Abstract
The present study investigates whether there is any evidence of club convergence for incidence of per capita road accidents across Indian states. The study uses the quantitative technique of the Phillips and Sul convergence approach to discover state convergence patterns based on the per capita accident cases. Although the findings suggest a divergence for aggregate panel of states, but there is existence of convergence among few states together as a group. Furthermore, our analysis shows a divergence in the number of road accident cases in both urban and rural areas, but there is evidence of club convergence across states in both urban and rural areas. The divergence finding clearly indicates that traffic accidents occur more often in certain states. The government must pay attention and should come up with more stringent policy to reduce the cases of road accidents in India.
Introduction
Road traffic crashes (hereafter, RTCs) are one of the important public health issues because of the enormous human and economic costs involved. According to the World Health Organization, road accidents kill around 1.36 million individuals every year, making them the top cause of death among those aged 5–29. Non-fatal injuries afflict another 20–50 million individuals worldwide, with many becoming disabled as a result. Road traffic fatalities in low and middle-income economies account for about 90 per cent of all casualties (Gopalakrishnan, 2012; Perel et al., 2007). RTCs remain a concern not only in low and middle-income countries but globally. The United Nations Sustainable Development Goals (UNSDG) also focus on road safety issues. The UNSDG’s 11.2 target states, ‘By 2030, provide access to safe, affordable, accessible and sustainable transport systems for all, improving road safety, notably by expanding public transport, with special attention to the needs of those in vulnerable situations, women, children, persons with disabilities and older persons’. The continued presence of RTCs has been a cause of anxiety on the Indian subcontinent. Despite having around 1 per cent of the global automobile population, India was responsible for over 6 per cent of road traffic accidents worldwide. Nearly 70 per cent of the accidents involved young Indians. It should be no surprise that RTCs continue to be the major cause of injury, fatality and hospitalisation in India, ranking first among 199 nations in terms of road accident mortality (MoRTH, GoI, 2019). In 2019, there were over 151,000 fatalities due to road accidents. Dash et al. (2020) empirically assessed some of India’s major causes of road accidents. They highlighted that reckless driving is primarily to blame for increasing road fatalities in Indian states. In addition, a lack of safety awareness among pedestrians and cyclists increases traffic accidents throughout Indian states. Other human factors, such as drunk driving, are mostly to blame for the annual rise in traffic fatalities. Poor road conditions and faulty automobiles are significant non-human factors contributing to traffic injuries and fatalities. There are a reasonable number of studies on RTCs, their risk factors and their determinants (see, for example, Dash et al., 2020, 2021; Dutta et al., 2020; Febres et al., 2020; Panda, Dash et al., 2022; Panda, Mishra et al., 2022; Raja et al., 2021; Rolison et al., 2018; Sebego et al., 2014; Tsala et al., 2021). Among the studies that have looked into the linkage between RTC fatalities and economic development are Akinyemi (2020), Law et al. (2011), Sinha et al. (2021), Söderlund and Zwi (1995) and van Beeck et al. (2000).
The classical economic growth models, such as those proposed by Solow (1956) and Romer (1986), contend that technical innovation, population growth and capital accumulation play essential roles in long-term economic growth. The rate of vehicular ownership (e.g., vehicles per capita) and the fatality rate per vehicle (deaths per vehicle) represent the two components of the RTC fatality rate (i.e., traffic deaths/population). While the first component reflects capital accumulation, the second component is more susceptible to technical development (e.g., improvement in road building and maintenance, vehicle safety and trauma care technologies). Capital accumulation is important in the early phases of economic development because it increases vehicle ownership, which increases the incidence of RTC deaths. However, in later phases of economic development, when numerous states have achieved the common production frontier, technological progress is the dominant driver of growth. As a result, technical improvements in areas such as road safety, vehicle safety and medical treatment are projected to be the major drivers driving significant decreases in the RTC death rate at higher levels of economic development (Nghiem et al., 2013).
However, there is little research on analysing the trends of traffic accident fatalities. To the best of our knowledge, Castillo-Manzano et al. (2014) and Nghiem et al. (2013) have examined the trends of number of death cases due to road accidents using convergence analysis. Their studies primarily focused on OECD countries and looked at how the RTC fatality growth rate changed over time by employing sigma and beta convergence. The present study complements Castillo-Manzano et al. (2014) and Nghiem et al. (2013) by focusing on Indian states. Our research varies from the previous literature on the following grounds. First, although numerous studies have examined the convergence analysis of various key indicators such as per capita income, productivity, government revenue, debt and prices across India (see, for instance, Akram & Rath, 2019; Akram & Rath, 2021; Akram et al., 2020; Bhattacharya et al., 2018; Jangam et al., 2021; Mishra & Mishra, 2018; Rao et al., 1999; Rath & Madheswaran, 2010; Sahoo, 2020; Sofi & Durai, 2016), none of the studies has looked into state-level convergence analysis of rising road accident cases in India. Though some literature has discussed the issues of rising accident cases in the countryside, the studies related to state-level per capita accident cases and their convergence paths are scanty. As a result, the focus of our research is on the convergence path of aggregate accident cases in India. The examination of this issue is important in the context of India because it records around 412,432 road accidents and ranks number one globally with 153,972 deaths due to road accidents (MoRTH, GoI, 2021). Thus, it is imperative to check whether there is a common transition path for number of per capita road accidents across Indian states, which resulted in such high number of accidents and fatalities in India. Second, we investigate this issue by focusing on both urban and rural areas. The number of cases of per capita road accidents not only varies across Indian states but also varies between urban and rural areas. The number of cases in urban areas is higher than in rural areas. Therefore, it would be interesting to examine the transition path for road accident cases in both urban and rural areas. To do so, we use panel club convergence technique developed by Phillips and Sul (2007, 2009).
The study finds a divergence in case of road accidents for aggregate panel of states, whereas there is existence of convergence among few states together as a group. Further, our study finds a divergence in the number of road accident cases in urban areas and no evidence of convergence for road accident cases in rural areas.
The rest of the article is organised as follows. The data and methodology are presented in Section II. Section III discusses the empirical results and Section IV concludes.
Data and Methodology
Methodology
The Phillips and Sul (PS) convergence is a novel way to identify the ‘club convergence’ in the literature. The PS (2007, 2009) model analyses convergence in the situation of study cross-section averages in the case of heterogeneous time-varying components, focusing on relative convergence rather than absolute convergence. Let us discuss the PS convergence in detail with the following equation:
where,
Equation (2) is segmented further into two components, (a) Ait stands for a ‘systematic component’; (b) Bit indicates a ‘transitory component’.
Equation (3) is transformed by multiplying and dividing
where
The PS approach shows that the null and alternative hypothesis is represented as:
where
Data
The current research examines 28 states and 6 union territories (UTs) from 2001 to 2020. In the year 2000, three new states were formed: Chhattisgarh, Jharkhand and Uttarakhand. Data were available beginning in 2001. The state of Telangana was formed in 2014, and for the sake of this research, it was clubbed with the erstwhile state of Andhra Pradesh. The research extracts data for our aggregate analysis from the Centre for Monitoring Indian Economy (CMIE) database. RTCs represent road accidents per million population. The study tries to investigate convergence in case of different regions of rural and urban accident cases. Due to paucity of data on urban and rural divide pre-2008, the study will use data from 2008 to 2020. We excluded the UT of Lakshadweep from our study owing to a lack of data. All road accident data are extracted from the CEIC India premium database.
First, we look at the average number of accident per million population instances across India’s sub-regional level. Figure 1 shows the average number of traffic accidents in India’s 28 states and 6 UTs. Goa has by far the most RTCs, followed by Puducherry and Kerala. Bihar, on the other hand, has fewer accidents despite being one of India’s most populated states. The economic development (increased vehicle ownership) can be regarded as one of the reasons for higher RTCs. According to the NFHS-5 study, Goa (46 per cent) and Kerala (26 per cent) have the greatest proportion of car-owning households, while Bihar has the least car ownership (2 per cent).
State-wise Average Road Accident Cases.
Following that, we look at the club convergence of accident instances from Indian states as well as their rural and urban areas.
The club convergence of aggregate road accident cases in India is shown in Table 1. The period of the analysis was from 2001 to 2020. It reveals that the null hypothesis is rejected at a 5 per cent significance level, suggesting divergence in RTCs among Indian states. The dispersion of RTCs across Indian states reveals that the states are not converging to a single path equilibrium, leading us to study if the states are converging to numerous equilibria, signalling the potential of clubs. The statistics indicate 4 clubs among the 28 states and 6 UTs, which are significant at 5 per cent, indicating non-rejection of null hypothesis of club convergence. Here, the log (t) test is used to find the small club and merge it with the appropriate large clubs. From our finding, we see a merger of one club, and finally, they framed three final clubs where different states followed their transaction path. Club 1 comprises 14 states, namely Assam, Chhattisgarh, Goa, Haryana, Himachal Pradesh, Jammu and Kashmir, Karnataka, Kerala, Madhya Pradesh, Nagaland, Odisha, Puducherry, Tamil Nadu and Uttar Pradesh. This suggests that these states are formed on a steady-state path with a large number of accident instances. Similarly, the other two clubs represent states where the relative number of RTCs was lower than the first club (as seen in Table 1). Club 2 has 17 states, whereas Club 3 has 3 states. Then we investigate whether there are any patterns or trends of RTCs in rural and urban areas. Tables 2 and 3 show the convergence of urban and rural road accident cases in Indian states, respectively.
Club Convergence of Per Capita Aggregate Road Accident Cases.
Club Convergence of Per Capita Urban Road Accident Cases.
Club Convergence of Per Capita Rural Road Accident Cases.
Table 2 illustrates that there is a disparity in the case of RTCs at the sub-regional level in urban areas. Due to a lack of data before 2008, the research considers from the period 2008 to 2020. The analysis finds divergence at the aggregate panel level but traces convergence at the club level. The results further indicate that all 34 states and UTs are forming 4 different clubs and each club follows a separate transition path. However, total of three clubs remain post merger. Similarly, Table 3 depicts the club convergence of accident incidence in rural Indian states. The findings also demonstrate a panel divergence for all states together, but presence of club convergence at sub-panel level. In case of a rural road accident, the data from clubs show the availability of three transition routes within Indian states after merging. The five states, that is, Chhattisgarh, Goa, Jammu and Kashmir, Tamil Nadu and Uttar Pradesh on the other hand, do not converge or diverge with any of the aforementioned clubs, and instead follow their own transition path.
For more clarity on the clubs of per million accident cases of different regions, the study plots the transition paths of different clubs in Figures 2, 3 and 4. It shows the club-wise ‘average relative transition paths’ of different states based on the per capita accident cases of Indian states. From Figure 2, we validate that the states are following different transition paths of accident cases in the case of Indian states. Similarly, the relative clubs of other urban and rural regions are plotted in Figures 3 and 4. According to the data, the figure shows that road traffic collisions in Indian states are on a noticeable upward trend.
Relative Transition Path of Aggregate Accident Cases.
Relative Transition Path of Rurban Accident Cases.
Relative Transition Path of Rural Accident Cases.
The number of road accidents and number of death associated with road accidents in India has been increasing over the years. While few studies examined the factors associated with the high number of road accidents, none of the previous studies analyse its trend patterns across Indian states. The analysis of the patterns of trend for the cases of road accidents across states is important from policy perspective because it will help to understand whether the severity of RTCs originates from only few states or whether the patterns are common for all the states. The present study made an attempt to investigate this issue. We employed a panel club convergence technique on the number of road accident cases across 34 states and UTs using annual data from 2001 to 2020. The results based on Phillips and Sul panel club convergence method found a divergence in case of road accidents across Indian states. This implied that the number of road accidents across Indian states do not have any common transition path and hence the patterns of such road accidents are different. Further, our results also find similar results in case of urban areas and rural areas of Indian states. The divergence results further motivated us to identify the states that follow a common transition path. We noticed that there is evidence of convergence for RTCs at club level, which implied that all 34 states and UTs are forming three clubs with some commonalities among the states within a club. From policy perspective, it is important for the government to come up with a more stringent policy to reduce the cases of road accidents in India. One plausible solution is to create rules and regulations by focusing on those states who are on the higher side of the convergence club.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The authors received no financial support for the research, authorship and/or publication of this article.
