The economic cost of fatal workplace accidents in Sweden – A methodology for long-term decision analysis

1.1Causes of workplace accidents

Many of the studies have focused on work-environment and socio-cultural causes of accidents in the workplace, identifying a number of factors in- and outside the workplace that carry a high risk of accidents. A lack of safety programs has, for example, been cited as the cause of chemistry accidents in laboratories including, in one case, the death of a UCLA (University of California, Los Angeles) researcher [2].

Alcohol and drugs are known general factors that significantly increase the risk of accidents at work. A study of a large railway transport company in Portugal showed that random and unexpected testing for alcohol and drugs (A&D) in the workplace had a statistically significant preventive effect on safety and reduction of individuals’ accident risk, where the fraction of prevented accidents was 59% for on-board train workers, 72% for those working near trains, and 85% for white-collar workers [3].

Studies have also looked at internal risk factors of workplace accidents in different industries. For instance, a study of the cause of an increase in fatal accidents in Malaysia’s construction industry found unsafe methods, the unique nature of the industry, and job-site conditions to be the top three causes, with working at high elevations, incorrect work procedures, and structure failures as sub-causes [4]. However, the high number of fatal accidents and the causes in the Malaysian construction industry are also likely due to the lack of a safety culture and non-compliance with the country’s Occupational Safety and Health Act [5]. The impacts of human factors in fatal workplace accidents have also been studied in Brazil (for the period 2007 to 2011), where three organizational factors were found to be important determining risk factors – non-compliance with safety standards, deficiencies in assessing the work risk, and supervision failures [6]. Outsourcing has also been found to increase the risk of accidents [7] and, in a study of Mexican workers, socio-demographic factors (age, sex and occupation) as well as work environment and workplace conditions were all associated with fatal accidents in the workplace [8].

Studies of the causes of workplace accidents often lack an in-depth analysis of the conclusive risk factors. However, the work environment is a multidimensional reality [9, 10], and the research area should identify which of the psychosocial, ergonomic and physical risk factors at workplaces in each industry are most decisive in causing the accidents in question. This has been done, for example, for lost work time, work impairments, and labour productivity [11–16]. Crucial and suspected risk factors of workplace accidents include conflicts, violent behaviour, alienation, discrimination, stress, job insecurity and dissatisfaction, job demands, long work shifts, instability, handling heavy objects, and working in poorly lit environments. Some of the suspected work environment risk factors have also recently been studied, with one systematic review having shown extended working hours to be significantly associated with an increased risk of accidents [17], and another study showing higher suicide mortality rates among Korean workers [18]. Non-standard employment such as through a temporary agency and seasonal work was also associated with increased mortalities among workers [19], and the severity of workplace accidents despite an underreporting of severe injuries [20]. Further, emotional contagion in the form of anger and joy has also been associated with greater and fewer cognitive failures, respectively, which are linked to higher rates of subsequent workplace accidents [21]. Workplace accidents caused by workplace failures have also been shown to be associated with psychological distress [22]. In addition, a correlation has been shown between self-reported occupational accidents and employment conditions, job demands and workplace justice [23].

1.2The cost sources of accidents in the workplace

The costs identified in the workplace accident literature consist of direct administrative and medical costs, indirect income reduction and loss of productivity, and intangible human costs [24]. There are at present not many scientific research and commercial reports that attempt to estimate the costs of workplace accidents, and those that do exist are impeded by unclear cost sources and accident outcomes at the same time as using either poor or obscure methodologies [25]. Capturing these economic costs, however, is a must for making rational decisions regarding any investment to prevent accidents in the workplace. Certain, rational and effective decision-making in this context also requires an appropriate cost assessment methodology. The misreporting of these costs can lead to a knowledge gap or suggesting insufficient preventive measures.

There are, however, some studies in the relevant literature that have estimated the cost of workplace accidents. The Liberty Mutual Workplace Safety Index (WSI), for example, ranks the top ten causes of serious workplace injuries according to their direct cost to U.S. businesses. The 2019 WSI lists overexertion involving outside sources, same-level falls, and being struck by an object or equipment as the three most costly non-fatal injuries reported, at 13.11, 10.38 and 5.22 billion USD, respectively, while striking an object or equipment was the least costly, at 1.15 billion USD. With respect to non-fatal injuries, the WSI further states the construction sector to be the most costly industry, while the leisure and hospitality sector was the least costly [26].

In the case of accidents with fatal outcomes, the costs can be substantial if we include the indirect economic cost to society as a whole [25]. The total cost of all work-related injuries may also be much higher if we take into account long-latency disabilities, the causes of which are difficult to establish [27]. There is thus a potential to underreport causes (i.e. risk factors of workplace accidents) as well as to underestimate the cost of diseases. The European Agency for Safety and Health at Work conducted a review of studies evaluating the cost of work-related injuries, finding that a variety of methods and approaches were used to assess this cost [25].

There are challenges surrounding the approaches and methodologies used to measure the cost of workplace accidents before making decisions regarding safety and security in the workplace. The social costs associated with workplace accidents are usually estimated using one of two well-known cost models: accounting costs, and economic costs based on the human capital approach (HCA) [28]. The accounting cost estimation model includes all registered or measurable costs, no matter who in society pays. The main costs measured in this model include: administrative costs, social security and social insurance payments, healthcare costs, and friction costs for the affected firms [29]. A firm’s circumstances and status in the market can also come into crisis after the occurrence of work-related accidents, leading to an additional cost. An indirect economic or opportunity cost model based on HCA focuses instead on potential societal income losses [30, 31]. There are also intangible human costs associated with workplace accidents, such as physical pain and suffering, and the impact on an accident victim’s family [25]. These costs can dramatically decrease the quality of life and life expectancy of workplace accident survivors with permanent or temporary disabilities, further diminishing social welfare and sustainability. As mentioned, the human costs are intangible, which means they are difficult to identify, to measure and, especially, to value in monetary terms. Because the existing intangible costs associated with workplace accidents can be decisive, however, there is an indirect economic model – the willingness to pay (WTP) model – for valuation of decreased quality of life following non-fatal accidents at work [32, 33].

When conducting a cost assessment for economic decisions at the societal level, we should also add in the indirect costs of lost government tax revenues and employer friction costs to the direct accounting costs [30]. When it comes to workplace accidents, however, this cost assessment model contains a big uncertainty in that no one knows how many accidents initially listed as “non-fatal” may change to fatal or how many of the affected workers may become permanently disabled during the friction period. Both workplace accident outcomes and the length of the friction period are unknown. In practice, the cost model only entails a short-term assessment of social costs associated with workplace accidents. Conversely, the indirect economic cost assessment model takes a long-term societal perspective and does not need to add any other costs. However, in the human capital approach the definition of labour productivity contains both market productivity at work and non-market productivity at home, and optimally also at social enterprises engaged in during leisure time [30].

1.3Modelling economic evaluation of preventive measures

Because studies of workplace accidents often lack cost data about safety programs and accident-prevention measures, the current literature offers no models for decision-making and economic evaluations of alternative safety intervention programs, though models have been developed for work-related disorders [30, 31, 34–36]. Researchers in the field prefer to highlight the importance of minimizing the risk and costs of accidents, rather than focusing on the cost of implementing different measures to improve safety and security in the workplace.

1.4The importance of societal perspective in reports of workplace accidents and accident cost assessments

Scientific research in this area should include a societal perspective since the whole society pays for workplace accidents that occur in an industry. All potential costs incurred by all stakeholders, irrespective of who pays, should be considered. In order to do this, researchers need complete information about workplace accidents that have occurred – about the productivity of the affected workers at work and at home, in addition to how their families and surrounding community, government, and social organizations and enterprises are affected. Researchers and decision-makers also need information on the spaces, equipment and tasks with a high risk for accidents in the exposed industry [6] in order to suggest possible preventive measures and to assess their costs.

The socio-economic consequences of occupational accidents are characterized by long time horizons and a broad and diverse number of social areas [25]. The costs of workplace accidents, especially those with fatal outcomes, are thus allocated across time and across society through different stakeholders at different levels. Policy-makers should therefore know that, even if they frequently require costs, measures undertaken to prevent accidents also generate utility over time via the potential human (non-economic) and economic cost savings – for both industry and society as a whole – of preventing accidents [7]. The human capital approach addresses the potential societal income lost through workplace accidents – an economic cost to society that could be saved by preventing the accidents in the first place.

1.5Fatal workplace accidents in Sweden

According to the Swedish Work Environment Authority, workplace accidents with fatal outcomes in Sweden for 2010–2019 had three top causes (AV/ISA). The first cause was lost control – primarily of vehicles, but also of machinery, tools, animals or other objects. The second top cause of fatal workplace accidents in the country was bursts, explosions and landslides or sliding objects, and the third falling or slipping. These three causes were responsible for 52%, 17% and 14%, respectively, of all fatal accidents in workplaces during the period noted.

When it comes the industries with occupational accidents leading to death – the construction, transport and agricultural sectors topped the list, representing 57% of all fatal workplace accidents in Sweden for the same period, 2010–2019. At the opposite end of the scale – hotels, restaurants, and the information and communication sectors were the safest workplaces in Sweden during the period.

In the period looked at in the current study using registered data and information by Swedish Work Environment Authority (2008–2019), 73 people who died in workplace accidents in Sweden (an average of 6 workers per year) were foreign workers working for foreign enterprises. The information for these workers is therefore registered in other countries and not included in our cost assessment model here.

1.6Objective

The purpose of this study was to introduce an appropriate economic cost assessment model that can be used prior to rational decision-making at the macro level regarding prevention of accidents at Swedish workplaces. The economic cost could therefore be assessed from a long-term societal perspective.

2.1The economic cost assessment model

The indirect economic cost (EC) of each fatal workplace accident is estimated based on the marginal revenue product of labour per month (MRP_L/M), which indicates the work function’s potential contribution to the industry’s output per month:

The deceased worker i would have increased national income during his/her effective working years (number of months j remaining until retirement) by:

The total economic cost of a fatal workplace accident (TEC_i), including productivity losses at work and at home, is thus estimated as the sum of EC_i and HPL_i:

2.2Adjusting for differential timing

As the economic costs of fatal workplace accidents are allocated across a period of several years that varies for each deceased worker, a decision to improve workplace safety and prevent accidents must consider the present value of the cost, estimated as:

4.1Internal and external causes of workplace accidents

Organizational factors in workplaces can lead to emotional reactions and stress, in turn leading to accidents at work. For instance, the ever-fluctuating and high quantity and quality demands on production made by profit-maximizing firms in the market as they battle for competitive advantage lead to high stress among workers [18, 23]. Swedish workplaces and the exposed workers cannot be isolated from Swedish society, however, so some causes and sub-causes of accidents at work are not due to the work environment but to individual, demographic and socio-cultural conditions that mainly develop in the workplace. An effective workplace safety program can thus not prevent all accidents and their related societal costs. As earlier studies have shown, human factors, safety culture and the misuse of alcohol and drugs, for example, have led to workplace accidents [3, 5, 6]. However, it is extremely difficult to determine the extent to which fatal work accidents in Sweden are caused by internal or external factors. When making economic decisions at the macro level regarding the prevention of workplace accidents, decision-makers should therefore consider all inside and outside environmental factors since non-work-related factors also affect people’s socio-economic functions and thus the organizational and social value of health and safety [8].

4.2The human capital approach and its limitations in the cost assessments

The assessment of the economic cost of fatal workplace accidents was based on the concept of human capital. Introduced by Schultz (1960) and further defined by Becker (1987), the concept of human capital encompasses potential forces and capabilities embodied in humans that lead to the creation of individual and social well-being [28]. Applying the concept of human capital substantially is a useful approach when assessing lost opportunities and potential productivity from a societal perspective, as it has the potential to consider determinants such as structures of the economy and its labour market. The human capital approach upon which the economic cost of fatal accidents in Sweden is assessed, however, is salary-based and assumes that, in a perfectly competitive labour market, wages should reflect the workers’ marginal contributions to a firm’s output. The equilibrium price of labour (salary), which is determined in the labour market, is accordingly equal to the marginal revenue product of labour (MRPL) [30]. There are at least two conditions that should be met for this equality. The first is the existence of equal job opportunities, and the second the existence of the “same salary for the same job” [30, 37]. In the case of monopsony and high unemployment, a worker’s salary is less than the MRPL because of workers’ low bargaining power [30, 37]. When a market fails, salaries can also be affected by the health status of workers and can vary between genders, and ethnic and age groups [37]. It is not clear, however, that the Swedish labour market is under perfect competition and free from discrimination. When assessing the MRPL on the basis of salaries, one must thus consider the state of the country’s labour market in terms of unemployment and market failures (monopsony power and discrimination) that lead to allocative inefficiencies (deadweight losses) [37].

In addition to the assumptions in the underlying economic theories, such as constant salary increases and stable inflation of around 2% and a perfectly competitive labour market, there are other limitations in the cost assessment model:

4.3Prior to an economic decision to prevent accidents at work

The reports of fatal accidents published by the Swedish Work Environment Authority did not contain all of the data and information needed to make economic decisions on preventing workplace accidents in Sweden. The first step of the decision-making process for a safe and secure workplace is to identify the spaces, equipment and tasks with a high risk for workplace accidents in high-risk industries. This type of data and information is useful in order to propose suitable preventive measures. In the case of alternative measures, the second step is to assess the cost and effectiveness of the measures, whereas the cost of a unique preventive measure should be compared to its opportunity cost (i.e. the economic cost of fatal workplace accidents) or the social utility of preventing accidents at work. For the former, the approaches of cost-effectiveness analysis (CEA), cost utility analysis (CUA) and cost-minimization analysis (CMA) can be applied. For the latter, social return on investment (SROI), socio-economic cost analysis (SECA) and cost benefit analysis (CBA) can be used. A calculation of willingness to pay (WTP) can also be used prior to evaluation of any preventive measure. The areas of application for these methodologies are as follows:

All of the approaches and methodologies described here have been successfully developed and applied in social sciences as well as occupational health economics for making decisions regarding implementation of proposed programs at different levels [32–38].

4.4Recommendations for future research

Occupational health economics as new developing science attempt to allocate the scarce resources to improve occupational health by providing rational decision-making models [9–11, 30, 31, 35, 36, 41]. The subject is thus in need of developing appropriate economic evaluation methodologies to assess future costs and benefits associated with alternative measures. The human capital approach used in this study is, however, a robust foundation for developing such methodologies. The economic cost of fatal accidents at Swedish workplace was assessed in this study using the approach. Economic decision-makings as well require an assessment of costs associated with different available preventive measures using the same approach. Studies of workplace accidents with an economic perspective should therefore be developed in this way while trying to model the employers’ efforts to improve the work environment [10].