence on cars for commuting trips, relative to other transport modes, carries significant social and environmental costs, including greenhouse gas emissions, construction and maintenance of dense road networks, provision of parking spaces, time loss in traffic congestion, and negative externalities on health (Sprumont et al. 2014).
Conceptual Model of Daily Home–Work Travel
De Witte et al. (2013) explain how making a choice for daily travel engages individuals in a very complex decision-making process involving an extensive variety of factors from various disciplines including economics, sociology, geography, and psychology. Figure 1 presents a conceptual model to illustrate that mode selection for home–work trips is determined by a whole range of physical and psychological (objective and subjective) factors that are interrelated, to a greater or lesser extent, and that involve the spatial characteristics of both home and work locations. The interconnections between indicators corroborate that commuting behavior often results from a compound decision-making process that is performed consciously or unconsciously, and habitually or intentionally, and that depends on one’s access and competences (Kaufmann, Bergman, and Joye 2004). Access refers to the range of mobility alternatives and is limited by spatial, temporal, and other contextual constraints. Accessibility to mobility options is a function of factors such as cost, logistics, weather conditions, and so on, and is highly dependent on the actors’ socioeconomic characteristics. Competences refer to the skills and abilities required to make use of access; these include physical abilities, acquired skills, and organizational skills. From the various possibilities, that is, access and competences, a commuter decides on a transport mode based on their psychological characteristics (Van Acker, Van Wee, and Witlox 2010). In other words, commuters interpret and evaluate their travel options both at home and the workplace, and then react, based on their habits, preferences, attitudes, and other factors. Kaufmann, Bergman, and Joye (2004) refer to this as the appropriation phase and suggest that the individual’s interpretation varies based on their needs and aspirations, which are inherently linked to their motives, attitudes, values, and habits. Appropriation helps to explain how different choices are made when access and competences are identical among different individuals. It also elucidates the “irrationality” of travel behaviors when decisions are not based on utility maximization attainable by minimizing travel time and costs (Manaugh and El-Geneidy 2015). It is, indeed, the complex relationship between these attributes that can explain the large amount of variation in commuting behaviors.
Conceptual model of daily commute mode choice. Up: h stands for home and w stands for work. Down: Capital letters subscripted by “h” and “w” refer to alternative travel modes available at home and work locations, respectively. Source: Author; inspired by De Witte et al. (2013) and Kaufmann, Bergman, and Joye (2004).
Countless conceptual and empirical studies have tried to explain travel behavior changes from the residential environment perspective, which is often considered the trip origin (e.g., Bohte 2010; Cao, Mokhtarian, and Handy 2009; Ewing and Cervero 2010; Krizek 2003; Schwanen and Mokhtarian 2005). Much less is known about the influence of destinations on daily mobility, whereas travel behavior is highly affected by the land use and transportation characteristics of the traveler’s destination(s), particularly the location of work (Chatman 2003; Clark, Chatterjee, and Melia 2016; Scheiner and Holz-Rau 2013a; Vale and Pereira 2016). It is suggested that employment centers with high accessibility to quality public transport stimulate a switch away from car commuting and that mixed land uses at work locations encourage a mode shift to walking and cycling. Apart from this, home and work can be experienced as both the origin and the destination entity, and thus, commuting behavior is a function of both home and work location characteristics, at almost equal weights. In addition, subjective experiences develop not only in relation to the geographical space of the place of residence but can also be formed by a constant presence in the workplace, its surrounding urban environment, and transportation services. People do not only develop, reaffirm, and change their self-conception by means of daily travel and activities in the home settings (Feldman 1990, 1996). A large part of their integration in social groups, the role they play in broader society, and thus their evaluation of their own unique self can be developed and formed in their workplace (Hassan 2012). Therefore, in a similar way to residential relocation, identifying the physical and psychological changes experienced during the period of a workplace relocation can help us understand the complexity of the travel-related decision-making process.
Finally, ignoring job location factors can lead to an over- or underestimation of the impacts of the built environment characteristics of the trip origin (home location) on commuting behavior. Where job relocation takes place “involuntarily,” as in the merger of several organizations into a single employment site, a large community of employees and their households are likely to experience more significant or unexpected changes regarding their daily mobility, as compared to voluntary job relocations. Furthermore, it is not always feasible to intervene in residential environments to moderate travel behaviors, due to economic restrictions and the size and scale of the territory to be considered.
In this context, the main aim of this article is to review studies that provide insight into (1) the factors that affect sustainable commuting decisions when an involuntary workplace relocation impacts the home–work travel patterns of a large community of employees and (2) how this understanding can help practitioners and policy makers to develop effective measures to incentivize sustainable commuting behaviors, that is, less automobile dependency and more public and active (walking and cycling) transport use. The outcomes of this article are of high practical relevance, especially with respect to sustainable travel demand management. The underlying principle is to derive benefit from identifying the motivations and barriers to using green transport modes when behavioral contexts change and to offer new transport opportunities in situations where there is increased attentiveness to alternative modes.
The remainder of this article is organized into four parts. The second section describes the methodology used to select the relevant literature that is reviewed. The third section presents the results of the review: first by discussing the data collection and methods used by the selected papers, second by categorizing the papers into different groups, and finally by summarizing and analyzing the factors that affect commuting behavior following a workplace relocation, based on the conceptual model. Next, this article discusses key findings and makes recommendations for encouraging sustainable travel behavior and presents the strengths and weaknesses of the current evidence (fourth section). Finally, suggestions for future research are provided (fifth section).
Research Methodology
Search Strategy and Data Extraction
A search of six electronic databases (Urban Studies Abstracts, Scopus, Web of Knowledge, Transportation Research Board archives, Transport Research International Documentation, and Google Scholar) was undertaken. The search syntax was limited to terms for travel and terms for workplace relocation, as outlined in Figure 2. The study searched for published and unpublished reports from the earliest possible start date to February 2018 and limited the language of publication to English. Studies were included and excluded according to the following criteria: they were included if they (i) investigated the impacts of involuntary workplace relocation on employees’ commute mode choice and 1 (ii) compared before and after journey-to-work characteristics of workers; and they were excluded if (i) job relocation was voluntary or a result of preceding life events such as residential relocation and (ii) commute mode shift was a result of other life events such as external interventions in land use and transportation characteristics of the home or work location. No study was excluded on the grounds of research design, study population, or type of relocated job. The study evaluated the identified articles on their suitability for data extraction, first by screening the abstract and then eventually by reading the full text (Scheepers et al. 2014).
Search strategy. Abbreviations: FT = free term, TI = title, AB = abstract, CT = controlled term, UT = uncontrolled term, * = truncation sign: stem word + all possible endings.
In total, the search strategy resulted in 1,452 records—1,316 after removal of duplicates. Based on the examination of the titles and abstracts, 1,242 publications were excluded. Of the remaining seventy-four publications, full texts were retrieved, thoroughly read, and again evaluated with regard to matching the inclusion and exclusion criteria. A total of fifty-four studies were excluded, which left us with twenty studies. In addition, the reference lists of all seventy-four publications were screened, which resulted in two additional studies. In total, twenty-two publications were eligible for data extraction.
From the included studies, information was extracted regarding (i) study characteristics: year of publication, location of study, study population, and response rate; (ii) methodology, including number and time of surveys; (iii) before-the-move and after-the-move transportation and land use characteristics of the employment locations (note that these variables include density, diversity, design, distance, and availability; accessibility; and quality of transportation services and surrounding activity opportunities); (iv) socioeconomic and psychological characteristics of the respondents; and (v) results and significance by evaluating the link between identified variables and commute mode shift.
This systematic review is carried out in the form of a narrative review rather than a meta-analysis. Since the number of published studies on this particular topic is small, and the scale and context of the relocation projects vary substantially, a significant statistical analysis was not feasible. Performing a meta-analysis requires that the researcher makes choices that can affect the results, including selecting a large number of studies based on a set of desired objective criteria. This is often possible when a statistical overview of the results from one or more systematic reviews is carried out. Therefore, this narrative review relies on the interpretation of results made by the authors of the selected papers, without any opportunity for the current authors to question them.
Findings
Based on the changes that took place in the land use and transportation characteristics of the new employment locations, the reviewed studies fall into three categories (Table 1): (1) relocations from the Central Business District (CBD) to a suburb or subcenter—nine cases, (2) relocations from a suburb or subcenter to the CBD—two cases, and (3) relocations within the city (or relocations associated with insignificant changes in the site’s land use and transportation attributes)—twelve cases. 2 No study was found on workplace relocation from a non-CBD location to another non-CBD location.
Characteristics of Included Studies and Overview of the Results.
Note: Studied variables. Abbreviations: BE = built environment (spatial); T = transportation; AT = attitudinal; SE = socioeconomic; PT = public transport; CBD = Central Business District.
This review of the literature is not geographically constrained, and papers from different continents have been included. The majority of the papers come from Europe (twelve), but studies from North America (three), Australia (three), and Asia (four) are also included. The logic behind the inclusion of studies from different regions is that assessing a variety in the built environment and transportation supply of cities with developed CBDs as well as a variety of people with different sociocultural traits will provide more valuable and inclusive results.
Data Collection and Methodological Approaches
Among the twenty-two reviewed papers, the collection and analysis of the data vary based on the studies’ research questions and the projects’ scale and geographical context. Some studies, such as Cervero and Wu (1998) or Aguiléra, Wenglenski, and Proulhac (2009), used census data to evaluate the impacts of employment suburbanization on a regional scale as a long-term trend rather than focusing on a single workplace relocation as do Waygood, Kitamura, and Nakai (2007) or Walker, Thomas, and Verplanken (2015).
In ten studies (Aguiléra, Wenglenski, and Proulhac 2009; Bell 1991; Hanssen 1995; Meland 2007; Sprumont and Viti 2017; Vale 2013; Van Wee and Van Der Hoorn 2002; Walker, Thomas, and Verplanken 2015; Waygood, Kitamura, and Nakai 2007; Yang et al. 2017), surveys were carried out before and after the relocation, while the rest of the studies used retrospective surveys or collected information through interviews with representatives of the companies or by reviewing the relevant planning documents. Some studies collected commute behavior information using a general classical travel survey, while others, such as (Hanssen 1995) and Sprumont and Viti (2017), collected commute mode choice information using a one-day and a two-week travel diary, respectively.
Some older studies mostly relied on descriptive analyses (Bell 1991; Daniels 1972, 1981; Wabe 1967), whereas some more recent studies had a focused research objective leading to a specific data collection process and less conventional methodological approaches (Sprumont and Viti 2017; Walker, Thomas, and Verplanken 2015; Yang et al. 2017). Few studies used statistical models and exploited the data to identify the explanatory variables of the observed impacts on commuting behavior (Sprumont and Viti 2017; Sprumont et al. 2014; Vale 2013; Yang et al. 2017). Yang et al. (2017) applied multinomial logit models, with revealed preference (RP) and stated preference (SP) data to explain the variation between anticipated and actual travel mode choice. Using standard deviational ellipses (SDE) combined with cluster analysis, Sprumont and Viti (2017) also tried to capture activity spaces of a group of university employees whose activity-travel routine was disrupted to a great extent. Finally, Burke, Li, and Dodson (2011) applied a long-term forecasting approach and discussed different decentralization scenarios for the year 2031, using strategic transport modeling to estimate aggregated modal shares, vehicle kilometers traveled, and vehicle hours traveled. For detailed information, including rates and percentages, refer to Table 1.
Relocations Associated with Significant Changes in Land Use and Transportation
Relocation from CBD to suburb or subcenter
Nine studies concerned transportation changes caused by employment suburbanization/decentralization, which started to occur as a subsequent phase of residential decentralization or urban sprawl, noticeably from 1960s (Aarhus 2000; Cervero and Landis 1992; Cervero and Wu 1998; Daniels 1972, 1981; Fernandez 1994; Parolin et al. 2001; Wabe 1967; Waygood, Kitamura, and Nakai 2007). From this date onward, a number of studies in the United States and the United Kingdom devoted attention to changes in the journey-to-work as the most dramatic consequence of employment decentralization. A common observation concerns the significant increase in the average commuting distance, referred to variously as “wasteful commuting” (Hamilton 1982), “jobs-housing imbalances” (Bookout 1990; Cervero 1989), and “spatial mismatches” (Kasarda 1988)—the impacts of which were greater on residents of the city center who experienced reverse commute. This dramatic change, in addition to inefficient accessibility of public transit, induced significant shifts from active transport, mass transit, and collective forms of travel to drive-alone automobile travel. In a study from London, UK, accessibility to public transit was difficult, even for suburban residents, as the new employment site was located almost a mile from the center of the suburban town where the train station was situated (Wabe 1967). Since the new suburban employment sites were typically office parks outside the urban core and poorly planned as regards accessibility to public transit and services, these results were not unexpected. However, it should be considered that the increased share of private transport was partly a product of the secular trend toward car ownership in that era (Daniels 1981).
Relocation from suburb or subcenter to CBD
Only two papers fall into this category (Meland 2007; Van Wee and Van Der Hoorn 2002). Both studies highlighted reduced access (80 percent reduction reported in Meland 2007) to free car parking and increased access to public transport as the principal underlying factors for a mode shift from private to public transport.
There are three main reasons why this category includes a smaller number of studies. Workplace relocations (1) often follow on the phenomenon of urban sprawl, moving jobs closer to workers’ residences (Cervero and Landis 1992), (2) aim at reducing peak-hour commute trips and traffic congestion in the urban core (Yang et al. 2017), and (3) are usually associated with company expansion as a result of organizational mergers, hence requiring larger tracts of land, which are often unavailable or inaccessibly priced in CBDs.
Relocations Associated with Insignificant Changes in Land Use and Transportation
Twelve cases concerned workplaces that moved from one or multiple addresses to one joint location, all within the city center or from the CBD to a similar peripheral area, as regards the transportation infrastructure and land use characteristics (Aarhus 2000; Aguiléra, Wenglenski, and Proulhac 2009; Bell 1991; Burke, Li, and Dodson 2011; Hanssen 1995; Loo and Chow 2011; Sim, Malone-Lee, and Chin 2001; Sprumont and Viti 2017; Sprumont et al. 2014; Vale 2013; Walker, Thomas, and Verplanken 2015; Yang et al. 2017). Seven studies showed almost no positive effect on commute mode shift. Although in most cases the new sites were centrally located and linked to the transport network, many factors hindered their accessibility. For instance, the rail station was not on a line with frequent trains, in contrast to the former locations, which were well served by public transport (a main station on a main line) and with office buildings situated within a few minutes’ walk of the station and of major activity opportunities.
The rest, on the other hand, reported a mode shift in the opposite direction, mainly as a result of relocation to a formerly planned site where intense commercial centers surrounded by high-density housing were integrated with efficient public transport systems. In one case, where relocation took place within the city, lack of car parking spaces at the new workplace induced more public transit use by commuters, even though public transit accessibility and quality remained unchanged (Aarhus 2000). Walker, Thomas, and Verplanken (2015) indicated that an eighteen-minute-shorter walking distance to public transit, when other variables are unchanged, can result in a 37 percent increase in public transit use and a 33 percent decrease in car use. These findings imply that people are likely to continue their commuting habit or improve it to more sustainable ones if land use and transportation characteristics are designed and planned in a way to stimulate this behavior.
Factors Affecting Commuting Behavior Changes during the Process of Workplace Relocation
The included studies examined a variety of factors that affect commuting behaviors and relevant decisions during the relocation process. These factors are extracted and discussed below. According to the conceptual model (Figure 1), these factors are categorized into three groups: (1) location and journey characteristics, (2) socioeconomic characteristics, and (3) psychological factors. Table 2 indicates if each factor is related directly or indirectly to mode shift.
Factors Affecting Commuting Mode-switching Decisions Directly or Indirectly.
Note: + = If presence or change of factor results in mode shift from private to (PT) public transport; − = If presence or change of factor results in mode shift from public to private transport; 0 = If presence or change of factor has no impact on mode switching decision.
Location and journey characteristics
Transportation infrastructure and parking
Nearly all studies indicated that accessibility and quality of public transportation infrastructure are positively related to a mode shift from private to public transport. Accessibility to a quality road system and arterial road network, on the other hand, was positively related to driving for work trips. Overall, seventeen studies reported a considerable mode shift—up to 75 percent—from public transit to private automobile commuting. Although this modal shift helped a great proportion of workers to save substantial time on their work trips after the move, some studies reported travel time increases (Table 1). Jobs reporting staff travel problems most frequently mentioned the inadequacy of public transport (to cope with peak hour demand), the need to transfer between modes, and, occasionally, a complete lack of service in some parts of the office catchment area. One study also reported a noticeable disparity between office hours and public transport timetables, which resulted in discouraging many employees from using the service (Daniels 1972).
In a contrasting case, Meland (2007) showed that the new job site gave easy access to the entire public transport system, including in the surrounding municipalities, whereas the old location solely provided bus service to and from the city center. This resulted in a notable reduction in car use even though free parking was available for more than 30 percent of the workers. Walker, Thomas, and Verplanken (2015) indicated that an eighteen-minute reduction in walking distance to the train station resulted in a 37 percent increase in train use and a 33 percent decrease in car use. Aguiléra, Wenglenski, and Proulhac (2009) also discovered that office decentralizations that take place adjacent to efficient rail and bus facilities can reduce the proportion of private transport trips for companies with large numbers of managerial and professional staff owning one or two cars.
Nine studies evaluated the role of car parking availability on the likelihood of a mode shift toward car-based commuting (Table 1). Five of them found that accessibility to parking, whether free or not, was a stimulating factor for car-based commuting for a high share of travelers, even when jobs were located within walking distance of public transit. Aarhus (2000) and Bell (1991) indicated that the existence of parking spaces at new job sites increased the tendency for using cars, even among employees whose travel distance shortened after the move. Sprumont et al. (2014) found that a high monthly parking fee did not outweigh accessibility to a high-quality road system when the organization wanted to discourage driving to work. In a later study, Sprumont and Viti (2017) reported that despite a general increase in home–work distance, only 4 percent of respondents indicated a modal shift to the car. The authors relate this finding to the parking costs imposed at the new workplace. Finally, in cases where accessibility to efficient public transit was accompanied by a reduction in parking availability, employees were stimulated to choose public transport modes. This happened in three case studies where the workplaces relocated to an inner-city location (Aarhus 2000; Meland 2007; Van Wee and Van Der Hoorn 2002).
Residential location and commute distance
Ten studies looked at the relationship between workplace relocation and residential (re)location (Table 1). Five studies reported residential moves or future move intentions that were directly or indirectly related to the relocation of the workplace. Overall, studies suggested that for most commuters, a long home–work distance is less important than easy access to transportation and activity opportunities when choosing green modes over the car. In the cases of employment suburbanization, there was a dramatic shift (up to 75 percent) from public transit to car-based commutes, irrespective of the change in commuting distance, as home–work distance remained unchanged or even decreased in some cases. Ironically, findings from a sampling of decentralized offices throughout the UK by Daniels (1972, 1981) indicated that residential freedom at decentralized locations led to an increase in the use of private vehicles as a commute mode, at the expense of public transport. This finding suggests that long-term decisions, such as home relocation, do not always move in the direction of improving commute travel, as different people have different priorities. In another example from Sprumont and Viti (2017), one respondent who already lived near the new workplace moved their home to a farther location because they didn’t want to live and work in the same place. According to the theory of Redmond and Mokhtarian (2001), in this commuter’s perspective, travel, per se, might have a positive utility. In other words, they gain more utility in a longer travel time than in residing close to workplace (Redmond and Mokhtarian 2001). In a before-the-move study, Yang et al. (2017) reported that those workers who, after the workplace relocation, moved their residence closer to public transport anticipated using this mode more than those who moved their residence closer to the new workplace but far from a public transport service. However, the after-the-move survey revealed otherwise, as both groups indicated a statistically significant shift toward car-based commuting.
Overall, studies failed to discuss transportation and the land use possibilities and constraints of residential areas. However, the essence of home and work can be experienced as both origin and destination entities, and thus, commuting patterns are characterized by both home- and work location characteristics almost equally. Overlooking either home- or work location factors can lead to an over- or underestimation of the impacts of the other location’s built environment characteristics on commuting behavior.
Trip chain and presence of activity opportunities
One important factor that situated a considerable share of employees outside the comfort zone for active and public transport use, and induced them to switch to private automobile use, was the absence of activity opportunities at the new employment site (Bell 1991; Hanssen 1995; Meland 2007). In some cases, the modal choice for commute trips was affected by whether or not the participant scheduled out-of-office tasks during the day. Therefore, a personal vehicle was required to perform business on the journey to and from work or in the middle of the day. Moreover, “public transit is less convenient for complex trip chains whereas private car allows versatility and flexibility” (Parolin et al. 2001, 4). Meland (2007) indicated that the use of the private automobile was positively related to having plans to do out-of-office tasks, probably because the parking fees were paid for by the employer on those days. A study from Melbourne, Australia, found a 10 percent reduction in the number of nonwork daily errands per person at the new job site, as compared to the former inner-city workplace, which offered far greater opportunities such as shopping and leisure (Bell 1991). However, the number of trips to “serve a passenger” (e.g., taking children to school) increased as a result of switching to car use.
One unexpected result from the study by Parolin et al. (2001) was the slight increase in the number of recreational and social activities undertaken after the relocation from the CBD to suburb. Since the survey data provided no indication of the location of the activities, it was not possible to say whether an activity occurred close to the workplace or close to the residence. This may imply that the tendency to participate in activities may not be dependent on employment location. Sprumont and Viti (2017) investigated the employees’ activity-travel pattern before and after the move. Using an SDE, the authors showed that after a considerable change of distance (twenty kilometer) between the old and the new workplace, the majority of commuters tended to keep their activity space close to home rather than either of the workplaces, particularly the former.
Relocation to planned sites
In a study from Singapore, Sim, Malone-Lee, and Chin (2001) marked how land use characteristics can play a significant role in reducing the distance traveled and the reliance on cars for work trips as well as the number of commuting trips generated to the CBD. The authors indicated that workplace relocation to a planned regional suburban center resulted in 78 percent of workers using public transit for the commute and only 5 percent using car. An efficient transport system connected the regional center to the mass rapid transit and several bus lines linked the area to the CBD and other parts of the island at a reasonable price. This different impact on modal choice, however, is partially related to Singapore’s particular context, where urban transportation follows a different trend: generally, the vast majority of employees commute by public transport.
Burke, Li, and Dodson (2011) investigated the impacts on transportation of a planned decentralization of employment in Brisbane, Australia, by comparing the existing transportation model with that of the year 2031, which was proposed as the planning horizon. In contrast with most of the literature on decentralization, but similarly to the case of Singapore, this model suggested that strongly planned and guided employment decentralization may not be deleterious to public transit use if jobs are clustered tightly with key suburban activity centers, if transit links are more elaborate and interconnected, and if necessary cross-suburban bus links are provided. It is, however, acknowledged that this study is a prospective study and the actual impacts are not known.
A contrasting example comes from workplace relocation to a mixed-use transit-oriented center in the inner suburb of Lisbon, Portugal (Vale 2013). The site was created with the express objective of creating a new metropolitan center (Carrière and Demazière 2002), and the necessary transportation infrastructure was put in place. As a result, the site was very accessible by both private and public transport. However, despite a slight change in commuting time, the study found a significant increase in commuting distance and in car use. The results also revealed substantial car-use inertia, irrespective of place of residence. This demonstrated that people tend to maintain their commuting time within acceptable limits. Moreover, the number of active and public transport users considerably decreased, implying that the built environment characteristics of the new location could not trigger the expected changes in employees’ mobility patterns.
Socioeconomic characteristics
Studies that have taken socioeconomic variables into account are limited, as the majority of the studies were centered on modal shift as the main variable. A few studies, such as Waygood, Kitamura, and Nakai (2007), controlled for the relationships between multiple variables including gender, household composition, commute cost, car ownership, and occupation, as well as different weather conditions and personal preferences. Most studies included a list of factors such as age, gender, ethnicity, occupational level, income, and vehicle ownership but did not explain how each of these factors influenced commuting behavior during the process of job relocation. Meland (2007) and Waygood, Kitamura, and Nakai (2007) reported that married/cohabitating participants with children aged eleven or under tend to have a lower degree of change than single commuters, who were more flexible in changing both modes of travel and residential location. Aguiléra, Wenglenski, and Proulhac (2009) looked at the socioprofessional status of the respondents. They indicated that accessibility to a subway and rapid transit system of sufficient quality resulted in a 15 percent reduction in car use and a 12 percent increase in public transport use by managers (high-income residents) who owned at least one car. On the other hand, car use almost doubled for low-income laborers, whose workplaces were in areas poorly served by public transport. Using a different approach, Fernandez (1994) and Cervero and Landis (1992) indicated that workplace decentralization negatively influenced ethnic groups and minorities, who preferred not to relocate their home after the suburbanization of their work; thus, their commuting distance, travel time, and cost increased.
One study highlighted the role of socioeconomic factors in mode choice decisions (Yang et al. 2017). Using SP versus RP, the research found that, in reality, car availability and age are significantly related to the propensity to choose the car over other modes for the commute. Workers of midcareer age (twenty-six to thirty-five and forty-six to fifty-five) showed a higher probability of choosing a car over a nonmotorized mode than did workers just starting their career. Factors such as gender and income, having multiple wage earners, having a young child, and receiving a transport subsidy, which were expected to influence mode shift in the SP survey, were found to be insignificant in the actual mode choice setting.
Travel-related psychological factors
Similarly to the socioeconomic characteristics, travel-related habits, perceptions, and attitudes of the target population have not been studied thoroughly and explicitly. Five studies reported that a majority of commuters found driving to be faster, more reliable, less expensive, more comfortable and convenient, and cleaner. In a case of employment suburbanization from Sydney CBD, Parolin et al. (2001, 8) found that 70 percent of respondents reported high levels of dissatisfaction with public transport services and mentioned that bus services were “infrequent, late, and crowded.”
Workplace relocation, as an important life event, is associated with habit discontinuity and behavior change for different profiles of individuals. These research topics are complex, and the quantification and qualitative evaluation of their elements are difficult. Overlooking these factors, however, can lead to an over- or underestimation of the impact of the built environment’s characteristics on commuting behavior. Walker, Thomas, and Verplanken (2015) was the only study concerned with travel habit formation and decay during the process of office relocation, using the Self-Reported Habit Index (SRHI) method. The research found that travel habits were weakened immediately after the move (one week after), equally for those who changed modes and those who did not. However, in the third survey, which was carried out four weeks after the move, respondents showed stronger habits with regard to the new mode than they had immediately after the relocation. These observations indicated that premove habit strength and postmove habit weakness could not thoroughly predict or explain who changed behavior after the relocation. Additionally, demographic and attitudinal variables were found to be poor predictors of commuting behavior change. To this end, research on mobility does not provide concrete proof that internal cognition or sociodemographic traits are stronger determinants of travel behavior after a workplace relocation than are spatial and infrastructural factors.
Anticipation of the workplace relocation
Previous studies have found that anticipation of future events can affect travel-related decisions such as car ownership or residential relocation (Clark, Lyons, and Chatterjee 2016). For instance, people consider purchasing a car when expecting a child or anticipating a future increase in the home–work distance. The reason for this is that changes in car ownership require significant investments of time and money and only take place infrequently (Oakil et al. 2014). In addition, factors such as car ownership levels, the number of household members with a driver’s license, and car use in previous years are positively related to car ownership levels in the current year (Kitamura 2009). Some life events are found to be more predictable than others; thus, travel-related decisions are more likely to be taken in advance when these events are anticipated. However, for other life events, decision-making usually follows the event (Oakil et al. 2014). Examples of these event types are childbirth versus home or workplace relocation. These lead and lagged effects help analyze temporal relationships between events and, thus, play a prominent role in individuals’ decisions regarding mobility tool ownership after the relocation of their workplace.
Four studies concerned the anticipation phase, that is, the time between the employer’s announcement of the relocation and the actual occurrence of the move (Sprumont et al. 2014; Wabe 1967; Walker, Thomas, and Verplanken 2015; Yang et al. 2017). However, these studies failed to evaluate the impact of anticipation on the adaptation of mobility decisions (e.g., residential move and purchasing a car) that were possibly made by relocated employees. Walker, Thomas, and Verplanken (2015) found that during the months between the announcement and the move, the company took various actions, such as offering enticing subsidy programs, to prepare staff for the relocation, with a focus on stimulating sustainable commuting after the move. However, after the relocation had taken place, the organization did not continue its pro-environmental travel behavior, as it paid the parking costs of those who continued to drive during the first six months. In the study by Wabe (1967), four years before the relocation, the firm announced the move and informed all workers who joined the company during that period. While this information was an enticement to join the company for those who found the suburban location a suitable and convenient workplace, some employees left the company or moved their residence to shorten their commuting distance. Yang et al. (2017) reported that, in response to a hypothetical situation of the workplace being relocated to a new town, workers anticipated and chose commute modes almost irrespective of travel time, but largely based on their access and competence related to their sociodemographic characteristics. However, after jobs actually moved to the new town, workers’ mode-switching decisions were predominantly made with regard to their actual travel time and place of residence. This finding implies that individuals may have an inaccurate evaluation of the enablers/limits related to their own sociodemographic characteristics. They may also show a lack of awareness of the full implications of their stated choices or an imprecise estimation of travel time and distance. To this end, it is recommended that when the anticipation of future travel behaviors is studied, personal constraints such as financial ability to afford a mode or physical ability to walk be taken into account.
Discussion
This systematic review tried to map the influences of forced workplace relocations on the commuting pattern and mode choice of the employees, by reviewing twenty-two studies. Identified factors include built environment characteristics (density, diversity, distance, transportation service at the workplace, and accessibility to activity opportunities, road systems, and parking spaces), travel-related psychological factors (mainly habits and perceptions), and socioeconomic characteristics. Among these factors, four were found to be more crucial in determining the share of employees who choose a private car as their major commuting mode following the relocation: (1) access to high-quality public transit, (2) access to (free) parking, (3) access to roads system, and (4) home–work distance. Findings imply that the impact on car use of access to parking and high-quality public transit (observed in fifteen studies) is focused on more than the influence of access to road systems and the home–work distance (observed in ten studies). While the first two factors have resulted in up to a 43 percent and 75 percent modal shift, respectively, the maximum percentage reported for the latter factors is 7 percent in Cervero and Wu (1998). One reason for this observation is probably the feasibility of measuring the variable of accessibility to public transit through spatial analysis, even though few studies have defined what they mean by high-quality public transit. Studies such as Wabe (1967), Daniels (1972, 1981), and Hanssen (1995) brought up the issues of the number of transfers between public transit modes and of crowded public transit as hindrances to its use. On the other hand, access to the road network, the quality of the roads, the traffic flow on the roads surrounding the work locations, and the average travel time by car before and after the relocation are variables that require further evaluation in the future studies.
In the aggregate, most of the research (seventeen of the twenty-three studies) concluded that relocation is associated with a significant modal shift—from 15 percent to 75 percent—from public and active transport to car use. This finding is not surprising, as most cases involved relocations from the CBD to a suburb. However, it can be concluded that, despite the considerable differences between the North American, European, Asian, and Australian contexts and among various occupation categories, individuals tend to behave similarly when facing a change in the location of their work. In other words, adaptation to geographical change is associated with private transport use. One important concern is that the extent to which this trend would apply solely to workplace relocations is still questionable. Changes in car ownership and its use for commuting trips may be caused by a range of other changes occurring in a household, such as the addition of adults of driving age (Dargay and Hanly 2007; Scheiner, Chatterjee, and Heinen 2016). Therefore, it is highly advised that more studies like Sprumont and Viti (2017), which evaluated the consequences of coincidence of events, be produced in the future.
Finally, studies support the notion that the relocation of jobs can produce significant changes in transportation demand (number of trips, mode choice, and trip length) at the enterprise level and in the area of the new location (Aarhus 2000). Therefore, workplace localization strategies should consider optimal accessibility to transport services and activity opportunities at both the local and regional scales, based on the type of job to be relocated (e.g., employees’ income level, socioprofessional situations, and working hours). Furthermore, employment relocation strategies need to take into consideration all the different objective and subjective factors that influence the adaptation of short- and long-term decisions, such as commute-mode switching and residential relocation decisions. If these factors converge with strategic policies attempting to integrate land use and transportation, this can help reduce the substantial reliance on private vehicles for commuting journeys and thus mitigate the negative impacts on the built environment and on individuals’ health. Although the studied publications focused mainly on the travel-related impacts of relocation, some proposed, be it explicitly or implicitly, policy implications and effective measures that can be adopted to incentivize sustainable transportation and encourage less dependency on private automobiles. Table 3 provides a list of these measures. For instance, charging parking fees at workplace is recommended by most of the studies, although only Sprumont and Viti (2017) indicated that this strategy was adopted in the new workplace and was indeed promising, resulting in a slight decrease in car-based commutes. Studies such as Aguiléra, Wenglenski, and Proulhac (2009), which were carried out longitudinally, highlighted that efficient public transport service can increase its rate of use over the years. To this end, it is suggested that job organizations, in collaboration with urban and transportation authorities, implement as much of these measures as possible to maximize their positive impacts.
Effective Measures to Incentivize Sustainable Transportation and Encourage Less Private-automobile Dependency.
Strengths and Limitations of This Review
This review evaluated the extent to which the existing research has progressed toward clarifying the transportation impacts of involuntary workplace relocation. It can be argued that searching through six electronic databases means most relevant research is included in this systematic review. As there always remains a possibility of missing relevant studies due to poor indexing, we checked the reference lists of all included papers to overcome this problem.
Additionally, the inclusion of a whole range of studies from various countries around the world gives a comparative inclusivity to this review, which enabled us to make fair evaluations across the whole range of potential changes that take place following the relocation of a workplace.
This study focused solely on transportation changes associated with involuntary workplace relocation. A comparative study between voluntary and involuntary workplace relocation could reveal valuable insights into the complexity of mobility decisions during different life transitions. Additionally, workplace-related changes in commuting behavior may occur as a result of a change in an existing urban context, such as the opening of a new railway station adjacent to an employment location. For instance, Brockman and Fox (2011) and Wen et al. (2005) evaluated the effectiveness of transportation interventions at workplaces and studied how employees’ commuting behavior was subsequently affected.
Strengths and Limitations of the Available Evidence
Although the land use and transportation characteristics of workplaces greatly influence commuting choices, other factors, such as the commuters’ place of residence, and socioeconomic and psychological characteristics, also play significant roles in characterizing commuting patterns. However, the quality of the existing empirical studies is not very high, as they fail to include all factors simultaneously and do not control for correlations between them. In addition, information about the statistical significance of the results was often lacking. Most of the studies did not analyze the relationship between the relocation and socioeconomic characteristics of the studied population, for example, car ownership, number of household members who work outside the home, dependent children needing a school drop off/pick up, and so on.
Almost none of the studies discussed the changes and choices of respondents with respect to the concept of utility maximization. This concept, also referred to as the microeconomic approach, follows the assumption that travelers make a rational choice among discrete alternatives to minimize travel time and costs and maximize their utility (Shen, Sakata, and Hashimoto 2009). It is, however, important that utility not be only about minimizing time and cost (quantity) but also maximizing satisfaction and well-being (quality), even at the expense of a higher travel cost. Utility in the latter sense receives less attention in the relevant literature. For instance, some people may maximize their level of satisfaction by spending quality hours in the car with family members while fulfilling daily errands on the way to and from work. In fact, these individuals have accepted the monetary costs associated with car use (fuel, parking, maintenance, insurance, etc.) as they have other priorities when making decisions on their daily travel. Others may want to maintain a sense of security by not changing their place of residence (for those with no economic constraints) and to retain psychological bonds with their present home, even after an increase in commuting distance, time, and cost following a workplace relocation. To this end, it is highly suggested that aspects of these concepts be investigated in more detail in future travel behavior studies.
Additionally, few studies considered the role of transportation and land use characteristics at employees’ home locations. The choice of a transport mode is highly dependent on the way the residence and workplace are connected. Those papers that evaluated the commuting impacts of residential location only included the impact of home–work distance and took notice of neither the transportation possibilities and constraints nor the presence of activity opportunities close to the workers’ place of residence.
Only two studies were concerned with temporal alterations resulting from workplace relocation. Bell (1991) indicated that relocation resulted in a half-hour earlier departure in the morning. A case study from Norway also found some seasonal variation in the distribution of modes (Meland 2007). It is, however, essential to predict temporal changes as they may result in a misalignment between public transit timelines and office hours. In addition, anticipating such changes can help employers provide facilities to moderate employees’ travel demands.
Finally, in most cases, methodological approaches are limited to cross-sectional data sets and, thus, neglect the temporal and longitudinal dimensions of decision-making. Studies did not look at the short-term versus long-term influences of relocation and workers’ adaptation process. Some studies examined commuting behaviors a few months after the relocation while others analyzed the impacts over a two-decade period.
Unanswered Questions and Future Research
The relocation of a workplace, as a major life event, is associated with adaptations of long- and short-term mobility decisions, such as residential relocation or mobility tool ownership. Current theoretical models of the mobility impacts of major life events often evaluate the decision-making and adaptation processes after the occurrence of the event. However, as these events are often anticipated in advance, particularly in the case of forced job relocations, it is suggested that studies focus more on the period between anticipation, that is, when the move is announced by the employer and the occurrence of the event. This period is a vulnerable stage, in which habits are weakened and attentiveness to alternatives is higher; thus, it has important implications for planning policies aimed at changing travel behaviors toward green transportation.
In addition to changing their place of residence, people may change educational establishments or their children’s school. The chain of relocations spurred by the first relocation, that is, the workplace, can influence commuting behavior on a wider level. Few, if any, such studies, evaluating the macro effects, have been undertaken so far. Additionally, research in this area is dominated by the use of quantitative research methods; however, qualitative approaches can provide in-depth insight into the experiences and processes of changing commuting behavior after a workplace relocation. As part of a qualitative analysis, it is suggested that future empirical studies not only include psychological factors, such as habits, preferences, and attitudinal characteristics of employees and their households, in survey questionnaires but also conduct face-to-face interviews to ask people for the reasons for any change (or lack of change) in their travel-related decisions.
Finally, habit as a key element in the social psychology of behavior that affects individual decision-making needs to be evaluated and considered in studies that concern changes in travel behavior. Habits are helpful for explaining the extent to which changes occur or do not occur. For instance, an identical change in commuting distance for two different individuals may result in different travel mode switching decisions, based on the level of habit strength. It is, however, important to note that the level of habit strength cannot easily be measured or used in the quantitative analysis of large samples. The reason for this is that methods such as the twelve-question SRHI, particularly when used before and after an event, require respondents to have a vivid memory of their habit, its automaticity, and frequency. Such complexity can pose methodological issues and lead to unreliable responses. A group of transportation scholars (Aarts, Verplanken, and Knippenberg 1998; Klöckner 2004) suggests that habits can explain major variations in travel mode selection behavior. If habits take control of travel decisions (e.g., daily commute), the impacts of social factors and norms on moderating travel behavior decrease. Furthermore, the rationale behind continuing habits is often to be more efficient, controlled, and target oriented, which is more attractive than having to constantly make new decisions (Schönfelder and Axhausen 2010).
Previous theoretical models that portrayed the influence of important life events on different levels of habit formation (habitualization) often focused on this process only after the life event has occurred. Additionally, these models did not distinguish between voluntary and involuntary events or between those that are anticipated a long time in advance and expected only a short time before the occurrence of the event. On the other hand, individuals with different habit strengths are likely to behave differently in the face of changing circumstances and short- or long-term adaptation phases. Therefore, it is essential to measure habit strength when evaluating the impact of a major life event, such as an involuntary workplace relocation, on the daily commute.
Future analyses of the same type of life event and its influence on commuting behavior can refine and extend these results in various ways. First, it would be useful to explore various ways of segmenting the target population. For instance, the variables potentially explaining changes in travel habits after a relocation may be weighted differently depending on occupational category or household composition. Another basis for segmentation is the individual’s status as regards habit strength, flexibility, and level of reactivity to change and specifically to such life events.
It can be hypothesized that adaptation to change may be realized differently, as some people may go beyond moderate changes by moving their place of residence, while others may adjust their ideals to fit reality.
Footnotes
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.