The Trap of Going All-In on Semiconductors — Dissecting the Economics of Handouts to Chaebols
Source
Why this was selected
This article simultaneously fulfills three virtues rare in Korean semiconductor discourse. First, theoretical sophistication — it compares the conditional subsidy regulations of the US, Japan, and Taiwan with Korea's unconditional support for chaebols using the framework of import substitution industrialization (ISI), and points out the decline in the employment multiplier in the age of AI and robotic automation with specific figures. Second, alignment with reality — it lists actual indicators that puncture the 'success narrative,' such as the suspension of Samsung Electronics' P4, P5, and P6 factory construction in Pyeongtaek, a halving of foundry investment in 2025 year-on-year, and a collapse in exports from $83 billion in 2018 to $42.9 billion in 2023. Third, specificity hard to find through basic searches — figures such as domestic semiconductor production capacity projected to reach 21.9% of global demand while domestic demand is only 5.4%, indicating built-in structural overproduction; Samsung's direct employment of 39,000 against a 360 trillion won investment; and the tenfold discrepancy between the US CHIPS Act's 700 trillion won investment for 56,000 jobs and the Korean government's claim of 622 trillion won for 3.46 million jobs — these are rarely compiled in general reports or academic summaries. At a time when the entire South Korean political spectrum (including both ruling and opposition parties, and some progressives) stands on a consensus that 'semiconductors are a growth engine,' this is a rare article that deconstructs that very consensus through figures and an economic-historical perspective.
Context
From 2024 to 2026, South Korean political circles—across both ruling and opposition parties—moved under a consistent growthist consensus of 'semiconductor mega-cluster of 622 trillion won, Yongin national industrial complex, K-Chips Act, and Semiconductor Special Act.' President Lee Jae-myung also mentioned the need to introduce a 52-hour workweek exception when he was party leader in February 2025 (later, at the end of 2025, the exception clause was excluded from the bill through a bipartisan agreement, but a supplementary opinion acknowledging 'necessity' remained), and in January 2026, he formed a presidential special committee to push forward the follow-up to the Semiconductor Special Act. Hong Seok-man's article deconstructs the very economic foundation of this cross-party consensus — that overinvestment collides with the domestic-demand-oriented policies and supply chain restructuring of the US and China, and that a 30-year long-term investment plan could deteriorate into hundreds of trillions of won in non-performing assets due to 'moral depreciation' when alternative technologies like AI and neuromorphic computing emerge. This supports the 'dual dependence + chaebol-state reverse capture' thesis raised in this site's reference page 'Korea Semiconductor Geopolitics' (/p/korea-semiconductor-geopolitics) and the 'Lee Jae-myung Government 10-Month Evaluation' series at the level of corporate management and industrial organization. Other articles from the Chamsesang planned series that are worth reading together (Yoon Hyun-jung's 'Crisis Cannot Be Solved by Privatizing the Public,' Lee Hyun-jung's 'Semiconductors and AI Are Not Free,' and Lee Jong-ran's 'Unhealthiness of Semiconductor Workers') also remain candidates for future curation on this hub.