Land, Collateral, and the Missing Credit Market

A constitutional choice with monetary consequences

Article 40(3) of the 1995 Ethiopian Constitution reads: “The right to ownership of rural and urban land, as well as of all natural resources, is exclusively vested in the State and in the peoples of Ethiopia. Land is a common property of the Nations, Nationalities and Peoples of Ethiopia and shall not be subject to sale or to other means of exchange.”¹

This sentence — twenty-six words in the English translation, slightly fewer in the Amharic original — is the single most consequential provision in modern Ethiopian economic law. It encodes a decision the Derg made in 1975 and the EPRDF preserved in its founding constitution: that land in Ethiopia is not private property. It cannot be bought or sold. It cannot be inherited freely. It cannot, decisively for this series, be mortgaged. The land you live on, the land you farm, the land your business is built on — none of it can serve as collateral for a loan from a bank.

This article argues that this constitutional choice, which is usually framed as a question of social or agricultural policy, is in fact one of the central determinants of Ethiopia’s monetary disorder. The absence of land collateral makes the formal banking system structurally unable to lend at scale to the majority of the population. The absence of bank lending to the majority of the population forces the state banking system to substitute for what private finance would otherwise do, which deepens the financial-repression mechanism described in article 12. The absence of a deep private credit market keeps the informal economy and the parallel currency market large, which keeps the official monetary order weak. The land question and the money question are not separate questions. They are the same question, viewed from different angles.

What banks need from collateral

To see why land matters, start with the basic operation of a private banking system. A bank takes deposits from savers, pools them, and lends them to borrowers. The lending is risky: some borrowers will fail to repay. The bank manages this risk through several mechanisms, but the central one — across every functioning banking system in the world — is collateral. The borrower pledges an asset; if the borrower defaults, the bank seizes and sells the asset; the proceeds repay the loan. The arrangement allows the bank to lend at lower interest rates than it would if it were taking pure credit risk, because the downside is bounded by the value of the collateral.

What makes an asset good collateral? It needs to be specific and identifiable; transferable from one owner to another; valuable enough to cover the loan; and protected by a legal system that will, in fact, enforce the bank’s claim on it if the borrower defaults. Land is, in most economies, the canonical good collateral: it is geographically specific, legally identifiable through title registration, transferable through sale, durable, and (in most market economies) of stable or appreciating value.

The Ethiopian system fails on the second criterion in particular. Land cannot be transferred from one owner to another by sale because land cannot be sold. What can be transferred is the usufruct right — the right to use a particular piece of land for agricultural or residential purposes — and even this transfer is restricted: in rural areas, usufruct rights are typically allocated by peasant associations or kebeles and can be revoked or reallocated; in urban areas, the leasehold system allows transfer of leasehold rights under specific conditions but not freely.

What this means in practice is that a smallholder farmer cannot pledge his land to a bank to secure a loan to buy improved seeds or a tractor or fertilizer, because the bank cannot, in the event of default, foreclose on the land. A small urban shopkeeper cannot pledge the property his shop sits on, because he does not own the property; he holds a leasehold from the city. A medium-sized commercial farmer cannot pledge his farm, even if he has a long-term lease, because the lease is, in important legal respects, contingent on continued state approval.

The result is a banking system that cannot lend to the rural majority of the population on standard collateralised terms. The system either does not lend at all (which is what happens for most smallholders), or it lends on the basis of substitutes for collateral — group guarantees, character lending, government guarantees, restrictive covenants — that are workable in small-scale microfinance but do not scale to the levels of credit that a modernising economy requires.

What the data show

The empirical consequence is visible in any cross-country comparison of credit depth. Ethiopian domestic credit to the private sector, expressed as a percentage of GDP, was approximately 21 percent in 2022 — among the lowest in sub-Saharan Africa.² By comparison, Kenya’s figure was approximately 32 percent, Vietnam’s (a relevant Asian comparator) was approximately 134 percent, and South Korea’s was approximately 165 percent.³ Ethiopia’s banking system, in other words, is shallow not because Ethiopian banks lack deposits — total deposits in the banking system are substantial — but because they lack creditworthy borrowers to lend to. The collateral problem is the binding constraint.

The composition of what Ethiopian banks do lend is equally revealing. Of the credit that does flow through the formal banking system, a disproportionate share goes to large urban firms, state-owned enterprises, and a narrow set of politically-connected private borrowers. Lending to small and medium enterprises, to rural farmers, to women-headed households — the populations a healthy banking system in a low-income country should be reaching — is minimal as a share of total lending.⁴ The Ethiopian banking system is, in effect, a wholesale banking system serving a small set of large institutional borrowers, sitting on top of a country where the majority of the population is unbanked or has only deposit accounts.

The microfinance sector, which expanded substantially through the 2000s and 2010s, has filled some of the gap. Ethiopian MFIs — including ACSI in Amhara, OCSSCO in Oromia, and DECSI in Tigray — have substantial portfolios and have reached millions of rural households.⁵ But MFI lending operates at small scales (typical loans of 5,000 to 50,000 birr), at high interest rates (often 15 to 25 percent), and is not a substitute for the credit a functioning banking system would provide. The MFIs are doing important work; they are not replacing the missing credit market.

The political logic of the land regime

Why has the constitutional commitment to state ownership of land survived three regimes and three constitutions? Why was it not on the table for reform in 1991, despite the new regime’s substantial liberalisation in other areas? Why is it not on the table in 2026, even with comprehensive macroeconomic reform underway?

The political logic is worth examining because it tells us something important about why the disease this series tracks is so durable.

The land regime serves three political functions, each with its own constituency.

First, it serves as implicit guarantee against landlessness. In a country where the imperial-era memory of tenancy under exploitative landlords is still alive, the constitutional prohibition on land sale is, for many Ethiopians, a meaningful protection. If land cannot be sold, then a household in financial difficulty cannot be forced (or seduced) into selling its land and joining the urban poor. The fear of landlessness, in a country where rural-urban migration produces hardship and informal-sector precarity, is real, and the constitutional commitment addresses it directly.

Second, it serves as political infrastructure for the ruling party. The allocation of usufruct rights is, in practice, performed by local kebele and peasant-association officials who are, in most cases, members of the ruling party (the EPRDF, later the Prosperity Party). The party’s local presence is reinforced by its role in land allocation. Reforming the land regime — moving to private ownership, market-based allocation — would, mechanically, weaken the local political infrastructure of the ruling party. This is not a small consideration in a country where political control is, in many regions, exercised principally through these local structures.

Third, it serves as fiscal infrastructure for the state. Because land is state-owned, the state can re-allocate it for state purposes — for industrial parks, for commercial farms, for road and rail corridors, for the GERD reservoir — without the property-rights complications that would arise in a market-ownership system. The compulsory acquisition of land for development purposes has been a major mechanism by which the EPRDF-era state delivered its infrastructure programme. The cost has been imposed on the displaced, who have received compensation that has often been inadequate and sometimes not delivered at all.⁶ But the speed and scale of the land acquisition would not have been possible under a market-ownership regime.

Each of these political functions has costs. The protection against landlessness has the cost of foreclosing the credit market. The party-political infrastructure has the cost of politicising land allocation. The state fiscal infrastructure has the cost of displaced households without secure property rights. The series’s argument is not that the land regime is wholly wrong; reasonable people can disagree about the balance of costs and benefits. The narrower argument is that the monetary consequences of the land regime are systematically underweighted in Ethiopian policy debate, and that any serious reform of the monetary system has to confront the land question whether it wants to or not.

What reform would look like

There is a range of possible reforms that would address the monetary consequences of the land regime without abandoning the political commitments that have sustained it. The series mentions them here because the article 27 prescription will return to them.

One option is long-term, transferable leasehold. The Chinese system since 1988 has operated on this basis: land remains state-owned, but use-rights are granted for fixed long periods (40, 50, 70 years depending on use) and are freely transferable, including pledgeable as collateral. The system has produced both a functioning property-rights regime for credit purposes and the political flexibility of continued state ownership.⁷ An Ethiopian leasehold system of this form would not require constitutional amendment and could be implemented within the existing framework.

A second option is secondary-rights registration. The land would remain state-owned, but the use-rights would be formalised into a national registry sufficient to allow them to serve as collateral. The Ethiopian government has, since 2003, implemented a Rural Land Certification programme that has issued certificates to millions of farmers; the certificates have been shown, in research, to improve tenure security and to modestly increase investment.⁸ But the certificates have not been made pledgeable — they cannot serve as collateral for bank loans. Extending the certification programme to include pledgeability would be a substantial step.

A third option, more radical, is partial privatisation. The current regime could be maintained for the bulk of rural and urban land, with full private ownership permitted in specific categories (commercial agricultural land above a certain scale, urban real estate above a certain value). This would create a private market in the specific categories where the credit-collateral function is most binding, without abandoning the broader commitment to state ownership for the social-protection majority.

Each of these options has costs and trade-offs. The series does not advocate any one of them. The point is that the credit-market problem this article describes is not unsolvable; comparable reforms have been implemented in comparable countries; the Ethiopian state has the institutional capacity to implement one or another version. What is missing is the political will to confront the cost-benefit calculus honestly.

The connection to the rest of the series

Why does this article appear in a monetary history? Because the missing credit market is the deep institutional condition that makes the financial-repression machinery this series has been describing both necessary and ineffective.

It is necessary because, in the absence of a private credit market that can fund Ethiopian development at scale, the state-led credit system has to substitute. The CBE, DBE, and the directed-lending mechanisms of the EPRDF years were not unmotivated; they were responses to the real problem that the formal private banking system could not, given the collateral constraints, fund the country’s development on its own. The state had to do something. What it did was financial repression.

It is ineffective because the state-led credit system, however necessary as a substitute, does not produce the discipline that a functioning private credit market would. Loans get extended to projects that should not have been funded; non-performing loans accumulate without being recognised; the capital allocation that the credit system performs is driven by political rather than economic considerations. The state system is, in some ways, worse than a missing market because it produces the illusion of credit allocation without the discipline of it. A country with no banking system at all would, at least, not be accumulating bad loans at state banks.

The Ethiopian monetary disorder that this series tracks is, in this sense, downstream of the land question in a way that purely-monetary analyses miss. You cannot stabilise the birr without addressing the financial-repression machinery. You cannot end the financial-repression machinery without giving Ethiopian banks something else to lend on. You cannot give Ethiopian banks something else to lend on without reforming the collateral regime. You cannot reform the collateral regime without confronting the land question. The chain runs all the way through, and the series’s argument is that policy reform that does not address every link in the chain will continue to fail.

The next article turns to the period when this whole architecture was most loudly celebrated and the underlying disorder most aggressively concealed: the GTP years, 2010 to 2020.

---

Footnotes

1. Constitution of the Federal Democratic Republic of Ethiopia, 1995, Article 40(3). ↩

2. World Bank, “Domestic credit to private sector (% of GDP) — Ethiopia,” World Development Indicators, https://data.worldbank.org/indicator/FS.AST.PRVT.GD.ZS?locations=ET. ↩

3. World Bank, World Development Indicators, same series for Kenya, Vietnam, and South Korea. ↩

4. International Finance Corporation, Ethiopia Country Diagnostic (IFC, 2019); World Bank, Ethiopia Financial Sector Development Report (World Bank, 2019). ↩

5. Wolday Amha, “Microfinance Industry in Ethiopia: Performance and Challenges,” Association of Ethiopian Microfinance Institutions, 2018. ↩

6. Dessalegn Rahmato, The Peasant and the State: Studies in Agrarian Change in Ethiopia 1950s–2000s (Addis Ababa University Press, 2008), chapter 10; Daniel W. Ambaye, Land Rights and Expropriation in Ethiopia (Springer, 2015). ↩

7. Klaus Deininger and Songqing Jin, “The Potential of Land Markets for Productivity, Welfare, and Equity: Evidence from China,” World Bank Policy Research Working Paper 5345 (2010). ↩

8. Klaus Deininger, Daniel Ayalew Ali, Stein Holden, and Jaap Zevenbergen, “Rural Land Certification in Ethiopia: Process, Initial Impact, and Implications for Other African Countries,” World Development 36, no. 10 (2008): 1786–1812. ↩