Hollowing Out

The image is of a tree that looks solid but is empty inside. In labour markets, hollowing out means employment growth concentrates at the two ends, rising for highly paid professional work and for lower-paid manual services, while routine middle occupations shrink.

Economists call the pattern job polarisation, and research across advanced economies documents it over several decades. Two forces drive most of the change.

Automation replaces routine tasks that sit in the middle of the wage distribution, such as clerical processing and repetitive production work. Offshoring moves whole categories of production to lower-cost locations, removing the domestic plants and supplier networks that once anchored middle-wage employment.

The consequences reach beyond the workers directly displaced. Middle-skill jobs historically provided the ladder from low income into stable prosperity.

When the middle rungs thin, workers either move up into roles that demand new credentials or move down into lower-paid service work, and communities built around the lost industries can decline for a generation. The second usage concerns national industrial capacity.

A country that outsources production may keep design and branding while losing the factories, tooling and process knowledge underneath. Critics warn that this hollowing out eventually erodes the ability to innovate, because product knowledge often lives close to production.

Not all the news is decline. The same forces raise productivity and lower consumer prices, and new middle occupations do emerge in care, logistics and technical services.

The policy debate is about speed and distribution: gains arrive broadly and slowly, while losses hit specific places and occupations at once. For a manager, hollowing out appears inside firms too: automating the routine middle of a process leaves a small expert tier and a large basic tier, changing hiring, training paths and wage structures in ways that need deliberate design rather than drift.

Measurement shapes what we conclude, since studies using wages find stronger polarisation than some using hours worked, and results differ between countries with different labour institutions, suggesting policy can blunt the pattern. Comparative OECD work shows the middle shrank almost everywhere but at very different speeds, which is evidence that technology set the direction while institutions set the pace.