Dani Rodrik, a Turkish–American economist and professor at the Harvard Kennedy School, spoke at a conference of the Hungarian Chamber of Commerce and Industry in Budapest about what governments must do to create many good jobs. In an interview before the event, Rodrik outlined why he doubts the ability of industry alone to lift countries, whether globalization has ended, and what the spread of artificial intelligence might mean for workers.
Transitional times and the lessons of hyperglobalization
Rodrik argues the world is in a transition away from what he calls the era of “hyperglobalization,” when policymakers simultaneously pursued deeper economic integration, strong national sovereignty and democratic accountability. Hyperglobalization produced many winners but also many social losers: workers and regions were left behind, generating political tensions, rising populism and authoritarian tendencies.
He stressed China’s rise: through globalization China lifted hundreds of millions out of poverty, but it did so under a political and economic model different from the West’s, with growing state control. That difference has geopolitical consequences. Rodrik sees an unresolved triangle between national sovereignty, international integration and democracy: sovereignty has strengthened in recent years, while democracy has weakened in many places.
Three scenarios for the future
Rodrik laid out three possible outcomes. The “bad” scenario would be a 1930s-style collapse of international economic ties, which he finds unlikely given today's high level of interconnection. The “ugly” scenario would be geopolitics framing every economic question as zero-sum, undermining international cooperation on issues like climate change and AI governance. The “good” scenario would see governments balance integration and sovereignty to strengthen democracy, rebuild the middle class and address distributional conflicts caused by hyperglobalization.
Beyond welfare: creating good jobs
Rodrik argues the post–World War II welfare-state model once addressed distributional problems effectively, but structural changes in economies make simple redistribution insufficient today. Industry’s declining ability to employ large shares of the workforce, global competition and technological change mean governments must focus on creating productive, dignified jobs—particularly in service sectors such as care, retail and other services that will employ most people in the future.
He emphasizes that people seek dignity, recognition and identity through work, not simply cash transfers. Therefore policy should prioritize raising job quality and worker autonomy rather than relying solely on redistribution.
Platforms, technology and work organization
Rodrik warned that many current platform companies apply a Taylorist fragmentation of tasks that depresses wages, strips worker autonomy, raises turnover and yields low productivity. Yet technology could be used differently: platforms can equip workers with new capabilities, increasing autonomy and productivity. There are international examples where firms, often with support from social enterprises, local governments or national policies, reorganize work to boost skills, job quality and productivity.
He recommends governments use incentives, regulations, subsidies and training to steer firms toward better work organization. The aim is not to forcibly change firms but to make the right choices more attractive—for example, encouraging platform companies to share technological tools with restaurants so those businesses are less dependent on apps and capture more value.
Governments have driven past technological revolutions
Rodrik reminded listeners that major transitions—like the shift to electric vehicles or the renewable-energy revolution—were driven not solely by markets but by government investments and policies. He cited China’s strong role in renewable technologies as an example. Similar public action is needed to foster good jobs in the service sector.
AI: risks and opportunities
Rodrik warned that leaving the application and governance of AI entirely to a handful of giant firms could produce social shocks far worse than past globalization. He rejects the idea that redistribution, shareholder participation or unconditional basic income alone would solve the problem.
Still, he is somewhat more optimistic about AI than about earlier automation: AI can potentially make valuable knowledge and capabilities available to less-skilled workers. For example, in the care sector AI could allow lower-skilled caregivers to perform basic diagnostic tasks, improving access to care. Small business owners could use AI to access advisory services previously affordable only to large firms.
The industrial development route has narrowed
Rodrik observes that industrialization was once a reliable path for developing countries: relatively low-skilled workers could become productive industrial employees. Today, modern industrial technology demands more capital and skills, global value chains require extensive regulation, infrastructure and training, and the peak share of industrial employment in later industrializers is lower and shorter. Consequently, the historical “escape route” through mass industrial employment is largely closed.
He noted that countries like Hungary benefited from proximity to European markets, but that future jobs will largely be outside manufacturing. While industry remains important for national security, innovation and the green transition, any policy that ignores the fact that roughly 80 percent of future jobs will be outside manufacturing is misguided.
What governments should focus on
The central question, Rodrik says, is how to employ state tools so that the service sector provides good, productive jobs. He calls for a new kind of “industrial policy” for services: government measures that induce firms—from SMEs to large platforms—to create quality jobs. Tools include incentives, regulation, training programs and arrangements encouraging technology sharing.
Rodrik concluded with a warning: without active public shaping of the technological transition, society risks ending up in a dystopia where a few giant firms control technological progress and most workers are stuck in poor-quality jobs. That outcome can be avoided only if governments take a decisive role in steering technology toward broadly shared benefits.
