The study shows that most of the people view it more favorably when workers are replaced by other people than by robots or intelligent software.
This preference reverses, however, when it comes to people's own jobs. When that is the case, the majority of workers find it less upsetting to see their own jobs going to robots than to other employees.
In the long term, however, the same people see machines as more threatening to their future role in the workforce. These effects can also be observed among people who have recently become unemployed.
Over the coming decades, millions of jobs will be threatened by robotics and artificial intelligence. Despite intensive academic debate on these developments, there has been a little study on how workers react to being replaced through technology.
Researchers were able to identify the causes behind these seemingly paradoxical results, too. People tend to compare themselves less with machines than with other people. Consequently, being replaced by a robot or software poses less of a threat to their feeling of self-worth.
This reduced self-threat could even be observed when participants assumed that they were being replaced by other employees who relied on technological abilities such as artificial intelligence in their work.
"Even when unemployment results from the introduction of new technologies, people still judge it in a social context," said Christoph Fuchs, a professor of TUM School of Management, one of the authors of the study.
Fuchs continued "It is important to understand these psychological effects when trying to manage the massive changes in the working world to minimize disruptions in society."
"For people who have lost their job to a robot, boosting their self-esteem will be less of a priority. In that case, it is more important to teach them new skills that will reduce their concerns about losing out to robots in the long term," Fuchs added.
The study could also serve as a starting point for further research on other economic topics.
Fuchs said "It is conceivable that employee representatives' responses to job losses attributed to automation will tend to be weaker than when other causes are involved, for example, outsourcing."