The debate on robots usually evolves around labour market outcomes. While economist have been occupied with the impact of robots on employment and wages for decades, there is another question which deserves just as much attention: can robots save lives and prevent workplace injuries?

Our recent study addresses this question by focusing on robot adoption in European industries over the period 2011–2019, using IFR data on industrial robots. We find that the increase in robot adoption is associated with a reduction in fatality and injury rates of about 4.3% and 3.2%, respectively. Importantly, this result is not just about substituting workers. Since we account for changes in both employment levels and occupational composition, the result suggests that industrial robots take over some hazardous tasks, changing the content of work rather than simply reducing headcount.

This is the first cross-country study on robots and workplace safety, which is especially important in a highly heterogeneous context such as the EU. Countries differ in sectoral specialisation, labour market institutions, occupational safety regulations and reporting practices. Industries, in turn, differ in their technological capabilities, organisation of production and exposure to risk. Considering different EU countries, rather than focusing on one alone, allows to assess not only whether industrial robots improve safety, but also under which conditions they do so.

How robots may improve workplace health and safety

The main intuition behind this positive result is straightforward. In many production stages, industrial robots can substitute workers in the most dangerous, repetitive and physically demanding tasks. Thereby reducing exposure to hazardous materials, heavy loads, or extreme heat. They may as well improve ergonomics, lowering the probability of accidents caused by fatigue, malfunction or poor coordination. In this sense, automation is not only a matter of increasing efficiency, but can help prevent workplace injuries.

Nevertheless, ‘more robots, fewer accidents’ is far from self-evident. Integrating industrial robots into a production process does not guarantee safer workplace per se. New machinery requires training, redesign of workflows, clear communication, and safe worker–machine interaction. Without complementary organisational changes, automation may bring few benefits or even create new risks. This is important because it shifts the focus away from technology alone, towards the broader organisational and institutional environment in which it is adopted. Our findings point in this direction.

Asymmetric impact: industries are different

The positive impact from robotisation is not evenly distributed across industries: it is concentrated in technologically advanced sectors (e.g., electronics, pharmaceuticals, machinery). In these sectors, the reduction of fatalities and injuries is substantially stronger. By contrast, in more low-tech manufacturing activities, where most of the workplace injuries and fatalities take place, safety benefits do not materialise.

A plausible explanation for this asymmetric impact is that firms in high-tech sectors, on average, tend to compete through quality, innovation, skills and the accumulation of capabilities (i.e., technological competitiveness). In these sectors, therefore, the adoption of industrial robots is more likely to be embedded in broader processes of organisational upgrading. This is less likely to be the case in more traditional sectors, where competition often relies more heavily on cost reduction.

Asymmetric impact: institutions matter

The reduction of fatal and non-fatal workplace injuries associated with robotisation are stronger in countries where workers have more voice inside the firm (i.e., where works councils enjoy co-determination rights), such as Austria and Germany. Conversely, where worker representation is weaker and works councils have only consultation rights, the safety effect disappears.

This finding suggests that institutions do not simply react to technological change; they may actively shape its direction and trajectory. After all, when workers are involved in the decision-making processes, including decisions regarding which technologies are introduced, for what purpose and in which context, then the expected effect is likely to be more labour friendly. Put differently, institutions help determine whether automation becomes a tool for safer production or just another device for intensifying and fragmenting work.

Beyond techno-optimism and -pessimism

Too often, the debate on automation is framed in binary terms: either industrial robots are good (because they raise productivity), or they are bad (because they displace labour). Our results challenge this, suggesting that industrial robots can improve working conditions, at least when it comes to physical health and safety. However, these are not to be generalised: only some forms of automation, i.e., industrial robots in manufacturing, under some conditions, in high-tech sectors and in countries with labour-friendly institutions, reduce workplace injuries and fatalities. However, even where they improve physical safety, they may still reshape work in ways that increase pressure and negatively affect workers’ well-being. Other technologies, such as algorithmic management systems and digital monitoring, may have even more ambiguous effects, raising stress, intensifying work and reducing worker autonomy. The point, then, is not whether automation is good or bad, but how it is implemented, who governs it and whose interests it serves.

Co-authors

• Dario Guarascio, Associate Professor, Sapienza University of Rome
• Marco De Simone, PhD Candidate in Economics, Sapienza University of Rome

Full paper

De Simone, M., Guarascio, D., & Reljic, J. (2025). Do Robots Save Lives and Prevent Workplace Injuries? SSRN Working Paper.

References

• Dosi, G., & Virgillito, M. E. (2019). Whither the evolution of the contemporary social fabric? New technologies and old socio‐economic trends. International Labour Review, 158(4), 593-625.
• Abeliansky, A. L., Beulmann, M., & Prettner, K. (2024). Are they coming for us? Industrial robots and the mental health of workers. Research Policy, 53(3), 104956.