A new NBER working paper by Joshua S. Gans explores how automation can harm workers even without replacing them. The study focuses on jobs where workers value both producing useful output and knowing their contribution matters. When a credible machine alternative exists, it weakens the meaning derived from personal contribution, even if the worker stays employed.

Gans's model shows that this loss of meaning raises compensation when wages adjust fully. However, if wages adjust only partially, workers bear some of the loss themselves. The paper also finds that an external developer may profit by publicly demonstrating a machine before licensing it, because the demonstration lowers the perceived value of the human alternative.

This creates a "meaning externality" that can drive demand for automation and make profitable development socially harmful. The effects differ based on the machine's quality versus its public salience: better technical quality improves output, while greater salience alone weakens human work's value. Thus, automation can reduce the value of work before it eliminates jobs.

The paper, titled "Replaceable but Employed: Automation and the Meaning of Work," is available as NBER Working Paper 35559, published in July 2026, with DOI 10.3386/w35559.