Predicting personal injury crash risk through working conditions, job strain, and risky driving behaviors among taxi drivers

Today, taxis are an important component of China’s urban public transit network. However, taxi drivers frequently work under extremely stressful and hazardous conditions, including long working hours, frequent driving, and occasional disputes with passengers, all of which increase drivers’ physical and mental stress. A survey in Beijing reported that taxi drivers worked up to 11 h per day and at least 27.8 days per month [1]. Taxi drivers in Tlahuac, Mexico City were found to drive 11.1 h a day behind the wheel [2]. Another survey in Sydney showed that 67% of taxi drivers drove at least 50 h per week [3]. Obviously, taxi drivers’ e daily earnings rely heavily on their driving distance within a scheduled period, so they must spend long hours driving in order to carry more passengers and earn more money. Therefore, it is no surprise that taxi drivers engage in occupational risks while driving.

Many studies have been conducted to examine the significant effect of exposure to hazardous working conditions on taxi drivers’ health and to identify potential risk factors contributing to road crashes, such as drivers’ age, job experience [4], license, type of employment [5], and physical and mental health [3, 6, 7]. Generally, taxi drivers are regularly obliged to work long hours into the late night or early morning and are therefore vulnerable to fatigue or sleepiness while driving [3, 4, 8, 9]. Moreover, occupational illness [10], irregular shifts [3], and income dissatisfaction [11] have also been significantly correlated with taxi drivers’ risk of serious crashes.

Because taxi drivers’ daily earnings depend on their driving distance with occupants, they try to drive faster to save time and carry more passengers; as such, they often commit traffic violations. A study of taxi collisions in Taiwan [12] revealed that 25.6% of the observed drivers committed at least one speeding violation during a one-year period, and the frequency of speeding behavior was significantly related to the driver’s age, years of job experience, daily driven kilometers, number of off-duty days per month, etc. Moreover, taxi drivers who exhibited more passive-aggressive driving behaviors were found to be easily distracted and to have an increased crash risk [13]. A survey in Wuhan, China showed that taxi drivers’ attitudes about driving violations had a serious negative influence on their risky driving behaviors [14].

To date, there have been very few studies regarding taxi drivers’ work-related stress and risky driving behaviors. Understanding the relationship between these factors is essential to devising potential strategies to reduce and prevent road accidents and injuries for taxi drivers. Thus, this research aimed to 1) examine the association between their sociodemographic information, self-reported work-related stress, frequency of risky driving behaviors, and involvement in personal injury collisions; and 2) propose a quantitative model of predicting the potential crash risk among professional taxi drivers.

Table 2 presents the variables used in the regression model. Overall, there were high levels of self-reported workload due to long hours of work (M = 9.385, SE = 1.23), inadequate rest (M = 0.325, SE = 0.516), and high management fees (M = 166.74, SE = 17.93). A more detailed examination of responses to workload-related categories revealed that 76.37% of the sample reported having to work 9 h or more per day; 16.67% admitted having to work 8 h per day, and only 6.96% indicated working 7 h or fewer daily. Additionally, more than two-thirds (69.83%) of respondents worked 7 days per week; only 30.17% reported having time to rest. On the other hand, 64.56% were charged 151~ 180 CHY per day as a daily management fee, and another 20.68% were charged 180~ 210 CHY every day.

Moreover, respondents reported that they drove their taxi while sleepy or fatigued more than “quite often” but less than “frequently” (M = 3.272, SE = 1.270), followed by showing annoyance (M = 2.611, SE = 1.083), speeding (M = 2.494, SE = 1.312), running a red light (M = 2.241, SE = 1.359), parking at will to pick up/drop off passengers (M = 2.210, SE = 0.871) and overtaking other drivers under dangerous conditions (M = 2.177, SE = 1.254). Mobile phone use while driving was rated the lowest of all items (M = 1.082, SE = 1.018), consistent with findings from Wang et al. [14].

With regard to taxi drivers’ involvement in crashes during the past 3 years, respondents reported having been involved in 0.308 injury crashes (SE = 0.627). Among the 948 respondents, 17.83% admitted having been involved in one PIN collision and only 5.70% experienced two or more PIN collisions.

The mean of PIN collisions is 0.308 and the variance is 0.392, and these data therefore contain a small amount of systematic variation. Then a Negative-Binomial regression model was introduced to examine the relationship between taxi drivers’ sociodemographic information (age, gender, and educational background), work-related stress (hours of work per day, off-duty days per week, and daily management fees), frequency of daily risky driving behaviors, and PIN collision involvement, as Table 3 shows.

In Model I, all 14 self-reported variables were collectively entered to determine the predictor of PIN crash rate of taxi drivers, and 7 risky driving behavior variables except parking at will (coef. = 0.025, p = 0.915) were statistically significant predictors of PIN crash risk with all p < 0.05. For simplicity, the significant variables identified in Model I were together collected to determine Model II to predict the PIN crash risk of taxi drivers. Clearly, all risky driving behavior variables were again found to be significant. Compared with Model I, Model II had the better performance of fit (AIC = 755.0 vs. 764.1), but with fewer input independent variables (14 vs. 7).

As mentioned in Model II, positive coefficient is an indication of high risk for injury crashes [15], and thus the frequent behaviors while driving, including running a red light (coef. = 0.521, SE = 0.065), speeding (coef. = 0.241, SE = 0.074), showing annoyance (coef. = 0.154, SE = 0.068), aggressive driving (coef. = 0.276, SE = 0.069), driving while fatigued (coef. = 0.313, SE = 0.031), mobile phone use (coef. = 0.220, SE = 0.057) and dangerous overtaking (coef. = 0.245, SE = 0.055), were identified as the risk factors contributing to PIN collisions among taxi drivers with all p < 0.05, which were consistent with the findings from Wang et al. [15]. These findings should be recommended to their respective taxi companies for occupational safety protection.

Since drivers who drive for work purpose are known to be more prone to a wide range of risks [16], and thus such professional drivers have been reported to have above average crash frequencies compared to personal vehicle drivers as reported in the literature [17]. In Britain, 31% of fatal crashes and 26% of serious injury crashes involve occupational drivers [18]. A survey in Xi’an, China revealed that the average number of taxi crashes per month was 8.500 over the period of 2006 to 2011 [4]. Our previous study also showed that 47.34% of the occupational bus, taxi, lorry, company car and shuttle drivers in Xining, China had caused crashes at some point [19]. Accordingly, it is quite necessary to identify the potential crash risk for these professional drivers.

This questionnaire study tries to investigate for the first time the association between work-related stress, risky driving behaviors and involvement in personal injury crashes using the self-reported data from taxi drivers in Xi’an, China and then determine their predictor of such potential crash risk. Unsurprisingly, taxi drivers aged between 30 and 40 were more likely to be involved in collisions while driving. Because drivers often had young children and elderly parents to care for, they felt compelled to use their time more efficiently by driving faster and transporting more passengers in order to earn more money within a limited time window. Similar findings were reported among taxi drivers in Shanghai and New York City [20].

Additionally, the results found that more than three-quarters of participants had to drive 9 h or more per day, and 18.56% sat behind the wheel up to 11 h or more every day. More concerningly, 69.83% of respondents reported working all week without a day of rest. About three-quarters (74.57%) admitted to drive while sleepy or fatigued “quite often” or more frequently than that. On the other hand, approximately 85% had to pay a daily management fee of 150 CHY to their respective taxi companies, equal to nearly half their total daily gross income; therefore, it is unsurprising that the taxi drivers surveyed engaged frequently in risky behaviors such as speeding, running red lights, overtaking other drivers dangerously, and so on. Since it is impossible to cancel the daily management fee charged by taxi companies, it is strongly demanded to be reduced to some extent.

The results also provide insight into how to prevent and reduce traffic crashes among taxi drivers, namely by increasing their income. Taxi drivers could also be offered taxi ownership and operating rights under the centralized administration of taxi companies. It is worth noting that the government should play a major role in guiding reform to change the taxi industry’s present managerial approach; for example, tax incentives could be given to taxi companies to help subsidize employee benefits (e.g., medical insurance, accident insurance, paid leave, etc.). Taxi companies are strongly suggested to set up special funds to help those engaged in injury crashes. Furthermore, both the government and taxi companies should develop stricter rules and regulations to limit registered drivers’ maximum daily and weekly hours of work, which may help to prevent fatigued driving [21, 22].

On the other hand, taxi drivers should receive education about the dangers of risky driving behaviors to encourage them to obey the traffic rules and avoid serious traffic violations [23], such as speeding and disregarding red lights, etc. Those who disobey the rules should be harshly penalized or have their taxi license revoked. Moreover, employers should be required to improve taxi drivers’ working conditions, including through the establishment of a trade union, purchase of insurance, periodic training and safety education, and mental health exams for all registered taxi drivers. It would also be worthwhile to open a complaint centre through which passengers and drivers could submit grievances and resolve disputes about service problems; a coordinated problem resolution procedure may help to reduce taxi drivers’ risky driving behaviors.

Professional drivers in the world confront the similar excessive workload conditions and considerably high crash risk. A survey in Australia showed that occupational drivers had a higher intention to speed in a work vehicle than their personal one [24], and thus were more likely to have crashes while driving for work. In Sri Lanka, taxi driver’s sociodemographic factors (i.e. level of education, marital status) are also found to be significantly associated with their aggressive driving and risk-taking behaviors [25]. 81% of taxi drivers in Ankara, Turkey reported talking by using hand-held phone while driving [26]. On the other hand, many previous studies reported in the literature show that professional driver’s crash involvement is significantly related to the sociodemographic factors, working conditions or risky driving behaviors [10, 15, 27‐29]; therefore, the proposed quantitative approach using all these possible contributory factors can be used to examine the potential crash risk for professional drivers (i.e., taxi, bus, coach, truck) in China, European countries and other areas in the world [15, 30]. The findings of this study provide valuable data with important implications for worldwide decision-making in public health policy and traffic regulation for professional drivers.