[Research] A proven link between accident rates and companies' economic performance

Why investigate the potential link between a company's economic performance and its commitment to accident prevention ?

Over the last twenty years or so, various approaches to evaluating prevention policies in relation to economic data have been studied.  I could mention work on analysing the cost of accidents and the associated disruption factors: social costs, but also administrative, equipment or production costs. Other approaches seek to measure a 'return on investment' in prevention and use cost-benefit analysis as an argument to encourage companies to undertake occupational risk prevention.

The work carried out in this INRS study does not belong in these categories, but adopts an econometric methodology. The aim was to investigate the statistical link between companies' economic performance indicators and their frequency of accidents at work. The aim of this work was to respond to the concerns of all those involved in occupational health and safety (“Assurance maladie - Risques professionnels”, prevention and occupational health services, etc.) who wanted additional arguments to encourage companies to take more preventive measures.

The aim of this study was to carry out a large-scale statistical evaluation, across all sectors of activity, in order to provide scientific proof of the existence and importance of the link between companies' accident rates and their economic performance. The ultimate aim was to answer the question: "Can it be said that the more resources a company devotes to occupational risk prevention, the better its economic performance?”

A large-scale statistical evaluation... Can you tell us more about the scope of your study ?

The strength of this work lies in the large panel of companies surveyed and the time period covered, both of which are fairly unprecedented. Based on INSEE and CNAM databases, the data used concerned 1.977 million French companies covered by the general scheme, belonging to 83 professional branches, and monitored over all or part of the period from 2003 to 2017, i.e. 15 years. In all, more than 14 million records were analysed.

The model used by the INRS team related, for each company, economic performance indicators (productivity, profit) to production factors (capital, labour). The scientific contribution of this study is to have incorporated accident frequency and severity variables into the model.

What do you learn from analysing this data ? 

The main finding is that there is a significant negative correlation between the number of accidents and a company's economic performance. A 10% increase in the frequency of accidents at work reduces a company's productivity by 0.12% and its profit by 0.11% over the same year. And this effect is still very much present the following year.

A second lesson is that the significance of this effect is highly dependent on the size of the company. For companies with fewer than 20 employees, this 10% increase in the frequency of work-related accidents leads to a 0.38% reduction in productivity and a 0.24% reduction in profit. Whatever the size of the company, a work-related accident disrupts production and thus reduces productivity. This effect is particularly significant for small businesses, which have more constraints in terms of staff and equipment to deal quickly with the disruption caused by a work-related accident. Additional tests carried out on companies with more than 150 employees (which may have greater resources to invest in prevention) show that their profits are even affected over two to three years.

For the first time, these results provide scientific validation of the existence and importance of the link between claims experience and a company's economic performance. "The lower the claims rate, the higher the company's economic performance".
 
Over and above the employer's legal obligation to protect the health and safety of workers, this study provides an argument in favour of implementing occupational risk prevention measures in companies.