Ethiopia’s developmental state is often discussed as though it presents a simple verdict. Supporters point to the roads, dams, universities, railways, industrial parks, and public services built during two decades of rapid growth. Critics point to foreign-exchange shortages, inflation, mounting debt, weak exports, constrained private investment, and too few productive jobs. Both accounts are true. That is precisely why the model’s experience matters now.
The relevant question is not whether the state should lead or the market should rule. No successful economy has developed through such a clean choice. The question is whether Ethiopia can build a state capable of correcting market failures without protecting public failure, and use markets to reveal costs and opportunity without pretending that prices alone will create productive capabilities.
Ethiopia’s old model deserves more credit than its current reputation allows. In a country with little private capital and immense infrastructure gaps, waiting for the market to finance transformation would have meant waiting indefinitely. As documented by statistical authorities, poverty fell from 39 percent to 24 percent between 2004 and 2016, while access to electricity, water, education, vaccination, and transport improved. The state demonstrated a rare ability to mobilize resources and execute large projects.
RELATED POSTS
But Ethiopia constructed the hardware of development faster than the productive system needed to use it. Power generation did not automatically produce competitive factories. Industrial parks did not automatically create domestic suppliers. Roads did not by themselves overcome costly logistics, unreliable customs, skills shortages, regulatory uncertainty, or rationed foreign exchange. By June 2021, the World Bank estimated that industrial parks employed roughly 90,000 people. That was meaningful, but small beside the approximately 1.8 million Ethiopians entering the labor market each year.
The first failure was macroeconomic. The model depended on imported machinery, fuel, construction materials, and technical services, but it never generated exports on the scale required to finance them. A managed and increasingly overvalued exchange rate made imported capital appear cheaper, while weakening incentives to export and deepening the rationing of dollars. Public borrowing covered the gap for a time. Eventually, debt service, industrial inputs, medicine, fertilizer, and fuel were all competing for the same scarce foreign currency.
The second failure was institutional. The state was strong enough to direct credit and build assets, but often too weak to evaluate returns, stop failing projects, or expose public enterprises to commercial discipline. Losses could remain hidden in state-owned companies, public banks, guarantees, and an administratively priced exchange rate. Success was measured in kilometers of road, megawatts of power, or hectares of industrial land, rather than in productivity, export earnings, supplier development, and sustainable jobs.
The third failure was informational and political. The best developmental states were not merely autonomous from private interests; they were embedded in dense networks through which officials learned from firms, workers, exporters, and technical experts. Political economist Peter Evans called this embedded autonomy. Research on Ethiopia by Camille Pellerin, affiliate researcher at the French Centre for Ethiopian Studies in Addis Ababa, describes something closer to dis-embedded autonomy: an administration that could act forcefully but distrusted independent business organization and relied heavily on state-owned enterprises, party-linked endowments, and foreign investors.
That distinction is decisive. When chambers and business associations cannot speak candidly, policymakers lose the information required to distinguish a genuine coordination problem from a request for privilege. Formal consultation gives way to personal access. Firms invest in relationships rather than capability. Closed politics then becomes an economic liability, because bad news travels slowly upward and failed policies survive longer than they should.
The fourth failure was the absence of a credible bargain between support and performance. South Korea did not industrialize simply because the government offered protection, cheap credit, and subsidies. Support was often tied to export targets, and firms that repeatedly failed could lose access. Economist Alice Amsden described this as a reciprocal control mechanism: the state provided rents, but demanded measurable results in return. Ethiopia was better at allocating support than withdrawing it. Without transparent targets, time limits, and the threat of exit, industrial policy easily became an entitlement.
Finally, the model treated political order as something development would eventually secure, rather than as an input into development itself. Centralized authority enabled rapid implementation, but it did not create a durable settlement over power, resources, and belonging. Conflict later destroyed assets, disrupted trade, displaced workers, weakened revenue, and frightened away long-term capital. Trust is also infrastructure, and Ethiopia invested too little in it.
None of this proves that developmental-state thinking was misguided. It shows that Ethiopia copied the visible instruments of East Asian development more successfully than the institutions that made those instruments productive. Korea combined support with export discipline. Singapore combined macroeconomic stability with a high-capacity public agency that coordinated foreign investment, skills, infrastructure, and local industry. Vietnam’s Doi Moi reforms liberalized prices, agriculture, trade, and foreign investment, while the state retained a strategic role in infrastructure, human capital, and integration into global production.
These countries are not templates that Ethiopia can photocopy. They developed under different geopolitical, demographic, and trade conditions, and each accumulated its own problems, including concentrated corporate power, inequality, financial risk, and inefficient state firms. Their common lesson is narrower but more useful: industrial policy works only when public support is disciplined by competition, performance, learning, and institutional competence.
Can such an approach coexist with Ethiopia’s current reform program? It must. The exchange-rate reform launched in July 2024, the effort to end monetary financing, the shift toward interest-rate-based monetary policy, debt restructuring, stronger revenue mobilization, financial-sector opening, and greater scrutiny of state enterprises address distortions the old model could no longer conceal. Restoring truthful prices and a credible budget is not an abandonment of development. It is part of rebuilding the state’s capacity to pursue it.
But stabilization is a foundation, not a development strategy. A weaker currency can encourage exports only if firms can obtain working capital, reliable power, inputs, logistics, certification, and market access. Financial liberalization can allocate capital better only if competition is real and supervision is capable. Privatization can improve performance only when assets are valued transparently and monopoly power is contained. Otherwise, reform may replace public inefficiency with private rent extraction.
The reform’s distributional design is equally important. Currency depreciation and tighter monetary and fiscal policy impose immediate costs, while the productivity gains they promise arrive later and uncertainly. The IMF’s January 2026 review found that Ethiopia’s contribution to targeted safety nets had fallen below the program target and that a transparency benchmark for Ethiopian Investment Holdings had been missed. These may appear to be technical deviations. They point to the central political-economy problem: reform loses legitimacy when its discipline is visible to households but less visible to powerful public institutions.
A renewed developmental state should therefore be smaller in its promises, more selective in its interventions, and tougher in its evaluation. Ethiopia lacks the fiscal space for another broad campaign of subsidized credit and debt-financed megaprojects. It should concentrate on a few tradable systems where coordination can produce foreign exchange and mass employment, including higher-value agriculture and agro-processing, selected manufacturing, digital services, tourism, logistics, and renewable-energy exports. The objective should be to solve specific constraints, not declare favored firms permanent national champions.
Every incentive should carry a reciprocal obligation. Preferential credit, tax relief, industrial land, customs facilitation, or foreign-exchange access should be conditional on published targets for exports, productivity, jobs, training, technology transfer, or local sourcing. Benefits should expire automatically unless results are verified. Firms should compete for support under transparent rules, and repeated failure should mean exit, regardless of ownership or political connection.
The state must also become more embedded without becoming captured. Permanent sector councils should bring together domestic firms, exporters, workers, universities, regional governments, and regulators to identify bottlenecks and publish agreed actions. Public banks and enterprises require audited accounts, professional boards, hard budget constraints, and a clear separation between the state’s roles as owner, regulator, and policymaker. Competition policy, bankruptcy procedures, and an independent judiciary are not neoliberal decorations. They are the mechanisms that keep a developmental state developmental.
Agriculture and social protection must remain at the center. With most Ethiopians still depending directly or indirectly on farming, productivity, irrigation, storage, extension, rural roads, land-use security, and competitive agricultural markets will do more for industrialization than another isolated factory shed. Well-designed safety nets, health, and education are not merely compensation for reform. They preserve human capability and the public consent on which a long transformation depends.
Ethiopia should not return to the old developmental state, and it should not swing toward the equally dangerous belief that liberalization will do the state’s work. The old model confused state power with state capacity. The current reform risks confusing market prices with market development. A better model would combine macroeconomic truth with industrial ambition, private initiative with public coordination, and support with accountability. That is not a retreat from reform. It is the only way reform can become development.
—-
Samson Berhane is an economics graduate with expertise in business and economic reporting and communications. He can be reached at [email protected].The views expressed in this article are his own and do not represent the opinions of the institutions he is affiliated with nor that of the magazine.
