Why Every Regime Repeated the Mistake: A Political Economy

The puzzle the series has been circling

This series has tracked, across twenty-four articles, a recurring pattern: Ethiopia’s monetary disorder has survived two complete changes of regime (imperial to Derg in 1974, Derg to EPRDF in 1991), one substantial political opening (Abiy’s 2018 reforms), and at least four ostensibly major monetary reform packages (the 1945 founding, the 1963 split, the 1992 devaluation, the 2024 float). Each new regime arrived promising to fix what its predecessor had broken. Each, with the partial and possibly temporary exception of the post-2024 arrangement, ended up running the same architecture in slightly different form.

This article asks: why? Why has the disease been so durable across radically different governments with radically different ideologies and political coalitions? The answer is not that Ethiopian leaders have been uniquely stupid or uniquely corrupt; the answer is structural. The article makes the case that the monetary disorder this series tracks is the equilibrium outcome of a specific political-economy configuration, and that the configuration has survived regime changes because each new regime found itself facing the same incentives that the old regime had faced.

This is the deepest argument of the series. It is also the one that distinguishes the series from standard policy advocacy. Standard policy advocacy assumes that if the right reform is identified and articulated, the political system will eventually adopt it. The series’s argument is that the right reform has been identified and articulated many times — by the IMF in Article IV after Article IV, by domestic Ethiopian economists, by visiting Western advisors, by international financial institutions — and the political system has not adopted it because the political system is structured in a way that systematically rejects the reform regardless of who is in power. Understanding why is the precondition for any reform that lasts.

Four constituencies, three regimes, one equilibrium

The political economy of the Ethiopian monetary regime has, across all three of the post-1974 governments, rested on the same four constituencies. Each gains from the disease. Each loses, in specific ways, from the cure.

State-owned enterprises and the public-sector workforce. The state enterprises that absorb subsidised credit through the state banks, that receive priority forex allocations, and that benefit from the implicit support of the financial-repression machinery — these are the largest employers in the formal Ethiopian economy and the largest single beneficiaries of the existing regime. Their employees (and the wider networks that depend on them) constitute a politically active constituency in urban Ethiopia. The post-2024 reforms have begun to expose this constituency to market pressures the previous regime had buffered. The constituency’s political response is, predictably, opposition to deeper reform.

The state political and security apparatus. The ruling party — whether the Workers’ Party of Ethiopia under the Derg, the EPRDF after 1991, or the Prosperity Party today — has used the monetary mechanism as part of its broader system of resource allocation. Control over forex allocation, credit allocation, and the timing of devaluations has been one of the principal levers by which the party has maintained patronage relationships with regional allies, ethnic constituencies, and politically-connected private firms. A genuinely independent monetary authority would, by removing the political discretion, weaken the party’s patronage architecture. This is not, the series would observe, a problem the party has any incentive to want to solve.

The politically-connected private sector. The private firms that have, across the EPRDF and post-EPRDF period, accessed the subsidised credit, the priority forex, and the state-favoured contracting arrangements — these constitute a small but politically influential constituency. The MIDROC group, the EFFORT-affiliated enterprises that have continued under various forms, the various conglomerates that emerged through GTP-era industrial policy — these are economic actors with substantial political voice and a direct interest in the existing arrangement. They have, in many cases, the resources to influence the political process through formal and informal channels.

The diaspora-connected and foreign-currency-asset-holding elite. The Ethiopian elite that has held substantial portions of its wealth in dollars (in foreign accounts, in foreign property, in foreign-currency-denominated assets onshore) has been partially hedged against the monetary disorder. The disorder has imposed real welfare costs on Ethiopian society but the costs have not fallen on this group with full force. This group also has substantial political influence, both through direct political activity and through the broader networks of business and family relationships that connect it to the ruling political coalitions.

Against these four constituencies, the populations who would benefit most from genuine reform — the rural poor, the urban formal-sector workers without political connections, the smallholder farmers, the pensioners — have, structurally, the least political voice. The Sowellian asymmetry of distribution that article 23 examined is, viewed from the political-economy angle, the same asymmetry. The constituencies who pay the cost cannot effectively organise to demand relief. The constituencies who receive the benefit can effectively organise to maintain the arrangement.

Why the regime changes did not change the equilibrium

The fact that three radically different regimes have produced similar monetary outcomes is, in this framing, intelligible. Each new regime arrived in power facing a fiscal situation in which: revenue could not be expanded quickly enough to meet expenditure ambitions; tax administration was insufficiently developed to support large discretionary expenditure increases; political pressures required maintaining specific subsidies and patronage flows; external financing was constrained by the country’s creditworthiness and by political tensions; and the central-bank mechanism was sitting there, available, as the residual financing option that no other instrument could provide.

The choice each regime faced was therefore not between alternative monetary policies in the abstract. It was between using the central bank to finance the residual fiscal gap or facing the political consequences of not meeting the expenditure obligations the political situation required. Given this choice, every regime — across very different ideologies — has made the same decision. The Derg used the central bank to finance its wars and its food subsidies. The EPRDF used the central bank to finance the GTP infrastructure programme. The post-EPRDF government used the central bank to finance the Tigray war. Each was making the same calculation under different circumstances. The calculation came out the same way.

This is the equilibrium claim in its purest form. The disease survives because, for the political coalition in power at any given moment, the alternative to using the central bank is worse than the costs of using it. The inflation that results is borne by populations without political voice; the alternative — visible fiscal discipline, sharp expenditure cuts, taxation of the politically-connected — would be borne by populations with political voice and would, with high probability, end the regime that imposed it.

This is also why external pressure has been the historical trigger for reform. The IMF programmes of 1992 and 2024, the Common Framework debt-restructuring process, the conditionality attached to bilateral aid — each of these has been an external constraint that forced the choice the political coalition would not otherwise have made. The political costs of refusing the external pressure (loss of financing, isolation from international markets, accumulating macroeconomic crisis) have, in each case, eventually exceeded the political costs of accepting the reform. But the trigger has always been external. The domestic political coalition has not, on its own, demanded reform.

The federal-ethnic dimension

A specific feature of the Ethiopian political economy worth naming, because it complicates the standard political-economy analysis, is the ethnic-federal dimension that the 1995 Constitution created.

The federal architecture distributes some fiscal authority to the regional states (the federal block grants are the principal vehicle), each of which has its own political dynamics and its own relationships with the federal government. The monetary disorder has been, in some respects, a federal phenomenon — the inflation is national, the exchange-rate regime is national, the central bank is federal — but the distributional consequences have varied by region. Regions with stronger access to federal patronage (typically those whose ethnic-political parties have stronger representation in the federal coalition) have, on average, fared better than those with weaker access.

This has political consequences that matter for the political economy of reform. A reform that disciplined the federal fiscal-monetary mechanism would, indirectly, change the distribution of patronage flows across regions. Regions that have benefited from the existing arrangement would, predictably, oppose the reform. Regions that have suffered most from it would, in principle, support the reform — but the regional political coalitions are themselves embedded in the federal-coalition arrangement and have limited capacity to demand a reform that would upset the broader patronage equilibrium.

This is a deeper layer of the political-economy puzzle. The federal-ethnic architecture, designed in 1995 to manage the diversity of post-Derg Ethiopia, has created political mechanisms that systematically reinforce the central-monetary architecture this series has been critiquing. The federalism and the monetary disorder are not separate phenomena. They are linked by their shared dependence on federal-patronage mechanisms that the monetary disorder enables.

The intellectual layer

A third layer worth naming is the intellectual one. The political-economy analysis above explains why the political coalition in power chooses the central-bank-financing option; it does not, by itself, explain why the intellectual class has, across the same three regimes, largely defended (rather than criticised) the choice.

The defence has shifted over time. In the Derg years, the central-bank financing was defended in Marxist-socialist terms as the mechanism by which the workers’ state mobilised resources for industrial development. In the EPRDF years, it was defended in developmental-state terms as the mechanism by which the state filled the gap left by underdeveloped capital markets. In the post-2018 period, it has been defended in pragmatic terms as the mechanism by which the country navigated successive emergencies. The vocabulary has shifted; the defence has remained.

The Ethiopian intellectual class that has provided this defence is not, in any caricatured sense, naive or corrupted. It includes serious economists who have, in good faith, argued for the developmental-state framework, the gradualist transition, the heterodox approach to monetary policy. The arguments have intellectual substance — much of which articles 11 and 26 have taken seriously. The series’s editorial position is not that the arguments are foolish; it is that they have not been tested empirically against the Ethiopian outcomes they have been used to defend. An intellectual position that defends a policy regardless of what the data show is, in the relevant sense, no longer empirical — it has become apologetic. The Ethiopian developmental-state defence has, in places, drifted into apologetics.

The series’s argument is that the empirical record now permits — indeed requires — a different reading. Three regimes, seventy-five years, comparable institutional choices, comparable monetary outcomes — these are sufficient data to draw the conclusion that the architecture does not deliver what its advocates claim. The intellectual defence of the architecture should, given the data, have evolved. It has not. The persistence is, in part, ideological inertia, and in part the result of the political-economy reality that intellectuals who depend on state patronage for their academic positions are not, structurally, in a position to challenge the state’s monetary architecture.

The implication

The article’s implication is that reform that lasts will require changing the political-economy configuration, not just changing the policies. This is the deepest pessimism — and the deepest realism — of the series.

The post-2024 reforms, on this analysis, are vulnerable in specific ways. They are vulnerable to political pressure from the constituencies who benefited from the previous regime, because those constituencies have not been disempowered by the reform itself. They are vulnerable to the next external-financing shock, because the political coalition has not internalised the discipline as its own commitment but rather accepted it as an external constraint. They are vulnerable to the next major war, because the war-finance pattern article 18 examined remains the path of least political resistance.

What would change the configuration? Three things, the series argues, in increasing order of difficulty.

First, visible inflation that destroys household purchasing power across politically organised constituencies, in ways that the elite cannot fully hedge against. The 2022–24 inflation episode came close to this; whether it produced the political mobilisation necessary to demand sustained reform is the open question of the next several years.

Second, the development of a domestic capital market deep enough to provide an alternative to the central-bank financing of fiscal pressures. The Ethiopian Securities Exchange is the institutional foundation; whether it grows fast enough to provide the alternative within a politically relevant timeframe is uncertain.

Third, the emergence of a political coalition that takes monetary discipline as a core commitment rather than as an external constraint. This is the hardest to see emerging. It would require a constituency, organised around a specific political programme, that prioritises low inflation and currency stability as policy goals worth fighting for. No such coalition currently exists in Ethiopian politics. Whether one will emerge in the next decade is, in the series’s editorial view, the central question for the future of Ethiopian monetary policy.

The next article turns to the strongest version of the case against the prescription this series has been building — the developmental-state defence at its full intellectual strength, taken on its own terms and engaged.

---

This article draws on the political-economy literature on Ethiopia, including: Sarah Vaughan, “Revolutionary Democratic State-Building,” Journal of Eastern African Studies 5, no. 4 (2011); Sarah Vaughan and Mesfin Gebremichael, “Rethinking Business and Politics in Ethiopia,” APPP Research Report 2 (2011); Christopher Clapham, The Horn of Africa: State Formation and Decay (Hurst, 2017); Tom Gardner, The Abiy Project (Hurst, 2024).

On the political economy of monetary policy in developing countries generally, see: Alberto Alesina and Allan Drazen, “Why Are Stabilizations Delayed?” American Economic Review 81, no. 5 (1991); Adam Posen, “Declarations Are Not Enough: Financial Sector Sources of Central Bank Independence,” in NBER Macroeconomics Annual 10 (1995); Christopher Adam, Stephen O’Connell, and Edward Buffie, “Aid, Exchange Rates and Trade Liberalisation in a Two-Sector Model,” Journal of African Economies 17, supplement (2008).

For the comparative-political-economy literature on developmental states, see articles 11, 12, and 26 references.