12. Literature Review.md 5.1 KB
2 Literature Review
Road‑traffic safety has traditionally been measured by a single indicator – the road‑fatality rate, i.e., the number of deaths per 10 000 motor vehicles (Elvik, Vaa, & Christensen, 2019). Although this metric provides a clear picture of mortality levels, its scope is limited: it ignores both the frequency of crashes and their severity, dimensions that can vary markedly across countries and affect the effectiveness of safety policies (Peden et al., 2020; World Health Organization [WHO], 2023).
To address this shortcoming, recent research has developed multi‑indicator and composite‑index frameworks that combine variables such as total deaths, the number of fatal accidents, and the overall volume of road crashes (Wang & Li, 2021; Liu, Wang, & Chen, 2018). These studies have shown that composite indicators improve the predictive accuracy of RTS assessments and enable the identification of specific risk factors that are not revealed by a single‑rate measure (Bennett & Lee, 2022).
Methodologically, structural equation modelling—and in particular path analysis—has proven to be an effective tool for quantifying the interrelationships among safety components (Bollen, 1989; Liu et al., 2018). Path‑analysis applications have been reported for evaluating both traffic‑flow intensity and crash severity in other European nations (Gkiotsalitis & Chassiakos, 2020) as well as in Asian contexts (Zhang et al., 2019). In the case of Estonia, however, research that integrates multi‑layered indicators with structural modelling is scarce. Earlier Estonian studies have largely relied on single statistical descriptors (Statistics Estonia, 2021) or simple regression models without mediation pathways (Kaasik & Tamm, 2017). Consequently, a clear scientific gap exists: the absence of a multi‑indicator framework that combines road‑fatality rate, total deaths, fatal‑accident counts and overall crash volume while assessing their relative effects on RTS through path analysis. The present study fills this void by offering an in‑depth multi‑indicator model for Estonia and testing its performance against the traditional single‑indicator approach.
In recent years scholars have increasingly turned to overcome the limitations of one‑indicator approaches by constructing multi‑indicator and composite‑index frameworks that better capture the complexity of traffic safety. For example, Weijermars, Van Acker, and De Vos (2014) emphasized the need to augment road‑fatality rates with transport‑related, demographic, and behavioural variables in order to create cross‑national safety profiles and explain mortality variations across populations of millions. A comparable European‑wide investigation employed an extensive indicator set—including crash frequency, severity and traffic volume—to pinpoint the most deterministic factors explaining deaths per million inhabitants; this enabled group‑based comparisons such as those conducted for Belgium (see also Weijermars et al., 2014).
From the Asian perspective, ASEAN‑region research (Chen, 2020) demonstrated that a readily adaptable holistic index—the Road Safety Assessment Index (RSAI)—can evaluate national road‑infrastructure quality, vehicle characteristics and user behaviour while monitoring progress toward strategic safety goals. A related safety‑system framework was advanced by Jameel (2019), who organised thematic sub‑domains—infrastructure, vehicles, and driver behaviour—and built subordinate models that allow rigorous cross‑country benchmarking.
Methodological advances in composite indicators have also been highlighted. Coll and Marshall (2013) showed how hotspot identification combined with weighted aggregation improves the detection of high‑risk areas, yet they warned about theoretical shortcomings of traditional techniques such as correlation bias and compensatory effects. An IRTAD report focusing on driver fatigue identified fatigue as a significant risk factor in 38 countries, thereby providing an additional metric for safety assessment (IRTAD, 2020).
International benchmarking has become an essential instrument for policy learning. Maeso (2021) stressed the diversity of exposure metrics—population size, vehicle fleet, travel distance—and their influence on risk‑profile identification, underscoring the need for standardized measurement dimensions. Likewise, Triwijaya et al. (2022) argued that successful benchmarking requires a structured cycle and clear frameworks in which composite indicators serve as central components for sharing best practices and fostering continuous improvement.
In sum, recent literature demonstrates that multi‑indicator models—whether applied at the European, ASEAN or global level—provide deeper and more actionable insights into traffic‑safety dynamics than traditional mortality‑rate measures alone. These approaches support policymakers in making evidence‑based decisions and contribute directly to achieving SDG 3.6.
References cited in this section are listed in the manuscript’s reference list.