Before the Birr: Salt, Silver, and the Maria Theresa Thaler

An argument by Zef Telahun

This is an editorial perspective — signed opinion, not the site's neutral analysis. Factual claims are footnoted; the synthesis, emphasis, and judgement are the author's.

Why begin so far back?

A series on modern Ethiopian monetary policy could reasonably start in 1974, with the revolution that gave the state full control of the currency for the first time. This series starts two centuries earlier, for one specific reason: Ethiopia has lived under sound commodity money, for longer than most countries in the world have lived under any kind of money at all, and the discipline of that older system is part of the story. Sound money for Ethiopia is not an alien import to be transplanted from Switzerland. It is a recovery of something the country had, and gave up.

The argument of this article is bounded and concrete: that for most of the eighteenth and nineteenth centuries, Ethiopia operated on commodity monies — salt bars and the Maria Theresa thaler — that were monetarily disciplined precisely because they sat outside state control, and that the modernising step of issuing a state currency was simultaneously a normal act of state-building and a quiet loss of the discipline the older system had imposed by default. The point is not nostalgia. The point is that the founding act of Ethiopian state-issued money already contained the institutional possibility of the disease this series is tracking. To understand what 1945 created, one has to understand what was given up.

The salt bars

Begin with amole tchew, the bars of salt cut in the Danakil depression that served as small-denomination currency across much of Ethiopia from at least the early modern period.¹ Travellers from the seventeenth century onward describe their use; nineteenth-century accounts give exchange rates against the thaler that varied between roughly eight and twelve bars per coin depending on the quality and freshness of the salt.² What the salt bars had — and this is the relevant feature for a series on monetary discipline — is that they could be produced only by cutting them out of natural deposits, transporting them along established caravan routes, and trading them at prices set by the market for what salt would buy.

In the language of modern monetary theory, the salt was a commodity money whose supply was set by the marginal cost of production. If too few bars were in circulation relative to demand, the price of salt rose, which made it worth more to cut and transport new bars, which expanded the supply until prices stabilised. If too many bars were in circulation, they got eaten — salt has, helpfully for monetary discipline, a non-monetary use. The system could not be inflated by political decision because no political authority controlled the depression or the trade. The Emperor of Ethiopia could no more print salt bars than he could print rainfall. The discipline was built in.

This is the feature, not a bug. The point is not that salt bars were a sophisticated monetary system — they were a primitive one, with high transaction costs and limited divisibility. The point is that the one discipline modern central banks struggle most to maintain, the discipline of not creating money faster than the economy can absorb, was provided automatically by the commodity nature of the medium.

The thaler

The larger story is the Maria Theresa thaler, the silver coin minted in Vienna from 1741 that became, by an accident of trade routes, the dominant currency of large transactions across Ethiopia, Yemen, the Red Sea, and much of East Africa for the next two centuries.³ The coin entered Ethiopia probably through Massawa and the coffee trade with Yemen and Arabia, was first recorded in circulation during the reign of Emperor Iyasu II (1730–1755), and dominated large-value transactions by the time the Scottish traveller James Bruce visited in 1768.⁴ Emperor Tewodros II formally adopted it as the standard coin in 1855, naming it the birr — the Amharic word for silver.⁵

The thaler’s monetary properties matter for the argument of this series. It contained a fixed weight of silver — 23.389 grams of 833 fine silver, or about 19.5 grams of pure metal — that did not change across two and a half centuries of restrikes.⁶ The Austrian mint continued striking coins dated 1780 (the year of the empress’s death) long after she was gone, precisely because the market valued the unchanging specification more than it valued a contemporary date. When the Austrian government tried to change the coin’s specifications or replace it, the East African market refused: traders would accept only the 1780 design. By 1858, the Maria Theresa thaler had lost its status as legal tender in Austria — but continued, undisturbed, as the currency of choice in Ethiopia.⁷

This is not a small fact. It is a near-perfect natural experiment in monetary discipline imposed by users rather than issuers. The Ethiopian market, by collectively refusing to accept any thaler other than the 1780 specification, was forcing the Austrian mint to honour a commitment about silver content that the mint had no legal obligation to honour. Joseph Kalmer and Ludwig Hyun estimate that of the roughly 245 million thalers minted between 1751 and 1931, more than 20 percent — over 49 million coins — ended up circulating in Ethiopia.⁸ When the British military expedition to Magdala in 1868 needed local currency to pay its expenses, it carried thalers.⁹ When Italian colonial authorities in Eritrea tried to introduce a tallero eritreo of their own design in 1890 to displace the Maria Theresa, the attempt failed.¹⁰ The market preferred the discipline of the proven coin to the convenience of the new one.

Two features of the thaler-based system are relevant to this series’s argument. First, the price level was anchored by the silver content of the coin. Inflation in the modern sense — a persistent rise in the general price level driven by monetary expansion — could not occur, because no Ethiopian authority controlled the supply of thalers. The supply expanded when more were imported through trade, contracted when they were hoarded or melted, and adjusted around the silver content the market insisted on. The Ethiopian price level fluctuated with the world silver price and with local harvest conditions, but it did not drift upward across decades the way modern fiat currencies do. The famines that periodically struck Ethiopia in this period — most notoriously the 1888–92 kifu qen famine — were terrible, but they were not monetary events. The Ethiopian peasant lost his children to crop failure, not to seigniorage.

Second, foreign exchange was not a state problem because there was no state currency to defend. A merchant in Harar selling coffee to a buyer in Aden did not need to navigate a foreign-exchange regime. He sold for thalers and used them. The convertibility that modern Ethiopian governments have spent fifty years rationing was, in the nineteenth-century system, simply the property of the money itself. There was no parallel market because there was no official rate to deviate from. There was no chronic shortage of foreign currency because the currency was foreign.

Menelik’s failed experiment

That all of this was about to be lost is visible in what happened next, and the episode is worth examining because it is the founding precedent.

Emperor Menelik II had considered issuing an Ethiopian currency as early as 1875, when he was still King of Shewa.¹¹ After the Treaty of Wuchale in 1889 and the Additional Convention of 1 October 1889, the right to issue a national currency was negotiated; the coins were to be struck at the King of Italy’s mint to specifications matching the Maria Theresa.¹² Menelik’s silver talari, equivalent in weight and fineness to the thaler, were minted in Paris from 1893; 200,000 coins were produced in 1894 alone.¹³ The reasoning was straightforward and recognisable: a national coinage was a marker of sovereignty, particularly important after Adwa, in the same way postage stamps and a national flag were markers of sovereignty. European advisers were enthusiastic. So was the international scene of coin collectors.

The Ethiopian market refused them. The Menelik thaler never circulated in significant volume. Various explanations have been offered — that Ethiopians mistrusted unfamiliar coinage, that the profile portrait violated Ethiopian artistic convention, even that the population was sentimentally attached to the décolletage on the empress’s portrait.¹⁴ But the explanation that matters for this series is the one nobody at the time articulated as such: the market had a sound coin already, and it had no reason to switch. The discipline the Maria Theresa imposed by its unchanging specification could not be matched by a coin issued by an Ethiopian sovereign, however technically equivalent, because the Ethiopian sovereign could change his mind. The Austrian mint had committed, by inertia, to a specification it could not back away from without losing its East African business. Menelik had no such commitment device. The market knew the difference.

This is the moment the argument of this series turns. The founding of the Bank of Abyssinia in 1905, by concession from the British-occupied National Bank of Egypt, gave Ethiopia for the first time an institution with the legal authority to issue paper money.¹⁵ The bank issued its first notes in 1915. The notes were redeemable for silver coin, on demand, and circulated at par with the thaler — that is, the founding paper currency of Ethiopia was disciplined by convertibility into a metallic standard it did not control. As long as the convertibility was honoured, the discipline held. The institutional possibility of breaking the convertibility, however, was now present in the system. It had not existed before.

What was lost

The point of this article is not that pre-modern Ethiopia was a monetary paradise. It was not. The salt-and-thaler system worked, but it worked under conditions — limited monetisation of the economy, narrow government budgets, no centralised industrial economy needing complex credit — that the country could not stay in if it was to modernise. The decision to move toward a state currency was not a mistake. It was a normal step in nation-building, taken for normal reasons.

What is worth seeing clearly is what the move cost. Ethiopia gave up an automatic discipline — that no political authority could create money faster than the silver mines could be worked — and accepted in exchange the discretion of a state to manage its own money supply. The trade was justified by the belief that this discretion, exercised by responsible authorities, would produce better outcomes than the commodity-money system: easier credit, smoother adjustment to shocks, more flexible fiscal policy in emergencies.

The empirical question this series asks, in twenty-six more articles, is whether the trade has worked. The answer, this series will argue, is no — not because central banking is wrong in principle, but because the specific institutional choices Ethiopia made about its central bank, repeated across three regimes, surrendered the discretion to the executive of the day in a way that converted it from a stabilisation tool into a deficit-financing tool. The salt bars and the Maria Theresa thaler could not be debased because no one controlled them. The birr can be, has been, and continues to be debased, because the state controls it.

There is a serious version of the counter-argument: that commodity-money systems impose enormous deadweight costs in good times — limited credit, deflationary pressure on a growing economy, vulnerability to commodity-price shocks unrelated to local conditions — and that those costs are larger than the costs of occasional inflation under a managed currency. Robert Mundell made versions of this case across his career; the Bank for International Settlements has documented the deflationary tendencies of strict commodity standards.¹⁶ The series takes this seriously. The reply, developed in article 19, is that the choice is not between a nineteenth-century silver standard and unconstrained fiat. It is between unconstrained fiat and a rules-bound central bank operating with binding fiscal constraints, current-account convertibility, and political insulation. The salt bars had those properties by accident. A modern central bank has to be built to have them deliberately. Ethiopia has never built one that does.

That is the institutional gap. The series’s prescription, when it gets there, will be about how to close it. The historical chapters between this one and that one show why it has to be closed.

The next article turns to the imperial monetary order that Bank of Abyssinia made possible — what it got right, and what its stability concealed.