By John Martin
The panel “Orchestrating Economic Transformation: From Local Experiments to National Strategy” brought together a group of scholars and practitioners to revisit one of development economics’ oldest questions: how developing economies move people into more productive work. Moderated by Faculty Co-Director Dani Rodrik, the conversation began with emphasizing the tendency for big national strategies to overlook the granular institutional and incentive details that determine whether firms invest, workers move, and programs scale.
Rodrik set the stage with a forward-looking employment picture for developing countries that emphasized the reality that most workers will remain in relatively sheltered, non-tradable activities, often with less formal education. That implies that a successful development strategy cannot hinge only on globally-exposed frontier sectors, but must also raise productivity where a majority of the population actually works. Yuen Yuen Ang, Alfred Chandler Chair Professor of Political Economy at Johns Hopkins University; Arkebe Oqubay, British Academy Global Professor at SOAS University of London; Santiago Levy, Nonresident Senior Fellow at The Brookings Institution; and Eliana Carranza, Global Lead for Labor and Skills in the Social Protection and Labor (SPL) Global Practice at the World Bank, offered their views on solutions to these central questions in development.
In an era of disruption, the central policy problem is designing for learning under uncertainty
Ang challenged the assumption that governments can reliably identify future growth engines in advance, as many successful industrializers once seemed to do. With deglobalization pressures, AI-driven disruption, and shifting security priorities, she argued that uncertainty is the defining feature of contemporary policy. This implies that development debates organized around a binary are increasingly unhelpful.
Ang’s alternative is what she terms “directed improvisation”. Under directed improvisation, the state acts less as a dictator of solutions and more as a director of a discovery process. Crucially, “direction” simply means setting guardrails and structuring a cycle of variation, selection, and retention so that workable approaches emerge and can be scaled.
An important aspect of her argument is that this approach should not be treated as China-specific. Ang noted how readily theories derived from European history are assumed to travel, while non-Western experiences are often presumed culturally bound. To best position a developing country irrespective of culture or regime type, they should accept uncertainty when doing experimentation.
Focus on building institutions that can manage heterogeneity
Carranza focused on finding a successful local innovation and taking it to scale. She argued that the difficulty of policy integration is managing the complexity that inevitably appears when interventions move across populations, geographies, firm types, and governance environments. Effectiveness often declines at scale because the conditions that made it work are not uniform.
Policy levers she recommended were developing tools for diagnosing constraints that identify “where and for whom,” platforms for profiling and referral, mechanisms for continuous employer engagement, coordination across providers and ministries, and feedback loops that convert implementation outcomes to be continuously refined. Without these, countries oscillate between rigid national blueprints that ignore local variation to fragmented local experimentation that remains disconnected and difficult to integrate.
Oqubay’s remarks reinforced this point from the vantage of state-led transformation efforts in Ethiopia and broader Africa. His examples were stories of past institutional failures. He emphasized the capability to admit mistakes. Practitioners should study other countries with specific questions; they should adapt rather than copy, and should iterate with attention to sequence, speed, and scale.
Inclusive transformation requires fixing incentive structures that lock workers and firms into low productivity
Levy argued against the notion that industrial policy, export growth, or human capital investment as sufficient by themselves. Using Mexico as an example, he highlighted the reality that Mexico is undergoing extraordinary manufacturing export success alongside weak overall productivity and persistent informality of work. In his account, the standard dynamic story, that productive firms expand and unproductive firms shrink, fails when the micro-incentives are reversed.
Levy suggested that policies around taxation, labor regulation, social insurance design, and contract enforcement can systematically subsidize low-productivity informality while taxing and discouraging growth among high-productivity formal firms. When that is true, even a booming export sector may not pull the rest of the economy forward. His emphasis was therefore on sequencing. That is, before layering on sophisticated productive development initiatives, countries must remove anti-productive policies that continuously reproduce segmentation.
Transformative development is ultimately institutional and political. Challenges arise, whether that be in managing openness amid instability or learning across varying geopolitics. Policymakers and practitioners, thus, should be focused on durable progress that can diagnose realities accurately, coordinate across actors, and adapt.
