A Tax on the Poor: The Distributional Arithmetic of Ethiopian Inflation

Who paid

Across the seventy-five years this series has covered, Ethiopian inflation has averaged in the low single digits, the high single digits, the low double digits, the high double digits, and (in Derg years and in 2022) the low to mid-thirties. The cumulative loss of purchasing power for the Ethiopian birr against any stable basket of goods has been enormous. Two birr in 1965 bought what perhaps 1,500 birr buys in 2026 — a depreciation of roughly 750-fold over six decades.¹

This article asks: who paid this depreciation? Inflation is not a neutral tax. It falls with different weight on different parts of the income distribution, on different occupational groups, on people with different patterns of asset-holding. The distributional consequences of Ethiopia’s monetary history are, the article argues, regressive in specific identifiable ways — and the regressivity matters for the moral evaluation of the monetary architecture the series has been critiquing.

The argument is: every Ethiopian inflation surge (1984, 1991, 2008, 2011, 2022, 2024) has fallen hardest on the population least able to hedge against it — wage workers paid in birr, smallholders dependent on local market prices, the urban poor, pensioners. The same monetary system that imposed the implicit tax on these groups channelled the resulting transfer to constituencies who could insulate themselves: state-owned enterprises receiving subsidised credit, holders of foreign-currency assets, the politically-connected private sector. This is a regressive transfer in plain sight, and the post-float real-wage data make the point unanswerably.

Why inflation is regressive

The general theory is worth setting out before the specific Ethiopian numbers. Inflation as a tax has three properties that determine its distributional incidence.

First, inflation taxes assets held in the domestic currency. A household whose savings are denominated in birr loses purchasing power as the birr loses value. A household whose savings are denominated in dollars, gold, real estate, or other inflation-hedging assets is partially or fully protected. The distribution of asset-holding in any society is unequal: wealthier households hold more diversified portfolios and more inflation-hedging assets; poorer households hold more of their (smaller) wealth in cash and birr deposits. Inflation therefore falls more heavily on the poor as a proportion of their wealth, even though wealthier households may lose more in absolute terms.

Second, inflation taxes income that is not fully indexed. Wage workers whose salaries are set in birr and adjusted infrequently (often less frequently than the inflation rate) lose real income as prices rise faster than wages. Self-employed traders whose prices can be adjusted continuously suffer less. Workers in the formal sector with weak bargaining power suffer more than workers with strong bargaining power. The agricultural smallholder who sells his crop once a year at prices that may not reflect the year’s inflation experience suffers more than the urban trader who can adjust prices weekly.

Third, inflation taxes the goods that constitute the consumption basket of the poor. Food prices typically lead the inflation cycle and rise faster than headline inflation in most inflationary episodes. Imported goods (medicines, fuel, consumer goods) rise rapidly when the currency depreciates. The consumption basket of the poor is concentrated in food, fuel, and basic goods; the consumption basket of the wealthy is more diversified and includes substantial discretionary expenditure. The poor therefore experience a higher effective inflation rate than the headline CPI suggests.

These three properties, taken together, mean that any inflation episode hits the poor disproportionately. The Ethiopian record shows the pattern in extreme form.

The Derg years

The cumulative purchasing-power loss of the Ethiopian birr through the Derg years was, on the most defensible reconstructions, in the range of 200–400 percent across the 1974–91 period.² The losses fell with particular force on three groups.

Wage workers in the urban formal sector. Salaries set in the 1970s and adjusted slowly through the 1980s lost the bulk of their purchasing power. A teacher’s salary that supported a comfortable urban household in 1975 supported, by 1990, a household on the edge of poverty in real terms.³ The protests against the Mengistu regime in its later years were, in significant part, driven by salaried urban professionals whose real income had been destroyed by inflation while the regime maintained their nominal wages at fixed levels.

Pensioners. The Ethiopian pension system, established in the imperial period and continued through the Derg, paid benefits in nominal birr terms with infrequent adjustments. The real value of pension benefits declined sharply through the 1980s. Pensioners who had retired in the 1960s and 1970s with what had been adequate retirement provision found themselves, by the late 1980s, unable to meet basic expenses on the same nominal payments.⁴

Smallholder farmers in non-surplus regions. Farmers whose limited cash income came from periodic sales of small surpluses at prices the AMC was suppressing experienced the worst of the inflation. The AMC procurement prices, fixed in nominal birr terms, became increasingly inadequate compensation for the grain extracted from drought-affected regions. The famine of 1984–85 was, in this sense, the lethal endpoint of the regressive inflation tax: those least able to bear the cost bore it most severely, with the worst-off paying with their lives.

The EPRDF years

The EPRDF-era inflation pattern was more variable. The 1992 devaluation imposed a one-off real-income shock but was followed by years of relatively low inflation; the 2008 and 2011 spikes were sharp but were partially reversed; the GTP-era pattern was one of cumulative drift rather than acute episodes until 2017–18.

The 2008 episode is worth specific examination because it shows the regressive pattern in detail. Inflation in Ethiopia reached 64 percent year-on-year in mid-2008, with food inflation exceeding 90 percent.⁵ The episode was partly driven by global commodity-price shocks but, as article 14 examined, was substantially amplified by domestic monetary expansion. The distributional incidence was sharp.

Households dependent on imported food (urban populations, particularly the urban poor) saw their food expenditure share of total income rise from approximately 50 percent to over 70 percent.⁶ Households with foreign-currency savings (a small fraction of the population, concentrated in the elite) saw their wealth preserved or, in birr terms, increased. The Gini coefficient of household income, by some estimates, widened substantially through the episode, with much of the widening attributable to the differential ability of households to hedge against the inflation.

The 2011 episode showed a similar pattern. The 2022 episode showed it again, in even more extreme form.

The 2022–24 episode

The most recent inflation surge, driven by the combination of Tigray war finance, global commodity shocks, and the post-float devaluation, has been the most severe in the post-Derg period. Inflation peaked at 33 percent in 2022 and remained above 20 percent through much of 2023 before falling to single digits in late 2025 and re-accelerating to 13.4 percent by May 2026.⁷ Food inflation was higher still, peaking above 40 percent in some months.

The distributional consequences have been documented in real-time more than any previous Ethiopian inflation episode. The Reporter Ethiopia’s January 2026 analysis, “Beyond The Numbers: Ethiopia’s Cost-of-Living Crisis Persists,” captures the central point: the December 2025 headline single-digit inflation figure did not reflect the experience of households whose consumption was concentrated in food, where prices remained sharply elevated.⁸

The Addis Standard’s March 2026 analysis, “Birr in Freefall,” makes the comparable point about the exchange-rate component: that the post-2024 depreciation is “not only larger in magnitude but also faster and more destabilizing, carrying far-reaching consequences for ordinary Ethiopians.”⁹ The article notes that the macroeconomic reform programme’s positive outcomes — exchange-rate market unification, increased reserves, IMF re-engagement — have been “accompanied by short-term inflationary pressures” that have fallen disproportionately on Ethiopian households dependent on imported goods.

The real-wage data, where available, are stark. Ethiopian real wages in the formal urban sector, in 2026, are estimated to be approximately 30–40 percent below their 2019 levels in purchasing-power terms.¹⁰ A Ethiopian household whose monthly income in 2019 supported a particular consumption pattern requires, in 2026, between 60 and 70 percent more nominal birr to maintain the same consumption — and many households have not had nominal income increases of that magnitude.

The remittance arbitrage

A specific feature of the 2022–24 distributional pattern worth naming is the remittance arbitrage that emerged. Ethiopian households with relatives abroad sending remittances were partially shielded from the inflation because the remittances were typically denominated in foreign currency, with the recipient’s birr purchasing power determined by the exchange rate at the time of receipt. As the parallel exchange rate accelerated through 2022–23, the birr value of a given dollar remittance rose; households receiving remittances saw their purchasing power decline less than households dependent purely on domestic income.

The effect was to partially exempt the diaspora-connected households from the inflation tax that the rest of the population was paying. The diaspora connection became, in effect, a hedge that some households had and others did not. The distribution of diaspora connections in Ethiopian society is not random: it correlates with education, with urban location, with prior wealth (because emigration requires resources), and with specific ethnic and regional patterns (the Eritrean and Tigrayan diasporas are particularly large relative to home-country populations; the Oromo diaspora is large in the US and elsewhere; the diaspora from rural southern and western Ethiopia is smaller).

This is one of the deep features of the Ethiopian distribution: the populations who paid the most for the monetary disorder — the rural poor, the urban formal-sector workers without diaspora connections, the pensioners — were precisely the populations with the fewest hedges against it. The populations with hedges — the diaspora-connected, the foreign-currency-asset-holding, the politically-connected — paid less.

The political-economy implication

The distributional analysis has a political-economy implication that the series has been building toward across multiple articles. The constituencies who pay the inflation tax are, structurally, the constituencies with the least political voice. The constituencies who benefit from the monetary disorder (or who hedge against it) are structurally the constituencies with the most political voice.

This is not a conspiracy theory. It is a description of the political economy of asymmetric distribution. The poor smallholder, the urban worker, the pensioner, the diaspora-disconnected household — these are populations whose individual welfare losses are real but small in absolute terms, who are dispersed geographically and politically, who have limited capacity to organise around the specific issue of monetary policy. The beneficiary constituencies — state-owned enterprises with cheap credit, foreign-currency-asset holders, politically-connected private firms — are concentrated, well-organised, and politically articulate.

The disease this series tracks survives, this is to say, because the people who pay for it cannot organise to demand its end. The political coalition for monetary discipline in Ethiopia is, structurally, weak. The political coalition for the maintenance of the existing arrangements is, structurally, strong. Each reform has come from external pressure (the 1992 reform from IMF conditionality, the 2024 reform from the Eurobond default and the IMF programme) rather than from domestic political demand for discipline. The pattern is durable because the political economy is structured to make it durable.

This is the Sowellian point at full strength. The economic policies that the Ethiopian state has run for fifty years are not the result of ignorance, of incompetence, or of evil intent. They are the equilibrium outcome of a political-economy structure that systematically rewards the policies and punishes the alternatives. Changing the policies, if it is to be more than the current external-pressure-driven adjustment, requires changing the political economy. The series will return to this in articles 25 and 27.

The honest counter-case

The strongest version of the counter-argument to this article is that the distributional analysis, however accurate, leaves out the gains that the monetary architecture funded. The state-led infrastructure investment, the rapid expansion of basic services, the poverty reduction achieved between 1995 and 2015 — these were real benefits that flowed disproportionately to the same populations that paid the inflation tax. The poor smallholder lost real income to inflation but gained access to schools, health clinics, roads, and electrification. The net distributional impact is, on this account, ambiguous rather than clearly regressive.

The argument has force. The development gains of the EPRDF period were real and were experienced disproportionately by the poor. The school enrolment rates that rose from 30 to over 80 percent, the child-mortality rates that fell sharply, the road network that connected previously isolated communities — these gains improved Ethiopian welfare in ways the inflation accounting alone does not capture.

The series’s reply is two-fold. First, the gains were gross benefits delivered, not net benefits after the inflation cost. Some part of what was gained was offset by what was lost to inflation, and the net figure for the worst-off populations was probably positive but smaller than the gross figure suggests. Second, the gains could have been delivered with a different monetary architecture — one that did not impose the inflation tax. A reform that maintained the development spending while disciplining the monetary financing would have produced larger net benefits for the same population. The series’s argument is not that the development gains were illusory; it is that they were achieved at a higher cost than was necessary.

The next article turns to that counterfactual. Other countries facing comparable challenges chose different monetary architectures and produced different distributional outcomes. What can Ethiopia learn from them?

---

Footnotes

1. World Bank, “Inflation, consumer prices (annual %) — Ethiopia,” cumulative calculation from World Development Indicators series 1960–2025. The cumulative figure is approximate; reasonable estimates vary by perhaps 20–30 percent depending on which baskets and which official-vs-parallel exchange rates are used. ↩

2. Alemayehu Geda, “The Macroeconomic Performance of the Ethiopian Economy,” in Befekadu Degefe and Berhanu Nega, eds., Annual Report on the Ethiopian Economy, Vol. 1 1999/2000. ↩

3. Berhanu Nega, “The Economics of Ethiopia’s Transition,” in Edmond J. Keller and Donald Rothchild, eds., Africa in the New International Order (Lynne Rienner, 1996). ↩

4. Tegegne Gebre Egziabher and Meheret Ayenew, “Pension and Social Security in Ethiopia,” Forum for Social Studies, 2005. ↩

5. World Bank, “Inflation, consumer prices (annual %) — Ethiopia,” 2008 figure; National Bank of Ethiopia, Annual Report 2007/08. ↩

6. Ethiopian Central Statistical Authority, Household Income, Consumption and Expenditure Survey 2010/11, CSA. ↩

7. StockMarket.et, “Ethiopia’s Inflation Climbs to 13.4% as Food Prices Accelerate,” May 2026; Trading Economics Ethiopia inflation series. ↩

8. The Reporter Ethiopia, “Beyond The Numbers: Ethiopia’s Cost-of-Living Crisis Persists,” January 2026. ↩

9. Addis Standard, “Birr in Freefall: Ethiopia’s struggle with record currency depreciation, imported inflation risk,” March 2026. ↩

10. Estimate based on cumulative inflation 2019–2026 from World Bank WDI series, against typical formal-sector wage adjustment paths reported by the Ethiopian Federal Civil Service Commission and the major private-sector employer associations. ↩