Social Safety Nets

The design problem

New Zealand's welfare system was built for a specific kind of unemployment: cyclical, temporary, contained within particular industries. You lost your job, the benefit bridged you to the next one, and the economy eventually absorbed you back.

That model assumed the economy would keep generating roughly similar numbers of jobs, even as the mix shifted. For most of the twentieth century, that assumption held.

AI-driven automation poses a different kind of challenge. Displacement may be structural rather than cyclical — not a gap between two jobs, but a permanent reduction in the number of jobs of a certain type. The radiologist, the paralegal, the customer service team: if AI can do most of what they do at a fraction of the cost, the jobs don't come back when the recession ends. They're just gone.

That's not certain. Economists disagree sharply about the net employment effects of AI. Some point to historical precedent — every wave of automation has eventually created more jobs than it destroyed. Others argue this time is different in speed and breadth. The honest answer is: we don't know yet.

But that uncertainty is itself a policy challenge. How do you design a safety net for a risk whose magnitude you can't predict?

Different traditions, different answers

This is where politics enters. The question of what society owes people whose skills have been displaced by technology is not a technical question — it's a values question. And New Zealand's political parties answer it differently.

A market-liberal tradition emphasises targeted, time-limited support that preserves work incentives, with the expectation that the market will create new opportunities. A social-democratic tradition emphasises universal entitlements and collective risk-sharing, with a stronger role for the state in managing transition. An indigenous tradition asks different questions entirely: about the relationship between people, land, and livelihood that doesn't reduce everything to employment.

This section doesn't adjudicate between these traditions. Instead, it maps the landscape of options across four domains — income, health, housing, and mental health — presenting what the evidence says about each approach and what trade-offs each involves.

The question for New Zealand: given who we are and what we value, which combination do we choose?

This section covers

• Income Support — From enhanced benefits to UBI: the spectrum of options for economic security
• Healthcare — AI transforming health delivery, and the health costs of economic disruption
• Housing — How displacement intersects with NZ's pre-existing housing crisis
• Mental Health — The psychological toll of workforce disruption, and options for response