Risk of AI Investment in Skill-Deficient Workforces of Emerging Economies
Key Findings
Job Loss From AI
Heavy AI investment causes lasting joblessness in developing countries because weak retraining systems prevent workers from moving into new sectors.
When a developing country invests heavily in AI and automation, the gains often go to a few capital-rich sectors. At the same time, many workers lose their jobs in traditional, labor-heavy industries. This happens especially when the country lacks strong social safety nets and job retraining programs. India’s IT boom in the early 2000s is a clear example. The IT sector grew fast but hired very few people. Less than one percent of the workforce joined it. Most workers stayed in farming or informal jobs with few transferable skills. Without government-supported training, they could not move into new sectors. As a result, joblessness rose in traditional manufacturing. The problem is not that machines replace workers. It is that no systems exist to retrain or place those workers elsewhere. This mismatch causes lasting unemployment. In most developing countries today, such support systems are missing. Therefore, large-scale job losses are likely and built into the system. They can only be avoided with major investment in helping workers transition.
AI And Jobs
AI causes job disruption, but historical patterns show institutions respond with training and policy, leading to workforce realignment and preventing lasting unemployment.
Historical changes show economies adapt when technology disrupts work. Rapid tech adoption leads to new job training and support systems. These develop over time, not overnight. Japan and South Korea used public training and tax incentives after industrial automation. Policy responses reduced job loss and eased worker shifts. Displaced workers found new roles within about ten years. Similar patterns appear in many industrializing economies. State and business coordination helps reabsorb workers. Today’s AI advances cause temporary job mismatches. But strong institutional responses prevent lasting unemployment. Such learning has succeeded in most past shifts. Permanent job loss happens only when these responses fail. That failure is rare in comparable cases. The risk of long-term unemployment is low if institutions adapt. The past shows this pattern clearly. Preparation matters, but panic does not. The system adjusts through proven steps. The result is not job loss but job change. History supports this outcome. The process takes about a decade. No pre-existing institutions guarantee success. They form in response to disruption. The key is timely policy action. Institutions learn from past transitions. Most countries have managed this well. The same path remains open today. Adjustment is hard but manageable. Past patterns suggest a positive result.
Deeper Analysis
Could widespread unemployment from AI and automation be avoided in emerging economies if informal sector growth absorbs displaced workers, even without formal retraining programs?
Informal Jobs Trap
Informal work cannot absorb displaced workers effectively because once it exceeds 85% of non-farm labor, further inflows sharply depress earnings and resilience, making it a congestion point rather than a dynamic buffer.
Many workers in poor countries lose their formal jobs to AI and automation. They then move into informal work like street vending or micro-enterprises. This shift only helps if informal businesses can grow and pay decent wages. But World Bank and ILO data show most informal jobs are barely survival-level. These small firms reinvest little and never raise productivity. So the informal sector does not absorb workers well. It just becomes a crowded holding area for surplus labor. Low barriers to entry hide real unemployment. Official job numbers look fine, but earnings and job quality crash. The main flaw in this argument is the unseen limit. Historical data from Southeast Asia and Africa show a clear pattern. Once informal work exceeds 85% of non-farm labor, extra workers heavily cut average pay and weaken household stability. Therefore, expanding informal work does not prevent real economic distress.
Informal Job Growth
Widespread job loss from automation may not raise unemployment in developing countries because informal work offers immediate alternatives where oversight is weak and demand for low-cost services remains.
In many developing countries, most people who work outside farming are in the informal economy. This sector is easy to enter and does not require special skills or licenses. When jobs are lost to AI or automation, displaced workers can often find quick alternatives in informal work. This happens because rules are weak and demand remains strong for cheap services. People start small businesses or take odd jobs to survive. There is no need for government programs to make this happen. In Africa and South Asia, over 80% of new non-farm jobs are informal. These jobs appear without official planning or training. As a result, high unemployment may not follow job losses caused by automation. The informal economy absorbs the shock.
Bank Lending Limits Job Shifts
Labor cannot easily shift after automation in developing economies because bank lending habits limit long-term investments in worker retraining and formal job growth.
In many developing countries, banks dominate finance. They lend mostly to big firms. When machines replace workers, people need new jobs. But banks do not lend for worker retraining. They prefer short-term loans with collateral. This habit comes from economic instability and profit motives. As a result, workers cannot easily move into new roles. Formal job markets grow slowly. Even with government efforts, job creation lags. Latin America shows this pattern since the 1980s. Workers end up in informal jobs. This is not due to poor cooperation between state and workers. The root cause is weak financial systems. Development banks in East Asia did better. They funded upgrades and worker transitions. There, job shifts were smoother. In most emerging economies today, finance blocks labor change.
Job Trap In Informal Work
AI and automation deepen job inequality in developing economies by pushing displaced workers into low-productivity informal work instead of causing official unemployment.
In many developing countries, most people work in informal jobs without contracts or benefits. Labor markets are largely unregulated, and governments struggle to enforce basic workplace standards. When AI and automation replace workers, the job losses do not show up in official unemployment numbers. This is because displaced workers often move into informal work. These jobs are unstable and offer low pay and no chance for career growth. The informal economy acts less like a safety net and more like a trap. It keeps people working but blocks their move into skilled, better-paying roles. Evidence from Sub-Saharan Africa and South Asia shows over 80% of workers are in informal jobs. New digital investments have not been paired with programs to help workers develop new skills. Without such support, people cannot adapt to technological change. As a result, automation pushes workers into low-productivity work rather than removing them from the job market entirely. This maintains the appearance of steady employment. But it weakens long-term growth in skills and wages. The absorption of workers into informal work prevents true economic progress.
Explore further:
- Would the observed financial constraints on labor reallocation still dominate if foreign direct investment in AI infrastructure bypassed domestic banks and directly funded workforce reskilling through international development partnerships?
- What conditions would cause the informal sector to fail as an absorber, converting underemployment into open unemployment and social unrest?
What happens in emerging economies where political institutions lack the capacity or incentive to coordinate with capital and labor during AI-driven displacement?
AI Job Losses
AI displaces workers in developing economies because weak institutions cannot coordinate training, wages, and investment to absorb job losses.
In many developing countries, governments lack the power to manage the shift caused by AI replacing workers. Historically, state action helped move displaced workers into new jobs. This worked in countries like Japan and South Korea because strong institutions coordinated training, wages, and industry needs. Those governments invested wisely and ensured costs were shared fairly. Without such institutions, AI disruption brings no coordinated response. Instead, more workers slip into informal jobs and inequality grows. The harm isn't just temporary dislocation. It creates a lasting divide between a small group in high-tech firms and the rest with few opportunities. Past industrial shifts succeeded because governments helped absorb shocks. Where that capacity is missing, AI leads to lasting joblessness. The problem is not the technology itself but the failure to coordinate policy and investment.
Informal Workers Under Pressure
Informal job markets cannot absorb displaced workers when automation targets the same low-cost services those workers rely on, because machine-based platforms can offer identical services at much lower prices, driving informal workers out of the market.
In many developing countries, people who lose jobs often end up in informal work. These jobs usually involve simple services or goods that require little money to start. But artificial intelligence and automation are now replacing jobs in exactly those areas. Sectors like food service, delivery, and basic repairs are common targets for automation. When companies use automated systems, they can offer services at very low cost. This undercuts informal workers who rely on slightly higher prices to survive. Data from India and Brazil shows that informal markets shrink when formal businesses automate nearby services. For example, in Southeast Asia, ride-hailing apps replaced many informal taxi drivers. Those workers did not find new jobs. Instead, their incomes dropped and many left the workforce. The key problem is that automation hits the same low-cost areas where informal workers survive. When machines offer the same service for almost nothing, people cannot compete. This means the informal economy cannot absorb job losses if automation removes its foundation. The idea that informal jobs will naturally take up the slack fails when technology disrupts the very markets informal workers depend on.
Political Capture Blocks Re-skilling
Political capture in emerging economies blocks the re-skilling loop, so AI-driven job loss creates permanent unemployment instead of a temporary spike.
The main argument depends on a risky assumption. It assumes emerging economies can copy Japan or South Korea. Those countries used strong state-business coordination after war. But this fails when ruling elites control political institutions. Brazil after its 1990s trade reform shows the pattern. Unemployment and informal work stayed high for a long time. The state could not tax protected industries for retraining. It lacked both the incentive and the independence to act. This blocked the normal cycle of re-skilling. Two problems work together. First, without a professional state bureaucracy, capital prefers cheap labor or automation. Second, labor has no power to demand new training. The result is a split labor market with a permanent surplus. The conclusion is clear. In such economies, AI-driven job loss is not temporary. It creates a lasting floor of structural unemployment. The needed cycle of mutual concessions never starts.
AI In Oligarchies
AI adoption in emerging economies reinforces oligarchic control because dominant firms use automation to cut labor's share and strengthen monopolies, weakening demand and locking growth into exclusive, capital-driven patterns.
In emerging economies, powerful business groups and state-connected firms control most investment. They adopt AI and automation not to boost broad productivity but to strengthen their market power. This shift favors control and consolidation over fair labor practices. International Monetary Fund reports from Southeast Asia and Latin America show rising digital spending alongside falling wages and growing wealth in few hands. The reason lies in how big firms operate. They use technology to cut reliance on workers, not because workers are unskilled but because weaker labor reduces pressure for higher pay. These firms thrive where rules are weak and enforcement lacks teeth. Automation thus deepens monopoly power instead of creating new opportunities. It does not split the economy into formal and informal sectors. Instead, it entrenches existing control within the formal market. The growth of informal work is not the cause but a sign of deeper issues. Workers remain excluded not due to lack of training but due to power imbalances. Political systems that fail to balance capital and labor allow this cycle to persist. The result is not mass joblessness but a steady decline in workers’ share of economic value. This weakens overall demand and steers growth toward capital-heavy paths that leave most people behind.
AI Job Displacement
AI-driven automation leads to lasting unemployment in countries where weak institutions fail to coordinate timely worker retraining and job placement.
When political institutions fail to coordinate capital and labor during AI-driven job losses, unemployment rises sharply. This happens not just because workers lack skills. The real issue is the delay between new technology and worker retraining. Brazil in the 1980s and 1990s saw this clearly. Weak labor systems and inconsistent government action slowed retraining. Workers shifted into informal jobs instead of new roles. Japan followed a different path. Strong ties between firms, workers, and the state enabled large-scale training. Job rotation and upskilling kept workers employed. The key difference is durable cooperation between government, business, and labor groups. These structures turn job losses into managed transitions. Without them, job displacement breaks worker careers. Unemployment then lasts far longer than a simple transition period. In emerging economies, weak institutions cannot align business and worker interests. AI automation in these places causes lasting joblessness.
Automation Without Safety Nets
Automation causes permanent job loss in countries without strong social and tax systems to manage worker retraining and support.
Many middle-income countries cannot fund large-scale job retraining or support unemployed workers. Their tax systems are weak and labor rules are poorly enforced. Without strong government coordination, automation does not cause a short-term job shift. It splits the workforce permanently. One part joins high-productivity jobs that use AI. The other fills low-skill informal jobs with no retraining. This happened across Latin America in the 1990s. Machines replaced workers. The state failed to coordinate retraining or social support. Displaced workers moved into informal services. These jobs paid less and offered no security. Inequality grew and never reversed. Employment rates did not recover. When governments lack tools like progressive taxes and job training, job losses from automation become permanent. The shock does not pass. It becomes lasting unemployment.
Explore further:
- Could strong external actors, like multinational corporations or international institutions, substitute for weak state capacity in coordinating labor and capital during AI-driven disruption in emerging economies?
- What would need to be true about the informal sector's internal structure for it to retain any capacity to absorb displaced workers despite platform automation's price collapse?
- What happens in emerging economies where decentralized tech hubs or foreign-led education initiatives bypass captured states to deliver AI-relevant skills training independently?
- Under what conditions would organized labor or small firms successfully resist the rent-seeking use of AI by dominant conglomerates in emerging economies?
Would the observed financial constraints on labor reallocation still dominate if foreign direct investment in AI infrastructure bypassed domestic banks and directly funded workforce reskilling through international development partnerships?
Foreign Aid For Training
Foreign funding for AI and training succeeds only when recipient countries already have strong educational systems capable of scaling reskilling, making prior state investment in human capital the decisive condition for effective labor adjustment.
Development finance often funds both job training and AI infrastructure through partnerships. This shifts the main problem from a lack of domestic credit to a mismatch between foreign goals and local labor needs. For example, World Bank training loans worked well in East Asia during the 1990s. Those countries already had strong technical schools that could expand quickly. The key mechanism is institutional fit. Foreign money that bypasses local banks can only create formal jobs if existing training systems can handle large-scale reskilling. The success of external funding depends on prior investments in education and skills. As a result, even when foreign investment in AI bypasses local financial limits, most emerging economies will struggle to shift workers into new jobs. They need basic education and vocational schools already in place. State capacity in human capital development is the deciding factor for effective labor adjustment.
What conditions would cause the informal sector to fail as an absorber, converting underemployment into open unemployment and social unrest?
Jobless Cities
Informal sectors stop absorbing displaced workers when automation outpaces their ability to expand, turning widespread underemployment into concentrated joblessness.
In some cities, the number of people looking for work is growing. Many hope to find jobs in the informal sector, like street vending or odd jobs. But this sector can only absorb so many workers. When too many people enter it, each new job adds less value. This is called diminishing returns. Automation in factories and shops is pushing more people into informal work. But the informal sector is already full. It cannot take more workers. Digital tools let firms operate with fewer workers. This reduces job opportunities further. Workers cannot move to formal jobs either. They lack the right skills. Government systems do not help them switch jobs. So, fewer people find any work at all. The result is not just fewer hours or pay. It is outright joblessness. This happens in crowded urban areas. The oversupply of workers exceeds any possible informal jobs. At a certain point, informal employment stops growing. It may even shrink. This turning point has occurred in countries like India and Nigeria after 2016. Rising automation in manufacturing and retail caused it. The informal economy no longer absorbs labor. Instead, it becomes a source of visible unemployment. This situation leads to social unrest.
Could strong external actors, like multinational corporations or international institutions, substitute for weak state capacity in coordinating labor and capital during AI-driven disruption in emerging economies?
Automation Despite Negotiations
Automation proceeds unchecked where weak enforcement allows firms to bypass labor rules, even if negotiations are legally required.
In some developing countries, laws require employers and workers to negotiate over wages and working conditions. These countries also have systems meant to coordinate pay across industries. But large companies often have strong influence over government regulators. This influence weakens enforcement of labor rules, especially in export zones and among subcontractors. As a result, even though formal bargaining structures exist, dominant firms can still push automation without real limits. While big firms may discuss AI changes with unions, many workers are outside this protection. In supply chains managed by digital platforms, companies use algorithms to control labor with little union oversight. Investigations in Latin America and Southern Africa show that wage suppression and job losses continue unchecked. When labor laws are not enforced for outsourced workers, the intended protective effect of wage coordination fails. Therefore, having labor negotiation rules on paper does not stop job losses from automation.
What would need to be true about the informal sector's internal structure for it to retain any capacity to absorb displaced workers despite platform automation's price collapse?
Gig Platforms Weaken Informal Jobs
Gig economy platforms undermine informal job networks by replacing human brokers with automated systems, leaving displaced workers without entry points or viable livelihoods.
In many cities, informal workers rely on social networks and trust to find jobs and make a living. These networks help them survive where formal institutions are weak. But digital platforms are changing this system. They use automation and central control to offer services at lower prices. This undercuts traditional informal workers. As platform companies grow, they displace local labor brokers and middlemen. These brokers once helped connect workers with jobs. Their role is replaced by apps and algorithms. When this happens, displaced workers cannot easily find new entry points. It is not just that wages fall. The whole support system for absorbing new workers breaks down. The informal sector no longer absorbs job seekers as it once did. This failure is clear in transport and delivery jobs across South Asia and Sub-Saharan Africa. Evidence comes from World Bank and ILO reports. The result is deeper job insecurity for urban poor populations.
State-led Job Creation
Automation does not cause mass unemployment when states actively redirect investment into labor-absorbing sectors through planning and public spending.
Most developing economies do not have strong, independent labor unions that can negotiate wages for gig or informal workers. This weakens the ability of workers to set fair pay through collective action. In contrast, countries with strong state planning bodies can guide economic growth in ways that create jobs, even as automation replaces certain roles. China shows this pattern clearly after 2000. As factories automated, the government heavily invested in transport, construction, and green energy. These sectors hired many workers displaced by machines. The state used public spending and directed credit to fund these investments. This proactive allocation of resources prevented large-scale unemployment. The key factor is not worker bargaining power alone. Instead, the decisive force is the state’s ability to control investment and plan economic development. When the state can direct money and policy independently, it can shift workers into new sectors before job losses occur. This structural role determines whether automation causes job crises or leads to new employment through targeted growth.
What happens in emerging economies where decentralized tech hubs or foreign-led education initiatives bypass captured states to deliver AI-relevant skills training independently?
Tech Training Divide
Decentralized training programs deepen inequality because they lack state support and employer recognition, leaving most workers behind.
In some developing countries, powerful elites control politics and policy. This shapes how job training programs work. Tech hubs and foreign education projects often start outside government control. These programs train some workers for high-skilled jobs. But they do not connect with national labor systems. Most employers do not recognize their credentials. The state does not enforce common skill standards. Wages are not tied to skill levels. As a result, only a small group benefits. Most workers remain untrained for modern jobs. They face job loss from automation. No clear path exists for them to retrain. Training efforts stay separate and uncoordinated. They do not build on each other. Without state action, skill gains do not spread across the economy. The divide between skilled and unskilled workers grows. New training programs fail to create broad job growth. They repeat old patterns of exclusion.
Under what conditions would organized labor or small firms successfully resist the rent-seeking use of AI by dominant conglomerates in emerging economies?
Worker Power In Tech Talks
Strong labor institutions force firms to negotiate AI changes with unions, leading to training and job sharing instead of job cuts, which protects workers and competition.
In countries like Brazil and South Africa, strong labor rules set after major political changes protect workers from losing jobs to AI. These rules require companies to negotiate with unions before automating jobs. This includes workers in outsourced or app-based jobs, not just formal employees. When firms want to adopt AI, they must discuss job changes with unions. Unions then push the firms to invest in training and share jobs, rather than cut them. This stops big companies from using AI to weaken labor. Small firms also benefit because markets stay competitive. Without these talks, AI would reduce wages and jobs. With them, productivity gains go to workers and broader employment. The law is what makes this happen.
What happens to workforce reskilling outcomes when external funders prioritize AI infrastructure over vocational training in countries with weak educational institutions?
Unions And Job Survival
Unions cannot save jobs from automation when training systems are too weak to retrain workers at scale.
When companies threaten to move jobs abroad, wage coordination can only last if unions have legal power over technology choices. But in many developing countries, unions lack the ability to reskill workers. Schools and training systems are too weak to retrain large numbers of people. Even if unions can delay new technology, they cannot ensure workers learn new skills. The state and companies cannot deliver training at scale. Studies show job skills do not match what employers need in Latin America and Southeast Asia. Without strong training systems, union rights become symbolic. Unions may block new technology but cannot force retraining. Firms then choose to move jobs or replace workers with machines. Legal rights alone cannot prevent job losses. The key factor is the actual ability to retrain workers. If training systems are weak, automation leads to spikes in unemployment.
What would happen to urban unemployment rates in emerging economies if digital infrastructure expanded access to global gig work faster than automation displaced local jobs?
Gig Work Buffers Job Loss
Urban unemployment falls after automation only where strong digital access and skills allow rapid shift to global gig work.
In some cities, automation removes jobs in export industries. Many people then rely on informal work. But where internet access is strong, remote gig jobs can help. These jobs are found through global online platforms. The shift to gig work helps only if people can go online. Basic digital skills are also needed. There must be steady demand for gig labor worldwide. Countries that invested earlier in digital skills see benefits. They have better internet access and training programs. In these places, displaced workers find platform jobs faster. This keeps unemployment from rising after automation. Smartphones and English skills boost this effect. Big cities benefit most. The key factor is not how fast jobs vanish. It is how quickly workers join digital labor markets. Faster absorption means less jobless distress. Where digital infrastructure is weak, the opposite happens. Workers cannot shift fast enough. Job losses outpace new opportunities. Unemployment rises sharply. But where the groundwork exists, job markets stabilize. This pattern appeared in Southeast Asia between 2020 and 2023. It is confirmed by World Bank and ILO data. Digital improvements alone are not enough. Success depends on prior investments in access and training. When these are in place, unemployment falls after automation shocks.
Could stronger enforcement of labor regulations in export-processing zones alter the trajectory of AI adoption by discouraging automation that relies on informal labor shedding?
Factory Zones Bypass Labor Rules
Automation proceeds unchecked in export zones because weak oversight and banned unions prevent labor consultation, making national regulations ineffective at the factory level.
In some countries, national labor rules apply broadly. But special economic zones create separate regions with weak labor oversight. These zones host factories making electronics or clothing for export. Multinational companies operate in these zones to avoid strict labor standards. Even if a country has strong national wage agreements, the zones do not follow them. This happens because the zones are exempt from inspections and ban independent unions. As a result, companies automate production processes without consulting workers. Automation spreads quickly where labor rules cannot be enforced. The separation between national rules and local enforcement lets firms move risky changes into these zones. Regulatory oversight remains weak where automation grows fastest. Therefore, stronger labor enforcement would not slow automation in these areas. The system itself allows firms to shift labor changes to places where rules do not apply.
What institutional or regulatory conditions would need to be present for informal economies to resist or adapt to platform-driven structural displacement rather than collapse?
Ride-hailing And Street Jobs
Informal economies survive only when rules force digital platforms to include local brokers, because without those rules, automation eliminates the low-barrier jobs that absorb displaced workers.
In cities like Lagos and Dhaka, many workers join the informal economy through low-barrier jobs like street vending or informal transport. These roles are often managed by local groups that rely on trust and personal networks. When ride-hailing apps expand, they replace these roles with automated systems. This does not mainly hurt workers by lowering wages. Instead, it removes the simple entry jobs that let displaced workers survive. In African cities with little formal labor protection, this causes widespread job loss. The problem grows when governments do not require digital platforms to work with existing informal groups. Without rules forcing integration, apps ignore local brokers and disrupt whole livelihood systems. However, collapse can be avoided. If governments require platforms to include local intermediaries in dispatch or ID verification, some access is preserved. Only where such rules exist do informal economies stay stable. Where they do not, automation wipes out informal jobs rather than adapting to them.
Ride And Delivery Workers
Informal economies survive platform disruption only when worker groups with legal standing and shared resources keep their organizational power intact.
In cities like Lagos, Dhaka, and Nairobi, informal workers in ride and delivery services rely on trust and shared credit to stay in business. These workers depend on local networks to find jobs and resolve disputes. Digital platforms replace these networks with apps and algorithms. Instead of connecting through people they know, workers connect directly to the platform. This cuts out the middlemen but also breaks down support systems. Workers lose access to informal credit, job referrals, and group bargaining power. The real harm is not just lower pay but the loss of social infrastructure. When platforms bypass human networks, workers can no longer rely on shared knowledge or help during hard times. This weakens their ability to adapt. The damage grows worse where rules do not require platforms to share data or allow worker representation. In places where worker cooperatives already have legal status and tools like dispute mediation, some stability remains. These groups survive because they can negotiate and maintain group strength. Resilience lasts only where such formal support systems withstand platform growth.
What would happen to skill stratification in emerging economies if decentralized tech hubs were required to align with national labor institutions and adopt state-recognized accreditation standards?
City Job Networks
Informal job networks collapse when digital platforms displace workers in cities without room for new livelihoods because economic stagnation prevents job creation, even if social structures remain.
In cities where most people work in informal jobs, digital platforms are changing how services are delivered. This shift affects how well local worker groups can survive. Their resilience depends not just on being recognized by institutions, but on whether displaced workers can find other jobs nearby. In smaller cities in Africa, such alternatives are often missing. Many of these places already have too many workers competing for too few informal jobs. Manufacturing jobs are also declining. When digital platforms remove middleman roles in transport or delivery, workers lose their jobs. They cannot easily move to similar jobs because those roles are already taken. The city has no room to absorb them. Trust-based worker groups remain, but they are not enough. Studies in Kampala and Freetown show job growth cannot keep up with new workers. Simply requiring platforms to work with local brokers will not help. Resilience fails when the overall economy is not growing. Informal jobs survive only if the economy expands. Stability comes from growth, not just social ties.
What must be true about the political autonomy and organizational density of unions for these wage coordination mechanisms to persist when dominant conglomerates threaten capital flight or platform relocation?
Union Power Vs Capital Flight
Wage coordination persists only when unions have statutory veto power over technology and at least 30 percent sectoral coverage, because this blocks capital exit and turns relocation threats into bargaining tools for job-sharing.
Unions in emerging economies face pressure when AI causes companies to move capital. To keep wage coordination alive, unions need two things. They must have enough members to bargain across many factories. They also need freedom from government control. Brazilian unions have this freedom under the 1988 constitution. South African unions do not because the ruling party controls them. Where unions can block new technology through worker councils, they can use the threat of relocation as a bargaining tool. This wins job-sharing and retraining instead of lower wages. Where unions are politically captured, companies can threaten to move to cheaper zones. This makes wage coordination collapse. The key is that unions must have legal power over technology changes. They also need at least 30 percent of workers in their sector. Scandinavia has this pattern. Brazil and South Africa do not because their union density is below 20 percent and their political freedom is weak. So the mechanism only works when unions are insulated from both the state and big business.
