What the Empire Got Right (and What It Concealed): Money under Haile Selassie

A boring chapter, deliberately

If a country’s monetary history is going badly, the symptoms are dramatic: hyperinflation, currency collapse, queues at the bank. If it is going well, there is nothing to write about. By that standard, Ethiopian monetary history under Haile Selassie was going well. From the founding of the State Bank of Ethiopia in 1942 to the end of the imperial period in 1974, the birr held steady against the US dollar at 2.07, inflation averaged in the low single digits, and the parallel exchange rate — the symptom that haunts every subsequent chapter of this series — was a minor curiosity rather than a national institution.¹

This article argues that the stability was real but the foundation was narrow, and that the foundation gave way for reasons that were partly external (the dollar’s own difficulties after 1971), partly structural (the limits of an economy in which most peasants were not yet monetised), and partly self-inflicted (a fiscal expansion the imperial government did not match with revenue). Understanding what propped up the stability matters because it tells us what would have been required to keep it going — and what the Derg would have to dismantle before its own monetary disorder could begin.

The article is, in other words, a portrait of the system the Derg inherited. The Derg’s economists, when they took power in 1974, did not invent monetary chaos out of nothing. They demolished a working system. To understand the demolition, you have to see the system first.

The institutional architecture, 1942–1963

The modern Ethiopian monetary system dates from the proclamation of 23 July 1942 that established the State Bank of Ethiopia as both commercial bank and central bank, and the parallel proclamation of 1945 that introduced the Ethiopian dollar — later renamed the birr — as the country’s sole legal tender at a fixed gold parity.² The arrangement was, by the standards of the time, conservative: the new currency was backed by reserves held in sterling and dollars, the State Bank had a statutory limit on the amount of credit it could extend to the government, and the exchange rate was fixed by international agreement under the Bretton Woods system to which Ethiopia, as an original signatory of the IMF Articles of Agreement, was committed.³

The 1963 reform separated the central-bank and commercial-bank functions, creating the National Bank of Ethiopia as a stand-alone central bank and converting the commercial side into the Commercial Bank of Ethiopia.⁴ The NBE’s mandate was, on paper, the standard one for the period: maintain the value of the currency, regulate the banking system, manage foreign exchange reserves. The statute included provisions limiting NBE lending to the government — though, as in most central-bank statutes of the era, the limits were less binding than they appeared.

What is worth seeing clearly is that this architecture, viewed in isolation, was not obviously superior to what came after. The NBE was an organ of the imperial government, its governor appointed by the Emperor, its mandate broad rather than narrow, its independence informal rather than statutory. It would have failed any modern central-bank-independence index. What kept it from being captured by fiscal demands was not its institutional design but the relatively modest fiscal demands placed on it.

This is the first feature of the imperial monetary order worth naming: it was disciplined by a small state, not a strong central bank. The Ethiopian central government’s expenditure averaged around 12–15 percent of GDP through the 1960s — modest by international standards even for the era.⁵ The military took a large share of that, particularly after Eritrean federation became Eritrean annexation in 1962 and the insurgency began, but the absolute size of the budget remained limited by the country’s tax base. There was simply not enough revenue, and not enough borrowing capacity, to require monetary financing of deficits. When the discipline came under pressure in the late 1960s and early 1970s, the system showed strain.

What stability bought

The Ethiopian birr held its 2.07 parity to the dollar for thirty years.⁶ What did that buy the country?

It bought, first, a coherent price system. Inflation averaged in the low single digits through the 1950s and most of the 1960s — World Bank data, where comparable, show Ethiopian consumer prices rising at roughly 2 to 4 percent annually through this period.⁷ By the standards of any developing country, that is good performance. By the standards of what was coming, it is remarkable.

It bought, second, a functioning if shallow banking system. Commercial credit was available, savings accounts paid positive real interest, and the parallel exchange market — though it existed — was a minor feature of the economy rather than its central organising principle. A trader importing goods could get foreign exchange at the official rate, through approved channels, with manageable delays. The forex rationing committees that became central institutions of Ethiopian economic life from the 1980s onward did not exist in the 1960s because they were not needed.

It bought, third, a credible currency for cross-border trade. The birr’s convertibility into dollars at a stable rate meant that Ethiopian coffee exporters and importers of capital goods could plan over horizons longer than the next quarter. The country’s integration into the international payments system, mediated by the IMF, gave it access to short-term liquidity through standard mechanisms — the Compensatory Financing Facility for coffee-price shocks, standby arrangements for balance-of-payments support — that did not require the country to grovel for emergency loans.⁸

These are not nothing. They are, in fact, the basic preconditions for the kind of long-horizon economic activity that pulls poor countries out of poverty. The empire did not deliver the growth necessary to do that, for reasons that are mostly not monetary. But the monetary foundation it provided was, on its own terms, sound.

What stability concealed

So why did this not last? Why, when the revolution came in 1974, did the new government not just inherit a working monetary system and run it?

Three answers, in order of weight.

The first is that the stability was narrow. It rested on a fiscal regime that took relatively little from the economy and spent relatively little on it. Government revenue averaged around 10–12 percent of GDP through the 1960s, against an OECD average above 30 percent.⁹ The Ethiopian state was, by international comparison, a small state. Its monetary discipline was easy to maintain because there was not much fiscal pressure on it. When the revolution brought a state that wanted to do more — nationalise industries, run state farms, fight insurgencies, maintain a vastly expanded military — the small-state fiscal base could not finance the large-state ambitions. The 2.07 peg could survive a modest government. It could not survive an immodest one without an immodest expansion of tax revenue, which the Derg never achieved.

The second is that the stability was externally subsidised. American military and economic aid to Ethiopia in the 1950s and 1960s was substantial, particularly after the 1953 Mutual Defense Assistance Agreement; through the 1960s, US aid alone covered a meaningful share of Ethiopia’s foreign-exchange needs.¹⁰ The collapse of this aid relationship in the 1970s — first with strains over Eritrea, then decisively with the 1977 break following the Derg’s Soviet alignment — removed an external prop that the official numbers had concealed. The 2.07 rate had looked sustainable in part because aid inflows were filling the gap between what the economy earned in dollars and what it spent. When the aid was withdrawn, the gap became visible.

The third is that the stability was vulnerable to the dollar’s own troubles. The Bretton Woods system, on which the 2.07 peg depended, collapsed in stages between 1971 and 1973. Nixon’s closing of the gold window in August 1971 ended the dollar’s gold convertibility; the Smithsonian Agreement of December 1971 attempted to restore fixed parities at new levels; the system finally broke down in March 1973 when the major currencies floated. Ethiopia, like most developing countries, had to choose a new anchor. It chose to maintain the dollar peg, but at a moment when the dollar itself was depreciating against gold and against major currencies.¹¹ The Ethiopian birr was now, technically, fixed to a floating currency — a regime no longer enforced by the international monetary order but only by the NBE’s willingness to defend it. The defence required reserves the NBE increasingly did not have.

The famine the empire could not see

The 1972–73 Wollo famine, which killed an estimated 200,000 people and provided the political fuel for the 1974 revolution, was not a monetary event in the technical sense.¹² Its causes were drought, the collapse of the grain trade with affected areas, and the imperial government’s failure to acknowledge the disaster in time to mobilise relief. But it is worth treating in this chapter because of what it reveals about the information system of the late imperial economy — the part that Hayek’s framework predicts will be the most consequential.

What the Wollo famine showed was that the imperial government had no functional system for transmitting information about food scarcity from the affected regions to the centre. The price of grain in Wollo had risen sharply through 1972 — the market was signalling. But the imperial bureaucracy did not act on the signal, because the signal was happening in a part of the country the centre did not bureaucratically penetrate. Foreign journalists discovered the famine before the Imperial government did, or admitted it had.¹³ The political consequences were fatal: the disclosure of the famine, and the contrast between the suffering in Wollo and the imperial court’s wealth in Addis, did more than any other single event to delegitimise Haile Selassie’s rule.

The Derg’s economic policies, which the next several articles examine, were in part a reaction to this informational failure. The new regime promised to do better — to know the country, to plan it, to feed it. The promise was sincere. The problem, as the Hayekian framing of this series will argue, is that the proposed solution — central planning, fixed prices, state procurement — worsened the informational problem rather than solving it. A famine the imperial government failed to see became, twelve years later, a famine the Derg’s policies were causing. The Wollo famine was a failure of perception under a weak state. The 1984–85 famine was a failure of policy under a strong one. Both reflect, in different ways, the cost of mismatched information and authority.

The steelmanned imperial defence

A serious case must answer the most serious objections. The most serious objection to this article’s framing is that the imperial monetary order was not really Ethiopian at all — it was a foreign-imposed system, maintained by foreign aid, designed for foreign trading partners, that did very little to extend monetary services to the actual Ethiopian population. The peasantry, three-quarters of the population in 1974, was barely monetised at all. The smooth NBE balance sheet was a fiction enabled by the smallness of the formal economy. Real Ethiopian economic activity happened in barter, in equb and iddir informal credit and insurance societies, in salt and produce exchanged outside the formal system. The state currency served Addis, the urban economy, the export trade, and the foreign embassies — and not much else.¹⁴

This is largely true. The series’s reply is that this is the reason monetary discipline was not the binding constraint on Ethiopian development, but it is not the reason to abandon discipline. The empire’s monetary stability was real but narrow; what was needed was an institutional architecture that extended the discipline to the rest of the economy while maintaining it. The Derg’s choice was to extend the state’s reach by destroying the discipline, which is not the only available choice and was not, this series will argue, the right one.

The next article turns to the Italian occupation and the 1945 reform that, more than any other single act, founded the institutional structure within which the rest of the disease incubated. The point is to see how the founding act already contained — in the form of the NBE’s statutory ability to lend to the government — the mechanism by which monetary stability would later be lost.

---

Footnotes

1. National Bank of Ethiopia, “Bank Notes,” https://nbe.gov.et/about-us/bank-notes/, recording the introduction of the modern birr and the 2.07 parity. World Bank, “Official exchange rate (LCU per US$, period average) — Ethiopia,” World Development Indicators, https://data.worldbank.org/indicator/PA.NUS.FCRF?locations=ET, confirming the 2.07 rate held continuously from 1949 to October 1992. ↩

2. NBE, “Bank Notes,” and Bahru Zewde, A History of Modern Ethiopia, 1855–1991, 2nd edn (Ohio University Press, 2001), chapter 8. ↩

3. International Monetary Fund, “Ethiopia and the IMF,” country page, https://www.imf.org/en/Countries/ETH, noting Ethiopia’s status as an original IMF member (joined December 1945). ↩

4. NBE, “About Us,” historical timeline; Eshetu Chole, Underdevelopment in Ethiopia (Organization for Social Science Research in Eastern and Southern Africa, 2004), chapter 4. ↩

5. World Bank, World Development Indicators, Ethiopia government expenditure series 1960–1974; Bahru Zewde, A History of Modern Ethiopia, 197–204. ↩

6. Mequanint B. Melesse and Sileshi Yitayih, “The Effects of Currency Devaluation on Ethiopia’s Major Export Commodities,” Cogent Economics & Finance 11, no. 1 (2023), https://www.tandfonline.com/doi/full/10.1080/23322039.2023.2184447, documenting the unbroken 2.07 parity from 1949 to 1992. ↩

7. World Bank, “Inflation, consumer prices (annual %) — Ethiopia,” World Development Indicators, https://data.worldbank.org/indicator/FP.CPI.TOTL.ZG?locations=ET. ↩

8. International Monetary Fund, Ethiopia: Recent Economic Developments, IMF Country Reports of the 1970s, available through the IMF archives. ↩

9. Eshetu Chole, Underdevelopment in Ethiopia, chapter 4, citing imperial budget records. ↩

10. David A. Korn, Ethiopia, the United States, and the Soviet Union (Croom Helm, 1986), chapters 2–3. ↩

11. Barry Eichengreen, Globalizing Capital: A History of the International Monetary System, 2nd edn (Princeton University Press, 2008), chapter 4. ↩

12. Mesfin Wolde-Mariam, Rural Vulnerability to Famine in Ethiopia, 1958–1977 (Vikas Publishing House, 1984); Jonathan Dimbleby, The Unknown Famine: Ethiopia, 1973, ITV documentary (1973). ↩

13. Dimbleby, The Unknown Famine; the documentary’s broadcast in October 1973 forced international recognition of the famine and was a major contributor to the regime’s collapse. ↩

14. Dejene Aredo, “The Iddir: A Study of an Indigenous Informal Financial Institution in Ethiopia,” Savings and Development 17, no. 1 (1993): 77–90; Daniel Ayalew Ali, Klaus Deininger, and Markus Goldstein, “Environmental and Gender Impacts of Land Tenure Regularization in Africa,” Journal of Development Economics 110 (2014): 262–275. ↩