The fear is understandable and sometimes justified. Productivity means producing more per worker, so if output stays the same, fewer workers are needed and some lose their jobs. That is exactly what happened in agriculture and manufacturing, where mechanisation reduced employment enormously over the twentieth century. The loss falls on identifiable people, often in particular regions and occupations, and those workers frequently cannot move to the new jobs because their skills do not transfer — the factor immobility of Topic 2. For them the harm is real and lasting.
But the conclusion does not generalise, for several reasons.
Output does not usually stay the same. Higher productivity lowers the cost of production, which lowers prices, which raises the quantity demanded. A firm producing more per worker often produces more in total, so employment may not fall at all. Where demand for the product is elastic, it can rise.
Productivity is the only sustainable source of higher wages. A firm can pay more per worker only if each worker produces more — otherwise the rise must come from prices or from profit, and neither can be repeated indefinitely. Historically, real wages have risen almost exactly in line with productivity, which means the alternative to rising productivity is not secure jobs at the same pay but jobs whose pay cannot rise.
The savings are spent elsewhere. Lower prices leave consumers with money to spend on other things, creating demand and employment in other industries. The jobs lost are visible and concentrated; the jobs created are dispersed and hard to attribute, which is why the losses dominate the public argument even when the gains are larger.
Work becomes less arduous. Much of what productivity growth has replaced is heavy, dangerous or tedious, and the jobs that replaced it are generally safer and better paid.
My judgement is that rising productivity is good for workers in aggregate and over time, and genuinely harmful to particular workers in the short run — and that both halves have to be stated, because each alone is misleading. Without productivity growth, real wages cannot rise at all; with it, some people lose jobs they cannot easily replace.
That is precisely why the policy response matters more than the phenomenon. The gains are large enough to compensate the losers several times over, but nothing automatic does the compensating — so retraining, relocation support and income protection are what determine whether displaced workers share in a gain that is real at the level of the economy and invisible to them. A country that lets productivity rise and does nothing for those it displaces produces exactly the resentment the raw figures cannot explain.