Current Issue

Transportation infrastructure has traditionally been regarded as a fundamental driver of national and regional economic growth, serving as a strategic policy instrument to enhance interregional connectivity and promote spatial development (Aschauer, 1990; Munnell, 1992; Rietveld and Bruinsma, 1998; Banister and Berechman, 2000). Investments in transportation infrastructure, including highways and railways, have long been closely associated with improved accessibility, expanded economic opportunities, and increased residential attractiveness (Rietveld and Bruinsma, 1998; Banister and Berechman, 2000; Vickerman, 2015). These improvements are, in turn, expected to ultimately stimulate population inflows and promote urban expansion (Garcia-López, 2012; Baum-Snow, 2007).

However, recent global demographic and spatial trends, such as population decline and widening regional disparities, raise fundamental questions about the continued validity of these assumptions. Empirical evidence increasingly reveals that the impacts of transportation infrastructure are neither uniform nor universally positive, but instead vary considerably depending on regional socioeconomic conditions and spatial characteristics (Boarnet, 1998; Chandra and Thompson, 2000; Banister and Berechman, 2001; Baum-Snow, 2007; Vickerman, 2015; Baum-Snow et al., 2017). In other words, while transportation infrastructure may enhance economic activities and residential conditions in certain regions, thereby inducing population growth, it can simultaneously cause population outflows and regional decline in neighboring areas, potentially leading to increased uneven development and regional disparities.

This dual nature of transportation infrastructure investments, serving as a catalyst for growth in some regions while accelerating decline in others, has emerged as a central concern in the field of urban and regional planning. Such uneven outcomes can be effectively explained by Myrdal’s (1957) concept of the backwash effect and Hirschman’s (1958) theory of polarization, and have more recently been referred to as the straw effect (Ono and Asano, 2005; Jo and Woo, 2014; Kim and Han, 2016). In Japan, for instance, the expansion of the Shinkansen network has contributed to mitigating population decline in certain cities, yet it has also been criticized for intensifying marginalization and disparities in areas excluded from its benefits (Reggiani and Ortiz-Moya, 2022). In China, the development of the high-speed rail network has been found to further aggravate population loss in shrinking cities, thereby exacerbating spatial imbalances between growing and shrinking cities (Deng et al., 2019). Similarly, in South Korea, although the initial introduction of the Korea Train eXpress (KTX) was expected to yield benefits through improved accessibility, empirical studies have shown that it contributed to stagnation in small and medium-sized cities and to population outflows from non-beneficiary areas, thereby widening of interregional disparities over time (Jung and Lee, 2011; Jo and Woo, 2014).

These cases demonstrate that transportation infrastructure does not function solely as a driver of regional growth but can also serve as a structural factor contributing to regional decline under certain conditions. Accordingly, this study aims not merely to challenge the conventional assumption that transportation infrastructure investment inevitably leads to population inflows and regional growth, but to systematically examine the conditions under which its effects may be either positive or negative. In particular, this study seeks to explore the mechanisms and contexts behind the ‘Win-Win effect,’ in which transportation infrastructure induces population growth across all affected areas, and the ‘Win-Lose effect,’ in which transportation infrastructure development fosters population growth in some areas while causing population decline in others.

For this study, a systematic literature review was conducted, focusing on empirical studies retrieved from major international academic databases that examine the interaction between transportation infrastructure and population dynamics. The reviewed studies are classified into two main types. Type 1 includes research that analyzes the effects of transportation infrastructure on population change, which can be further subdivided into studies identifying Win-Win and Win-Lose effects. Type 2 consists of studies that examine how population change influences the provision and expansion of transportation infrastructure. This classification aims to clarify the conditions and directions of the relationship between transportation infrastructure and population change, and to derive relevant policy implications.

The structure of this paper is as follows. Chapter II outlines the research methods and analytical procedures of the systematic literature review. Chapter III provides a descriptive overview of the reviewed journal articles, including their temporal and regional distribution and methodological characteristics. Chapter IV synthesizes the main findings by causal type and systematically analyzes the interaction between transportation infrastructure and population dynamics. Finally, Chapter V summarizes the key results and discusses policy implications derived from the analysis.

This study adopts a systematic literature review approach, which aims to comprehensively collect and analyze existing studies on a specific topic to derive synthesized conclusions. Unlike a narrative review, which relies on the subjective judgment of researchers, this method follows a rigorous and transparent selection process to ensure the objectivity and reproducibility of research findings (Petticrew and Roberts, 2008; McAlister et al., 1999). Given these methodological advantages, this approach has gained increasing prominence in recent urban studies (Kim and Yang, 2023; Seomun and Song, 2023; Oh and Jung, 2024). To enhance methodological rigor, this review followed the PRISMA 2020 (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, an updated evidence-based framework that improves reporting quality through a 27-item checklist and a four-phase flow diagram covering identification, screening, eligibility, and inclusion (Page et al., 2021). The entire review process was documented using the PRISMA flow diagram to ensure transparency and replicability.

A comprehensive literature search was conducted in two major academic databases, Web of Science and Scopus, using the keywords “transportation infrastructure” and “population change” with the AND operator to identify studies containing both terms. To ensure topic relevance and search precision, the final search string was refined through an iterative pre-testing process. This search resulted in a total of 1,817 records, including 771 from Web of Science and 1,046 from Scopus.

The selection process followed the PRISMA 2020 guidelines (see <Figure 1>). In the identification phase, 812 records were excluded due to duplication, language restrictions, or classification as non-academic sources such as conference proceedings and other forms of grey literature. Only English-language and peer-reviewed journal articles were retained, leaving 1,005 records for screening. In the screening phase, titles were reviewed first, leading to the exclusion of 906 records unrelated to transportation, population, or urban and regional planning, including those in fields such as environmental science, public health, energy systems, or disaster management. The remaining records were then reviewed by abstract, and 53 were excluded for not addressing both transportation infrastructure and population change. In the eligibility phase, full-texts of 46 articles were assessed. Of these, 3 could not be retrieved. Among the 43 accessible articles, 4 were excluded for employing virtual, theoretical, or predictive modeling approaches, and 11 were excluded for lacking empirical analysis evaluating effectiveness. As a result, only studies employing empirical research designs were retained for the final synthesis, yielding a total of 28 journal articles. This systematic and transparent selection process ensures that the reviewed literature provides robust empirical evidence on the relationship between transportation infrastructure and population dynamics.

Figure 1. 
PRISMA flow diagram of the systematic literature review

<Figure 2> illustrates the temporal and regional distribution of the 28 journal articles. The number of publications steadily increased over time, with one article (3.6%) published in the early 1990s, two (7.1%) in the mid-2000s, six (21.4%) in the early 2010s, eight (28.6%) in the late 2010s, and eleven (39.3%) in the early 2020s. Earlier studies predominantly focused on North America and Europe, but from the 2010s onward, the geographic scope expanded to include Asia. By the early 2020s, Asia, particularly China, emerged as the primary focus of scholarly attention. During this period, studies on Africa also began to appear, indicating a growing academic interest in non-Western contexts.

Figure 2. 
Temporal and regional distribution of journal articles

<Table 1> presents a descriptive overview of the reviewed journal articles, including their spatial scope, geographic focus, temporal coverage, methodology, keywords, and citation counts. Spatial scope is classified based on whether the analysis is confined to a single subnational administrative unit (e.g., state, province, or region) or extends across multiple subnational administrative units or national territories. This classification reflects the different scales used to examine the relationship between transportation infrastructure and population change, ranging from localized effects to broader regional dynamics. The studies also differ in their units of analysis, with some focusing on individual cities such as London, Barcelona, or Jeddah, and others on larger administrative units such as provinces, prefectures, or metropolitan regions. Most studies employ longitudinal data spanning 10 to 30 years, with some extending beyond five decades, underscoring a structural and cumulative understanding of transport-population dynamics rather than short-term fluctuations. A wide range of quantitative methods is applied, including ordinary least squares (OLS), spatial regression models (spatial OLS, spatial error model, spatial lag model), instrumental variable techniques (IV, 2SLS), panel data models (PCSE, GLS), and the difference-in-differences (DID) methods. These methodological approaches reflect increasing efforts to address spatial dependence, endogeneity, and causal inference. Citation counts based on Web of Science indicate varying levels of scholarly impact, from highly cited foundational works to more recent contributions reflecting emerging trends.

The keywords listed in each article were compiled and analyzed to trace the thematic evolution of the field. These author-assigned keywords reflect how researchers framed the core concepts of their studies. For this review, a keyword co-occurrence network analysis was conducted using the VOSviewer, based on keywords that appeared at least twice across the selected articles. <Figure 3> presents the results of this analysis, where each node represents a keyword and links indicate co-occurrence relationships. The color gradient represents the average publication year in which each keyword appeared, ranging from blue (early 2010s) to yellow (up to 2024). While the dataset extends back to the 1990s, the most prominent keywords cluster between the early 2010s and early 2020s, enabling a temporal interpretation of evolving research priorities.

Figure 3. 
Temporal co-occurrence network of keywords in journal articles

Across all periods, core keywords such as ‘transportation’, ‘infrastructure’, ‘population’, ‘growth’, and ‘impacts’ have consistently occupied central positions in the network, confirming their foundational role in shaping scholarly discourse. In the early 2010s, studies frequently focused on urban spatial structures and decentralization, as evidenced by keywords such as ‘cities’, ‘sub-center’, ‘expansion’, ‘sprawl’, and ‘suburbanization’. In the mid-to-late 2010s, attention shifted toward urban expansion and land-use change, with keywords such as ‘metro’, ‘rail transit’, ‘highways’, ‘transportation’, ‘infrastructure’, ‘urban growth’, and ‘land use’ becoming more prominent. Around 2020 and beyond, newer studies increasingly addressed high-speed transportation networks and their broader regional effects, with keywords such as ‘high-speed rail(way)’, ‘region’, ‘integration’, ‘costs’, ‘productivity’, ‘efficiency’, and ‘accessibility’ gaining visibility. These reflect a shift toward policy-oriented research that examines regional integration, economic spillover effects, and accessibility, while also addressing spatial inequalities and regional disparities. This thematic evolution suggests a broadening of the research perspective, from localized urban structure to macro-scale policy implications.

Lastly, a review of citation counts based on Web of Science data identified several key studies that have had a significant impact on the field. Notably, Aljoufie et al. (2013), Levinson (2008), Garcia-López (2012), and Goetz (1992) recorded high citation counts and are recognized as foundational works on the relationship between transportation infrastructure and population dynamics. Collectively, these studies cover a broad range of topics, including high-speed transport networks, land use, urban spatial structure, and the effects of air and rail infrastructure, and have served as critical references that shaped subsequent research in this field.

The reviewed journal articles are classified according to the types of interactions between transportation infrastructure and population dynamics, as shown in <Table 2>. Broadly, they fall into two main categories. Type 1 includes studies that examine the effects of transportation infrastructure provision on population change, whereas Type 2 comprises studies that investigate how population change influences the development of transportation infrastructure. Within Type 1, three subtypes are identified based on observed patterns of population change following infrastructure provision: the ‘Win-Win effect,’ the ‘Win-Lose effect,’ and the ‘Lose-Lose effect.’ Rather than applying a predefined theoretical framework, this study adopted an inductive approach, deriving a typology from recurrent causal patterns and narrative structures identified during the review process. Each subtype is defined according to the combination of regions experiencing population growth or decline after the provision of transportation infrastructure.

The ‘Win-Win effect’ refers to cases where transportation infrastructure expansion leads to population growth in all affected regions. This pattern typically results from improved accessibility and interregional connectivity, which stimulate local economic revitalization and enhance residential attractiveness, thereby fostering widespread population inflows. The ‘Win-Lose Effect’ describes situations where transportation infrastructure development promotes population growth in some areas while causing population decline in others. In such cases, improved accessibility in certain locations increases their attractiveness as economic and residential hubs, while adjacent areas with relatively lower competitiveness experience out-migration, resulting in spatial imbalances. The ‘Lose-Lose Effect’ refers to cases in which population decline occurs across all affected regions despite substantial transportation infrastructure investment. This subtype typically pertains to shrinking cities or persistently depopulating regions, where large-scale public funding for transportation infrastructure fails to reverse population decline due to limited socio-economic revitalization effects. However, no studies in the reviewed literature empirically confirm this effect, which may be attributable to the scarcity of research focusing on such regions, limitations in the scale or timing of infrastructure investments, or data constraints that hinder empirical analysis.

Studies examining the influence of population change on transportation infrastructure development (Type 2) can be categorized into two distinct scenarios. The first involves cases in which infrastructure expansion corresponds proportionally to population growth, thereby effectively accommodating increased mobility demands. The second scenario includes cases of mismatch, where infrastructure provision lags behind population growth, resulting in persistent congestion, overburdened networks, and spatial mismatches. These typological distinctions underscore that the responsiveness of infrastructure development to demographic shifts is not uniform, but is highly contingent upon regional socio-economic conditions and spatial planning contexts. The main findings of the systematic literature review are summarized in <Table 3> and are explained in detail in the following section.

This study conducts a systematic literature review to examine the interaction between transportation infrastructure and population dynamics. A total of 28 peer-reviewed studies were analyzed, with attention to how these effects vary by infrastructure type and regional context.

First, road infrastructure is generally associated with a positive relationship to population growth, demonstrating the ‘Win-Win effect’ (Voss and Chi, 2006; Aljoufie et al., 2013; Iacono and Levinson, 2016; Kasu and Chi, 2018; Jedwab and Storeygard, 2022; Faiyetole and Adewumi, 2024). However, this effect is not uniform across all contexts. In regions where road networks are already well developed, the marginal benefits of additional road expansion tend to diminish, suggesting that the impact of road infrastructure on population growth is not linear, but conditional upon temporal and spatial conditions (Voss and Chi, 2006; Iacono and Levinson, 2016; Kasu and Chi, 2018; Jedwab and Storeygard, 2022; Faiyetole and Adewumi, 2024). In particular, the effects of road infrastructure overall are greater in small and remote cities (Jedwab and Storeygard, 2022), and local roads have been found to have a statistically significant impact on population growth (Iacono and Levinson, 2016). In the case of highways, some studies have pointed to the ‘Win-Lose effect,’ in which population is dispersed from urban cores to suburban or outer areas, thus promoting suburbanization (Garcia-López, 2012; Yudhistira et al., 2019). Nonetheless, the degree of suburbanization varies depending on demographic characteristics, tending to be minimal or even reversed in areas with a high proportion of young adults who prefer the amenities of central locations, while becoming more pronounced in areas with many households with children seeking lower housing costs (Meijers et al., 2012).

Second, rail infrastructure also contributes to improved accessibility and mobility within cities, and thus tends to have a greater effect on population growth in less developed or newly developing areas (Loo et al., 2017; Coronado et al., 2019; Han et al., 2023). However, this process may result in a ‘Win-Lose effect,’ where the benefits are concentrated in specific regions or among certain social groups, potentially leading to unintended socio-spatial disparities such as gentrification. In areas that have undergone a certain degree of urbanization, rail infrastructure has been associated with the outward movement of population from city centers, contributing to suburbanization, a pattern observed in multiple urban cases (Levinson, 2008; Garcia-López, 2012; Kasu and Chi, 2018). Nevertheless, as a mode of enhancing public transit convenience within urban cores, railways are generally considered a weaker driver of population decentralization compared to highways (Yudhistira et al., 2019). In the case of HSR, several studies have reported a Win-Lose effect, where some transit regions experienced population growth, while many small- and medium-sized cities or surrounding underdeveloped areas saw population decline, leading to increased concentration in major cities (Li et al., 2020; Li and Chen, 2023; Wang et al., 2023; Han et al., 2023). This phenomenon is often explained by the siphon effect, in which improved accessibility to large cities leads to the outflow of population and resources from nearby smaller cities to major urban centers. Notably, the closer a region is to a large city along the HSR corridor, the stronger the outflow effect tends to be, whereas areas located beyond a certain distance generally experience population inflows, creating an inverted U-shaped pattern (Li and Chen, 2023; Han et al., 2023).

Third, airports and seaports have generally functioned as regional growth poles, contributing to population inflow and economic revitalization, thereby exemplifying the Win-Win effect (Goetz, 1992; Breidenbach and Mitze, 2016; Kasu and Chi, 2018; Sheard, 2021). By offering broader regional and international connectivity than roads or railways, they have played a key role in enhancing regional competitiveness.

Fourth, an important implication is that population change influences the provision and expansion of transportation infrastructure. Some studies have analyzed that population growth drives the expansion of road networks and public transit systems (Makkonen et al., 2013; Breidenbach and Mitze, 2016), suggesting the possibility of interactive, bidirectional causal relationships (Iacono and Levinson, 2016). However, when infrastructure provision lags behind the pace of population growth, it may lead to negative effects such as traffic congestion and decreased accessibility (Aljoufie et al., 2013; Aljoufie, 2021). Conversely, in areas experiencing population decline, oversupply of infrastructure may result in a mismatch between supply and demand.

Based on the synthesized evidence from the systematic literature review, the following implications can be drawn for future research and policy implementation. First, transportation infrastructure does not produce uniform outcomes across regions as its effects are conditional upon spatial, demographic, and temporal factors. For instance, the impacts of road infrastructure tend to be greater in small and remote areas (Jedwab and Storeygard, 2022), and the expansion of local roads can further stimulate population inflows (Iacono and Levinson, 2016). In contrast, highway development in already urbanized regions may accelerate suburbanization or lead to spatial dispersion of population (Garcia-López, 2012; Yudhistira et al., 2019). These findings suggest that transportation infrastructure investments should be tailored to specific regional contexts. In growth regions or areas anticipating future population increases, proactive infrastructure expansion may be appropriate. However, in regions facing persistent decline, strategies should prioritize the efficient operation of existing infrastructure and link transportation planning with employment and industrial development policies.

Second, the findings underscore the need to integrate demographic and economic indicators when evaluating the effects of transportation infrastructure. Population decline does not necessarily entail economic decline, and some studies report that infrastructure-induced decentralization has alleviated congestion or created opportunities for commercial redevelopment (Levinson, 2008; Garcia-López, 2012). Therefore, a narrow focus on population figures may lead to misguided conclusions. Future research should adopt a multi-dimensional evaluation framework that considers socio-economic outcomes alongside demographic trends.

Third, the effects of transportation infrastructure are typically realized over the medium to long term. For example, the inverted U-shaped impact of HSR on population redistribution (Li and Chen, 2023; Han et al., 2023) illustrates the importance of spatial proximity and timing. Accordingly, long-term monitoring and flexible planning systems are essential. Rather than relying solely on short-term indicators such as traffic volume or land value changes, infrastructure planning should address broader socio-spatial transformations, including gentrification, suburbanization, or regional disparity. In particular, in rural areas experiencing severe population decline and aging, transportation strategies should be aligned with integrated service systems encompassing housing, healthcare, and transport welfare, and should support balanced development between urban and rural regions.

Fourth, this study confirms not only the unidirectional causal relationship in which transportation infrastructure affects population change, but also the reverse causality in which population growth drives infrastructure expansion (Makkonen et al., 2013; Breidenbach and Mitze, 2016; Iacono and Levinson, 2016). Failure to account for such mutual interactions may lead to mismatches between demand and supply, for example, congestion and reduced accessibility in areas where infrastructure lags behind population growth (Aljoufie et al., 2013; Aljoufie, 2021), or resource inefficiency due to oversupply of infrastructure in shrinking regions. Therefore, adaptive transportation infrastructure policies that respond to real-time demographic trends are essential.

Lastly, most of the reviewed studies are concentrated in the United States, Europe, China, and certain parts of Asia and other developing countries. As such, the applicability of their findings to the institutional and spatial context of South Korea remains limited. In particular, considering the rapid population decline and the accelerating risk of local extinction outside the Seoul metropolitan area, there is a pressing need for empirical research focusing on South Korean cases. Moreover, this study only proposed the possibility of a ‘Lose-Lose effect’ at a conceptual level and did not include empirical verification or specific case analyses. Future studies should more thoroughly examine the structural barriers that hinder the effectiveness of transportation investments in shrinking regions. This will contribute to more precise and context-sensitive policy design in South Korea.