Robot Tax: The Fix AI Displacement Actually Needs?
AI keeps rewriting the rules of work, and the usual policy answer—retrain displaced workers—keeps falling short. In an excerpt from his book Innovate for Impact: A Roadmap to Sustainable Technology Beyond AI, author Alessandro Crimi argues that retraining places the burden on individuals and can't keep pace with automation. Instead, he proposes an automation impact levy, or robot tax, to slow displacement and fund safety nets. The debate is messy, but it's forcing a bigger question: who pays when machines replace people?
Why Retraining Alone Doesn't Cut It
Retraining programs are necessary but insufficient, according to Crimi. They put the full weight of adaptation on workers and often fail to match the speed of change. History backs him up: even in medieval times, guilds, churches, and monasteries trained workers for innovations like windmills and watermills, yet living standards for most people stayed low. With AI, promised benefits like new tasks and productivity gains are neither automatic nor equally shared. That's why Crimi says the conversation must expand to profit sharing, taxation, safety nets, and policies like a four-day workweek or universal basic income—structural shifts that change economic relationships rather than just patching them.
What a Robot Tax Would Actually Do
A robot tax—or more precisely, an automation impact levy—targets a market failure. When a firm automates, it keeps the savings from a smaller wage bill while society absorbs the costs of unemployment and community decline. The tax aims to slow automation to a socially manageable pace and generate revenue for transition policies like UBI. Experimental evidence supports its mechanistic function, showing such a tax can reduce the probability of worker substitution. But critics argue defining the taxable unit is messy, since automation is often software integration rather than a discrete hardware purchase. Prominent tax scholars say reforming broader capital taxation would be more effective than a new targeted tax. Still, the debate exposes a governance gap: no standardized metrics exist to quantify automation-induced displacement.
South Korea's Quiet Experiment—and What Comes Next
South Korea offers the closest real-world test case, though it never called it a robot tax. In 2017, the Moon Jae-in administration scaled back tax credits for automation investments: large firms saw deductions fall from 3% to 1%, mid-sized firms from 5% to 3%, while small businesses kept their 7% benefit. It was less a penalty than a subsidy withdrawal, sidestepping the thorny definition problem. Meanwhile, the European Parliament's 2017 robot tax proposal was rejected over fears it would hurt business growth and innovation. Crimi argues such interventions are still necessary to reduce inequality and prevent techno-feudalism, where power concentrates among a few tech giants. The real question isn't just funding welfare—it's redefining contribution and entitlement in a post-labor growth model.
Key Takeaways
- Retraining is necessary but insufficient—it puts the burden on workers and can't match automation's speed.
- An automation impact levy aims to slow displacement and fund safety nets like UBI or a four-day workweek.
- South Korea's 2017 tax credit reduction is the closest real-world test of a robot tax, though it wasn't labeled as one.
- Critics say defining the taxable unit is messy and broader capital tax reform would work better.
- Without standardized metrics for automation-induced displacement, any tax risks being blunt or easily gamed.
Source: Rest of World • 🌍
Keep Reading
