In the second book of Kautilya’s Arthashastra (a text that predates Adam Smith by roughly two millennia) there is a chapter devoted to the Superintendent of Metals. Kautilya argued that the state should not merely regulate mining. It should own the smelting works, control quality of output, fix the wages of artisans, and determine which products could flow through private trade and which must pass through royal workshops. The idea that a government could rationally direct an industrial economy is, in that sense, rather old.
China’s 15th Five-Year Plan is a restatement of that old idea. The plan covers 2026 to 2030. GDP is to grow within a “reasonable range,” calibrated annually, more honest than a fixed target. R&D expenditure is to grow at over 7 per cent per annum in real terms, pushing China towards the 3 per cent of GDP threshold that separates innovation-led economies from catch-up ones (the 14th Plan period recorded 2.8 per cent). High-value invention patents per ten thousand people are to rise from 16 to over 22. The share of core digital economy industries in GDP is to climb from 10.5 per cent to 12.5 per cent, implying roughly 2.5 trillion yuan of additional value-added. The plan carefully distinguishes binding indicators from expectational ones.
The industrial architecture has three tiers. Traditional sectors (steel, petrochemicals, shipbuilding) are to be upgraded through digital and green technologies, with the “orderly withdrawal of outdated and inefficient production capacity.” Strategic emerging industries, covering next-generation information technology, intelligent connected vehicles, robotics, biomedicine, and aerospace, are to be scaled. Future industries (quantum technology, bio-manufacturing, hydrogen energy, nuclear fusion, brain-computer interfaces) are to be seeded, with the plan candidly admitting uncertainty: “exploring diverse technological routes” and building “mechanisms for sharing risks.”
The plan’s most technically specific section covers industrial base reconstruction: breakthroughs in fundamental processes and materials; landmark machinery including high-end CNC machine tools, large LNG carriers, and the CR450 high-speed train. Particularly notable is the procurement policy requiring state buyers to adopt domestically produced advanced equipment even when imports are available. No market for high-end domestic capital goods develops without guaranteed initial demand. This is precisely what a World Bank policy research report published last month, by Ana Margarida Fernandes and Tristan Reed, calls “public procurement to incentivize quality improvement and innovation by local suppliers”, one of the sharper instruments in their taxonomy of 15 industrial policy tools.
Scores exceptionally high
That report is worth pausing on. Reviewing the growth strategies of 183 economies, Fernandes and Reed find that all of them target at least one industry. Industrial policy is universal. What differs is sophistication, and the three national characteristics that determine what a government can plausibly deploy: the size of its domestic market, its administrative capacity, and its fiscal space.
On all three, China scores exceptionally high, making it one of the very few economies that can legitimately reach for the full toolkit. Most developing-country governments resort to the bludgeon of sweeping tariffs and subsidies when they should use the scalpel of industrial parks, skills programmes, and procurement policy. China’s 15th Plan is broadly reaching for the scalpel.
Its ambitions are ordered on a rather different scale. Which raises the question of execution. Between 2016 and 2020, China eliminated roughly 150 million tonnes of crude steel capacity and shut down all illegal ground-strip steel producers, thousands of them, by administrative direction, within months in 2017. Coal capacity fell by over 800 million tonnes in the same period. The mechanism, when central objectives are aligned and local employment pressures manageable, demonstrably functions. The current overcapacity in electric vehicles and solar panels (sectors the 15th Plan itself identifies as suffering from involutionary competition) shows precisely where it breaks down: when provincial governments protect payrolls that Beijing wants to rationalise.
Fernandes and Reed make one further point worth noting. Industrial policy should be temporary, buying time while fundamentals are built, ideally within a decade. China’s planning tradition recognises no such sunset. Each five-year plan begets the next. Whether permanent planning can sustain the discipline that temporary interventions sometimes achieve is a question the 15th Plan does not engage.
The plan targets per capita GDP at levels of moderately developed countries by 2035. The compound arithmetic is not difficult. Getting the industrial policy right, sector by sector, is a different order of problem entirely.
Kautilya’s Superintendent of Metals was accountable to the king. Modern planners are accountable to targets. Whether the targets remain accountable to reality is always the harder question.
Sinha writes on macroeconomics and geopolitics
Published on April 8, 2026