Industrial policy is often sold as a fast way to rewrite an economic trajectory. Pick sectors. Channel money into them. Create jobs. Increase output. In China, this logic was pushed to its extreme. State support covers dozens of industries, from shipbuilding to electric vehicles and semiconductors. The scale of intervention is massive. The outcomes have been highly uneven.
China has shown that growth in production is not the same as growth in competitiveness. In a number of sectors, the country reached global leadership in volumes. Yet efficiency, profitability, and technological depth often lagged behind. State money accelerated capacity expansion, but did not always improve business quality. This is the key fork in the road for any country trying to follow a similar path.
For Kazakhstan, the risk lies in the surface appeal of the Chinese model. Rapid growth. Large factories. Export ambitions. But behind the façade sits a complex architecture of incentives, mistakes, and distortions. Copying the form without understanding the substance almost guarantees disappointment.
In China, industrial policy was not driven by central decisions alone. Regions played a huge role. Provinces competed with one another for projects, subsidies, and priority-zone status. This competition pushed capacity expansion even where markets were already saturated. Factories were built faster than sustainable demand emerged. The number of players increased. Industries fragmented. Productivity declined.
Shipbuilding is a telling example. China achieved dominance in global output, controlling more than half of the world market. But the cost was high. Hundreds of shipyards. Low utilization rates. Weak margins. State support allowed companies to survive, but did not force them to become more efficient. Market consolidation was slow. This is a case of policy creating scale without quality.
This logic is familiar to Kazakhstan. Regional authorities compete for projects. Every region wants a factory. Every region wants a launch event and a reportable success. Economic logic gives way to administrative incentives. Projects multiply. Their long-term viability becomes secondary. The state gains statistics, but not resilient businesses.
China’s experience also highlights another critical point. Metrics matter. When success is measured by output volume or market share, the system adapts to deliver exactly that. Companies expand even if efficiency deteriorates. Scale replaces innovation. Labor productivity ceases to be a priority.
In several sectors, growth was accompanied by falling returns on capital. Investment rose faster than outcomes. This was not an accident. It was the direct result of chosen benchmarks. If the target is output, output will be delivered. If the target is efficiency, policy must look very different.
In Kazakhstan, this problem is especially acute. Support programs are rarely tied to productivity, high value-added exports, or technological upgrading. More often, success is measured by the number of projects and the volume of funds disbursed. Such a design encourages launch, not development.
Electric vehicles in China became the exception rather than the rule. Here, state support coincided with rapid domestic demand growth, intense private competition, and integration into global value chains. Infrastructure expanded alongside the market. Companies fought for consumers, not just subsidies. As a result, competitive players and technologies emerged. This case is often cited as proof that industrial policy works. But it only works when several conditions align. A large market. Rapidly growing demand. Access to technology. Real competition. Without these factors, state support turns into rent.
Kazakhstan should not overestimate the universality of this example. The domestic market is smaller. Access to technology is different. Competitive density is lower. Attempts to replicate such stories without these conditions will lead to imitation rather than results.
The most revealing case was semiconductors. China poured enormous resources into developing the industry. Large funds were created. Flagship projects launched. A rapid technological leap was expected. It did not happen. The gap with global leaders remained. The sector ran into shortages of expertise, equipment, and managerial know-how. Some projects ended in scandals and closures.
This example illustrates the limits of industrial policy. Money does not replace knowledge. Technological complexity cannot be bought with subsidies. Without access to advanced ecosystems and accumulated experience, results remain constrained. For Kazakhstan, this lesson is especially important. Attempts to jump straight into complex technological niches look impressive in presentations. In practice, they risk freezing capital and producing disappointment. Industrial policy must start not with ambition, but with a sober assessment of domestic capabilities.
Another key takeaway from China is policy design. Where support was temporary, competitive, and conditional, outcomes were stronger. Where subsidies became permanent and unconditional, companies adapted to assistance and lost incentives to evolve.
In Kazakhstan, exit mechanisms are often missing. A project is launched. Support continues. Even if economic logic has faded. This creates lock-in. The state finances past decisions instead of future growth.
Industrial policy should not substitute the market. It can lower barriers, accelerate learning, and support infrastructure. But it performs poorly as a permanent income source for business. China’s experience confirms this at a scale that is hard to ignore.
For Kazakhstan, the main conclusion is not to abandon industrial policy, but to abandon illusions. Such policy requires hard rules, transparent goals, and a willingness to recognize failure. Without this, it becomes an expensive showcase. China has demonstrated that even with vast resources, success is not guaranteed. For countries with fewer resources, the cost of mistakes is higher. The question is not whether to intervene, but how and why.
Industrial policy does not work automatically. It works only where institutional discipline, market forces, and a realistic assessment of capabilities align. Everything else is a costly experiment.
Ruslan Sultanov, economist, President of the “PharmMedIndustry Kazakhstan” Association, specifically for www.economyKZ.org
