Mali Office du Niger

A hydraulic project began by compelling people to inhabit it

French planners proposed that water diverted from the Niger River at Markala could turn an enormous plain into a source of cotton for French industry and rice for West Africa. Created in 1932, the Office du Niger combined engineering, settlement, crop rules, credit, procurement, and administration. Jean Filipovich's archival history estimates that colonial authorities forcibly resettled about 30,000 Africans by 1945 after the scheme's architect treated the absence of voluntary settlers and a viable cotton crop as obstacles coercion could overcome.¹

The dam and main canals outlasted empire; the original production theory did not. Settlers carried debt and production risk while officials prescribed crops and supervised conduct. Cotton performance disappointed, and the realized irrigated estate was far smaller than the projected command area. The colonial record therefore joins technical infrastructure to forced labor, involuntary movement, and constrained refusal. Counting hectares without that history would misclassify coercive capacity as neutral organizational inheritance.¹²

Independent Mali retained the Office as a strategic public institution after 1960. The state continued centralized production, processing, marketing, land, and water control before abandoning cotton and concentrating on rice. By the late 1970s and early 1980s, yields, canals, and fields had deteriorated, while fixed procurement and agency monopolies gave farmers weak incentives and little voice. A World Bank political history describes a coalition in which government protected urban rice supply and agency employment while cultivators lacked a recognized place in management.²

Reform worked by narrowing the agency and enlarging farmer discretion

Reform began through pilot rehabilitation and institutional bargains rather than one complete blueprint. Fields were leveled, canals rehabilitated, credit and input arrangements changed, marketing and milling opened to private actors, and farmers gained more ability to choose, sell, and organize. Paddy yields rose from about 1.5 tonnes per hectare toward 6 tonnes in the development record, but the same source attributes the change to combined physical, technical, market, and institutional investment rather than to privatization alone.²³

The Office's current official mandate centers on water management, infrastructure maintenance, delegated project development, land administration, rural advice, and technical assistance. It plans through performance contracts with the state and producers rather than operating the whole rice economy.⁴ That boundary is consequential: a common hydraulic network still needs a public operator, while planting, learning, milling, and trade do not require one agency to command every transaction.

Farmer participation also became operational. Development reviews describe elected delegates on joint committees that decide annual maintenance programs, budgets, and procurement while the Office retains day-to-day system management. Village water research found that formal rules and informal negotiations coexist at tertiary levels; access to decision makers, canal position, social relations, and practical knowledge affect what a written allocation means.³⁵ Participation therefore changed the balance without erasing hierarchy.

Farmer field schools made local expertise visible and exposed unequal access

An FAO study followed 116 participants across 23 integrated-pest-management field schools and revisited villages after training. Farmers continued plot experiments, compared water, spacing, fertilizer, pest, cost, and yield outcomes, taught peers, and sometimes invited Office extensionists into their presentations. The result is evidence of distributed learning rather than a claim that every field school or village performed equally.⁶

The same study identifies an authority boundary inside participation. Most cooperative members and exploitation-right holders were male household heads; women were underrepresented even though they supplied transplanting, milling, and vegetable labor. Well-organized villages had better access to extensionists and outside projects than socially weaker villages. Farmer-centered learning could strengthen professional autonomy while inheritance, registration, literacy, time, and network position still determined whose expertise entered formal decisions.⁶

This pattern distinguishes delegation from abandonment. Farmers need discretion to test and choose. They also depend on coordinated main-canal operation, drainage, maintenance, land records, and drought rules. The productive question is not whether the state or the farmer should do everything. It is which level has the information, capability, and affected-party standing for each decision, and where an appeal travels when those judgments conflict.

Use permits improved production without settling tenure authority

The Malian state owns land in the developed scheme. A 2017 World Bank project record describes the Office distributing publicly developed plots to smallholders through land-use permits and investor-developed plots through long leases; it also notes shrinking average holdings, intense demand, rent-seeking risk, and politically connected acquisitions.⁷ Water-service fees fund a shared system and condition access. A productive household can therefore be treated as a partner in farming while remaining a permit holder in land and water administration.

This asymmetry became sharper as national authorities offered large areas to domestic and foreign investors after 2000. Peer-reviewed field and archival research reported hundreds of thousands of hectares in provisional or final allocations, closed negotiations that bypassed regional procedure, and projects whose perennial or dry-season crops implied year-round water claims. It also documented planned canals intersecting villages and cattle routes and modeled future competition between new investors and existing family farms.⁸

An Oakland Institute investigation adds interviews and site accounts from affected communities, including residents who said they learned of projects after decisions advanced, lost fields or routes, or lacked enforceable compensation and benefit commitments.⁹ Its advocacy position makes transparent why it sought these harms; it does not make each testimony representative or prove every contested legal claim. Read with the peer-reviewed allocation study, it supplies evidence the lease file alone cannot: how a formally available hectare may already support homes, cultivation, grazing, movement, or customary authority.

Water at Markala remains water in a downstream account

The Office's production account often begins at the intake and ends at a rice field. The river continues into the Inner Niger Delta, where flood extent and timing support fisheries, flood-recession farming, pasture, settlements, wetlands, and nonhuman habitat. Irrigation has high value precisely because it can withdraw water when rainfall or river flow does not meet crop demand. Expansion therefore creates an upstream–downstream allocation, not merely more efficient use of an unused stock.⁸

A 2026 peer-reviewed water-accounting study used satellite datasets to estimate the delta's water stocks and flows and explicitly modeled Office withdrawals. It found that irrigation expansion can materially change the hydrological balance without basin-wide coordination and recommended improving land and water productivity before continued area expansion. The study estimates and models flows rather than observing every livelihood or species outcome, and its climate and land-use scenarios remain uncertain.¹⁰

Legitimate allocation must therefore show more than planned hectares and expected harvest. It must publish the water budget, identify environmental-flow requirements, define priority during dry years, map existing land and movement claims, and specify who can challenge an allocation before irreversible work. Technical capability remains essential; the ethical error is to let capability decide which affected lives count.

Organizational profile

Authority combines state-bureaucracy, professional-expertise, the technical-substrate of Markala and its canals, and market-capital from buyers, lenders, donors, and investors. Decisions sit in a central-executive, a rule-bound hierarchy, and at the frontline-local level where farmers, cooperatives, extensionists, and water delegates manage fields and tertiary canals. Ownership is state plus a partnership-network: public land and infrastructure support permits, leases, farmer organizations, contractors, and commercial exchange.⁴⁷

Coordination uses hierarchy, standards, planning, markets, and mutual adjustment. Main-canal release and tenure remain hierarchical; maintenance, water fees, cropping calendars, and contracts provide standards; plans allocate construction and water; rice, input, and lease markets carry price signals; and field-level water requires negotiation. Knowledge moves top-down, bottom-up, through specialist staff, and through embedded practice.²⁵⁶

Measurement is financial, operational, and quality-oriented through water fees, budgets, irrigated area, delivery, canal condition, yields, crop quality, and field comparisons. Learning uses continuous improvement and market feedback; adaptation combines central reconfiguration, local iteration, and slow institutional change. The reform record supports all three while warning that a yield change alone cannot identify which mechanism caused it.²³⁶

The beneficiary profile includes the state and public, workers, communities, and customers. It records intended classes, not equal or net benefit. The evidence supports risks of capture, bureaucratic rigidity, suppressed voice, externalized harm, and fragility: land and water can be allocated through connected actors; central rules can lag field conditions; women and customary users can be unheard; downstream costs can leave the production ledger; and drought or maintenance failure can propagate through a shared network.⁷⁶⁸¹⁰

Concept emphasis separates the defining organizational questions

Delegation, decentralization, and responsibility, coordination, communication, and common understanding, structure, hierarchy, and scale, measurement, accounting, and control, knowledge, expertise, and professional autonomy, and governance, stewardship, and accountability score 3. Reform changed the agency boundary, local voice, hydraulic coordination, performance measures, and recognition of farmer knowledge without surrendering central land-and-water stewardship.²⁶

Purpose, mission, and institutional legitimacy, authority, legitimacy, and acceptance, decision-making, judgment, and bounded rationality, cooperation, incentives, and organizational equilibrium, work design, productivity, and automation, learning, quality, and reliability, and strategy, competition, and adaptation score 2. Food production supplies purpose; state title, permits, and fees supply authority; farmers and officials judge uncertain water and crops; and changing work, incentives, markets, and infrastructure explain adaptation without independently defining the case.⁴³⁵

Culture, informal organization, trust, and voice scores 1 because informal water decisions and cooperative ties matter, but the reviewed evidence does not support one culture across the full command area. Innovation, entrepreneurship, and renewal, executive attention, information, and organizational sensing, and organizational ignorance score 0 to keep the fingerprint discriminating. Local experiments and ignored claims appear in the record, but they are classified here under learning, governance, and voice rather than as separately measured mechanisms.⁶⁹

Comparative records test command, commons, and restoration

The Sudan Gezira Scheme compares another colonial irrigation system whose farmer autonomy changed before land, finance, and maintenance responsibility fully aligned. The Senegal River Basin Development Organization moves the water-allocation question to interstate infrastructure. Oman's aflaj institutions and Bali's subak governance compare longstanding local rules for shared water rather than one state estate.

Ethiopia's Gambella villagization program sharpens the distinction between planned development and meaningful consent. Farmer-managed natural regeneration in Niger shows public policy recognizing distributed farmer practice without owning one hydraulic network. These are editorial comparisons, not claims of direct influence, equivalent coercion, or universally transferable design.

Evidence remains thin for current water-fee burdens, eviction and appeal rates, lease performance, women and tenants without registered rights, seasonal labor, downstream fishers and pastoralists, chemical exposure, nonhuman welfare, and who receives value across the rice chain. A full water-and-land account would join agency records to farmer, village, basin, and ecosystem evidence before a new allocation becomes an accomplished fact.