Different Countries, Different Collapses (Part 2)

Part 1 argued that AI is arriving into societies where parts of the social contract were already fraying. The Hollow Middle will not appear everywhere in the same form.

Two dimensions explain where pressure travels first: visibility, meaning whether strain becomes publicly legible or remains private and indirect; and adjustment absorption, meaning whether the cost is carried by workers and households or mediated by firms and institutions. Mapping countries across these two dimensions produces four archetypes, each with a different signal pattern, a different risk, and a different political challenge. Countries can move between archetypes as AI adoption, labour-market pressure and institutional choices evolve.

Four archetypes:

Loud Fracture describes systems where pressure is visible and adjustment is individualised. The signal moves fast into layoffs, consumer sentiment, political anger, but visibility is no guarantee of the right response.

Private Absorption describes systems where pressure is real but carried quietly, by households, migrants, new entrants, or the informal economy, without surfacing as open political conflict. Firms can move fast, but the cost accumulates invisibly.

Negotiated Strain describes systems where pressure is public but adjustment is not left to individuals alone. Firms, unions, social insurance and public agencies all shape how the transition unfolds, and the risk is that bargaining becomes preservation rather than renewal.

Buffered Quiet describes systems that cushion adjustment well enough that deterioration can look manageable for too long. The danger is not rupture but a comfortable present making a smaller future politically acceptable.

Loud Fracture

High visibility · Workers and households absorb more

The United States is the clearest case of Loud Fracture. It is not necessarily the most distressed society in the global data. It is the place where the signal becomes visible quickly and where much of the adjustment is individualized. Recent US layoff announcements show how quickly AI investment can become part of the public labour-market narrative.

Loud Fracture is not only an American pattern. In India, AI could intensify the country’s job-creation challenge, especially as millions enter the workforce each year and services-led mobility becomes more exposed. South Africa is equally revealing: its official unemployment rate rose to 32.7% in the first quarter of 2026, making labour-market pressure impossible to hide. Jordan and Indonesia show fragility of the pathway AI may disrupt tomorrow with high share of informal economy, even where AI adoption is not yet the dominant cause.

Firms can reorganize roles, freeze hiring or reduce headcount relatively quickly. Workers carry more of the risk through job insecurity, weaker benefits, retraining costs, mobility costs and household financial pressure. That is why pressure can move rapidly from the labour market into consumer sentiment, political anger and public debate.

This visibility is both an advantage and a risk. The advantage is that the problem becomes harder to deny. The risk is that politics may reach for faster but weaker answers: protectionism, anti-immigration narratives, or pointing fingers to AI and robots. Redistribution and taxation matter especially if AI raises capital returns faster than wages, but they are not enough on their own. The harder task is to pair redistribution with worker transitions, care infrastructure, wage standards, training systems, school-to-work pathways and competition policy so that AI productivity gains do not simply accumulate at the top.

Loud Fracture guarantees audible stress. The central question is whether leaders use that visibility to rebuild the bargain around work, or whether they spend the political signal on faster narratives that leave the underlying fracture intact.

Private Absorption

Low visibility · Workers and households absorb more

Private Absorption describes systems where pressure is real, but much of it is carried privately by households, workers, migrants, new entrants or people outside strong protection systems. As a result, the signal is less likely to appear first as open political conflict or public labour-market rupture.

China sits closest to this archetype in the current map. Its AI transition is moving quickly and explicitly. Policymakers have framed AI as a productivity engine and a source of new jobs, while also acknowledging pressure on young people, graduates and migrant workers. While China is pushing society-wide AI adoption to create jobs and revive growth, the bet also raises the need for stronger welfare support if roles for workers are automated, reduced or reshaped. The pressure is already visible in indirect ways: household caution, savings behaviour, fertility decline, youth labour-market anxiety, cultural language and weaker consumption. AI adds a cleaner labour-market mechanism to that story. It also presents the policy dilemma: firms are being encouraged to innovate, but not simply to push the transition cost onto workers.

Vietnam, the UAE and Mexico show different versions of the same pattern. Vietnam combines rapid industrial upgrading with a more constrained public signal environment. With over 80% of the private-sector workforce in UAE made up of migrant workers on employer-tied visas, firms can restructure roles quietly, workers have almost no collective channel to contest adjustment publicly, making contract non-renewal or return migration the more likely pressure valve than visible labour conflict. In Mexico, informal employment already covers 55% of the workforce. It could absorb workers displaced from formal jobs by AI, but only at the cost of stripping them of social protections and removing them from official data. Pressure disappears from the headline numbers without being resolved.

That is the core opportunity and risk of Private Absorption. Because pressure is less publicly explosive, firms can move fast. They can redesign workflows, automate routine tasks, and experiment with new operating models before every adjustment becomes a national political fight. For business, that creates a real transformation advantage: as the unit of innovation, firms can move faster than in systems where every disruption is immediately contested.

But the same feature creates the risk. If households, young workers and migrants carry too much of the adjustment privately, weak signals can accumulate into a larger demand, trust and mobility problem. The system may look stable while workers delay marriage, save more, lower expectations, accept weaker jobs or withdraw from aspiration.

Private Absorption therefore has a sharper business implication than the name first suggests. It can enable faster firm-level transformation, but it also places more responsibility on employers. If firms use AI only to cut labour costs, the burden moves quietly onto households. If they use AI to redesign work, retrain workers, improve matching and build new entry pathways, Private Absorption can become the country archetype that captures the biggest transition advantage.

Negotiated Strain

High visibility · Firms and institutions absorb more

Negotiated Strain describes systems where pressure is public, but adjustment is not left only to individual workers or households. Firms, unions, social insurance, employment rules and public systems all shape how the transition unfolds.

In this archetype, AI becomes part of a broader social bargain. The pressure is already visible across several countries in this group. Spain is seeing early signs of AI-related adjustment: with declining employment in computer programming, telecommunications and information services, even as finance, another AI-exposed sector, continues to grow strongly. Korea points to AI pressure landing first on younger workers and hiring channels, while Brazil’s debate over reducing the working week shows how technology, productivity and worker protection can quickly become political questions about the future of work.

The opportunity is that these countries have bargaining machinery, using firms, unions, employers, public agencies and training systems to negotiate how AI is adopted, where productivity gains go, and how workers move into new roles. However, the risk is that negotiation becomes preservation. These systems can spend their political energy protecting existing jobs, existing rules and existing insiders, while younger workers face fewer entry routes and firms delay adoption. As a result, countries in this archetype may experience a slow loss of competitiveness, weaker productivity growth and a thinner middle-class ladder.

The unique battle in Negotiated Strain is therefore, how to make the bargain forward-looking. If these countries use social dialogue to accelerate reskilling and share productivity gains, they can turn visible pressure into managed renewal. If they use it mainly to defend the old settlement, they may keep the peace while losing the future.

Buffered Quiet

Low visibility · Firms and institutions absorb more

Buffered Quiet describes systems that many countries would normally envy. Firms are capable, institutions are trusted, social insurance is stronger, and adjustment is less likely to become a public rupture.

The advantage is real. These systems have more room to manage AI before it becomes a social crisis. Singapore is the clearest execution-speed example: it is investing heavily in public AI research through 2030, building AI talent pipelines, and expanding worker access to AI learning. Sweden also makes worker-training efforts focused on AI-exposed professional groups. In these systems, AI transition can be organized through firms, public investment and skills systems rather than only through household survival.

But that is also the danger. Because the system cushions adjustment, deterioration can look manageable for too long. Jobs may remain, but promotion slows. Tasks change, but career ladders narrow. Entry-level roles become thinner, but unemployment does not spike. The warning signs are already visible in the broader evidence: the ECB has argued that AI may be creating jobs in the euro area for now, but also noted that more than a quarter of German firms expect AI to lead to job cuts in the next five years. In Japan, AI may reduce labour shortages, but it can still absorb the routine and coordination tasks through which younger workers learn.

This is what makes Buffered Quiet different from its neighbours. Compared with Private Absorption, workers are less likely to carry the burden alone. Compared with Negotiated Strain, pressure is less likely to become immediate political conflict. That gives these countries an enviable advantage: they can act before crisis. But it also creates a specific risk: the absence of visible rupture can weaken urgency.

The unique tradeoff is between cushioning and renewal for countries in this archetype. Cushioning protects workers from abrupt shocks. Renewal rebuilds the pathways into good work. If these systems only cushion, they may preserve today’s comfort while allowing tomorrow’s ladder to thin. If they use trust, firm capability and public systems to redesign work early, they can turn AI adoption into a managed productivity transition rather than a delayed middle-class erosion.

Good news is, Buffered Quiet economies do not fail dramatically. However, they may fail when a comfortable present makes a smaller future politically acceptable.

Different collapses, different warnings

The mistake would be to wait for every country to produce an American-looking crisis. The Hollow Middle will not announce itself in the same language everywhere. Elsewhere, pressure may travel through quieter channels: delayed adulthood, cautious households, weaker consumption, thinner entry routes, overprotected insiders, underprotected outsiders, migration, informality or a slow retreat from aspiration.

These archetypes are therefore best read as warning systems. Countries can move between them as AI adoption, labour-market pressure and institutional choices evolve. The point is to ask where each society sends pressure when the middle begins to weaken, and whether leaders are looking in the right place before the pattern hardens.

AI will test countries through what they already do well. Loud systems may turn visibility into polarization. Fast-moving systems may turn execution speed into private household burden. Negotiated systems may turn bargaining into preservation. Buffered systems may turn cushioning into complacency. Each society has a familiar way of absorbing pain, and that familiarity is part of the danger.

The Hollow Middle may arrive as households expecting less, workers progressing less, firms offering less, and politics learning to contain frustration rather than resolve it. The final failure is wh