Keun Lee on AI, Chinese growth, innovation systems, and more by Keun Lee – Project Syndicate

Project Syndicate: Fears that generative artificial intelligence will cause mass unemployment in the long term are likely “overblown,” you argue, not least because the technology will probably bring considerable product innovation (which tends to boost overall employment), in addition to process innovation (which has more uncertain employment outcomes). Beyond re-skilling and up-skilling initiatives, what government interventions would support employment, and how can policymakers strike “the right balance of support and discipline” of AI companies?

Keun Lee: Perhaps most important is what governments must not do: impose excessive regulations on new waves of AI-related innovation. Countries that take a more flexible, open-minded approach, leaving plenty of space for experimentation, will benefit from more new ideas and experiences – and gain critical data revealing where regulation is most needed.

At the same time, governments must prepare for significant labor-market changes. As generative AI is applied in ways that boost the productivity of human labor, wages will rise and working hours will fall. The wage premium enjoyed by some white-collar professionals will also decline, leading to a reduction in some types of income inequality. The details, however, will depend significantly on how AI is applied. Labor-market reforms must proceed accordingly.

PS: In February, you highlighted some of the potential pitfalls of industrial policy, using examples from countries ranging from Malaysia to Turkey to South Korea. Are there relevant lessons for the United States and other advanced economies as they embrace industrial policy?

KL: A successful industrial policy requires not only monetary incentives, but also specific programs focused on building up the capabilities of firms and workers. Capability failures can doom a firm or industry just as easily as market failures.

For advanced economies, combining existing design (software) and new manufacturing capabilities will be essential to meet the goals of industrial policies. As it stands, these countries tend to lack manufacturing capabilities, including qualified workforces and supplier networks, which take time to develop. Simply transplanting foreign companies onto their territory will not do the trick.

PS: Last year, you called inequality a “serious problem” for China, but one the government had recognized and “committed to addressing” through the “so-called common prosperity campaign.” Since then, China’s government has significantly reduced its emphasis on common prosperity, focusing instead on boosting growth. What does this shift imply about China’s ability to avoid the so-called middle-income trap that has ensnared most emerging economies?

KL: China’s current leadership has always put more emphasis on growth than equity, and rising tensions with the US have reinforced this tendency. But equity is vital not only for social and political stability, but also long-term economic growth and development. China cannot afford to ignore it.

The Kuznets curve shows how industrializing countries initially experience a rise in income inequality, but at a certain point, inequality begins to decline. China has yet to reach this turning point. This is partly because, beyond significant differences in returns on labor (wages), there are massive discrepancies in asset-based incomes.

Access every new PS commentary, our entire On Point suite of subscriber-exclusive content – including Longer Reads, Insider Interviews, Big Picture/Big Question, and Say More – and the full PS archive. 

Subscribe Now

Compounding China’s inequality problem is rampant corruption. A transparent, credible monitoring agency is essential to tackle this issue, though the establishment of such an institution would probably have to go hand in hand with the embrace of other democratic practices, including freedom of the press and freedom of speech.

This means that, as long as China maintains its current system of developmental authoritarianism, it will struggle significantly to reduce corruption and thus inequality. But just as other Asian countries have moved beyond this phase, so, too, can China.

BY THE WAY . . . 

PS: Unfavorable demographics are another major economic headwind for China. How will an aging and shrinking population affect China’s quest for advanced-economy status – the central theme of your 2021 book, China’s Technological Leapfrogging and Economic Catch-up: A Schumpeterian Perspective – and what policies are needed to offset its effects?

KL: While China is alone in grappling with the legacy of its one-child policy, there is plenty of precedent for a country with rising income levels to experience declining fertility, in Asia and beyond. These experiences point to effective solutions, beginning with ensuring that families have access to decent housing, education, and childcare.

Europe’s experience indicates that, rather than placing the burden for meeting these needs fully on individuals, the state should regard them as essential social services, financed by taxes. Of course, striking the right balance between private and public contributions will not be easy, and there is no one-size-fits-all approach. The good news for China is that these fertility-boosting interventions would also support progress toward high-income status.

PS: Your book applies the insights of the twentieth-century economist Joseph Schumpeter to shed light on China’s rise. Which Chinese industries best exemplify the Schumpeterian perspective on innovation and growth?

KL: One Schumpeterian thesis is that, when it comes to technology, latecomers can “leapfrog” over more established players. The latecomer thus becomes the new frontrunner, while the old leader remains locked into existing technologies and business models, owing to factors like sunk investment and inertia.

When it comes to traditional manufacturing, China achieved rapid catch-up, but it is in new information-technology services that it has managed to achieve leapfrogging. Companies like Alibaba, Tencent, and Baidu did everything Amazon, Facebook, and Google did, but soon forged new paths as well. Their exceptional growth was facilitated and fueled not only by asymmetric state protection or business-model innovation, but also by real technological innovation, as shown by their accumulation of US patents. In particular, Tencent is generating and supporting hundreds of spinoffs and startups.

PS: Whereas “innovation is still a black box or residual” for neoclassical economics, you write, neo-Schumpeterian economics argues that it occurs in a relatively ordered pattern, owing not least to “innovation systems.” What are some of the key components of an effective innovation system for China and other emerging economies

KL: The effectiveness of an innovation system depends on the synergistic interactions among relevant actors and surrounding institutions, such as the financial sector, the intellectual-property-rights regime, and the education system. Weak interactions amount to system failures, which are different from neoclassical market failures. Whereas correcting a market failure might be as straightforward as paying tuition so that one may go to school, correcting a system failure might entail boosting the quality of teaching and expanding extracurricular offerings.

Individual and collective learning by trial and error is also essential to an effective innovation system. This requires some degree of redundancy, because one cannot know ex ante which experiments will succeed. Innovation systems in emerging economies must be flexible, encouraging both boldness and practicality in experimentation. And they must include mechanisms to identify and diffuse successful models and ideas.