There is a word that gets used to describe a large portion of the world's economies — developing. It is used in policy papers, in funding proposals, in the names of international institutions. It implies a direction of travel, a trajectory not yet reached, a condition that might one day be resolved. What it rarely names is the set of decisions that created the condition it describes. This article is not about development. It is about memory — specifically, about which memories were built into economic systems and which were not, and what the observable consequences of that choice look like today.
The term underdeveloped areas entered official global vocabulary on January 20, 1949, when U.S. President Harry Truman used it in his inaugural address to describe the regions his administration intended to assist.1 The framing was explicit: these areas had not yet reached the standard that the speaker's civilization had already achieved. A single linear scale was implied. Progress would be measured against it. What went unmeasured was who had constructed the scale, who had set the baseline, and — critically — who had decided what counted as economic activity worth recording.
Structural amnesia is not the absence of memory. It is the systematic exclusion of certain memories from the systems that allocate value, credit, and legitimacy.
That question — what gets recorded — is where the chain of evidence begins.
I. What the Ledger Was Built to Count
Colonial statistical infrastructure in Africa was not built to understand African economic life. It was built to manage the extraction of resources and the administration of labour. National income accounting as a framework did not arrive in most of sub-Saharan Africa until the late colonial and early post-independence period, and even then, it arrived in a form designed by and for Western European economies.2
The consequences of this are now documented in economic history. Morten Jerven's research across Ghana, Nigeria, Uganda, Kenya, Tanzania, Malawi, Zambia, and Botswana found that the statistical capacities of these economies were built on frameworks that treated large portions of actual economic activity as invisible — particularly the informal sector, where the majority of people in most of these economies have always operated.3 When Ghana updated its GDP calculations in 2010 by shifting the base year from 1993 to 2006 and introducing new definitions and basic data, the country's measured GDP roughly doubled — effectively repositioning Ghana from lower-income to middle-income status overnight.4 The economy had not doubled. The measurement had simply begun, for the first time, to partially see it.
The implications of this observation extend beyond statistics. If the measurement framework did not capture the economy, then every policy built on those measurements was built on an incomplete map. Every comparison between economies — which one was growing, which was stagnant, which was "developing" — was a comparison of ledger entries, not of economic realities. What the ledger could not see, it could not remember. And what it could not remember, policy could not value.
- Colonial statistical systems were designed to account for extraction flows — commodities, labour, tax revenue — not for the full range of economic activity in communities.
- National income accounting frameworks, designed for industrialized Western economies, were applied to African economies in the late colonial period without modification for the structural differences in those economies.
- As a result, informal economic activity — which constituted the majority of livelihoods — was systematically undercounted or uncounted in official national accounts.
- Policies, comparisons, and classifications built on those accounts therefore reflected a partial picture, framing incomplete measurement as evidence of limited economic activity.
II. The Structural Adjustment Interruption
The period immediately after independence represented, in many African countries, the highest investment in statistical capacity — the infrastructure for economic memory. This is not a contested point. Jerven's research notes that statistical capacity was significantly expanded in the late colonial and early post-colonial period.5 Then, that capacity was dismantled.
Beginning in the 1980s, IMF and World Bank structural adjustment programmes (SAPs) required governments across Africa to cut public services, reduce public-sector spending, privatize assets, and eliminate protectionist measures as conditions for receiving loans.6 Statistical offices — which had never been prioritized in international frameworks to begin with — were among the institutional casualties. As Jerven documents, the importance of statistical offices was specifically neglected in the decades of policy reform that followed the economic crises of the 1970s.7
The structural adjustment period is typically analyzed in terms of its effects on health systems, education, or agriculture. What receives less attention is its effect on institutional memory — on the state's capacity to know what was happening inside its own economy, to record it, and to retain that record across time. What SAPs introduced was not just austerity. They introduced a second layer of amnesia on top of the first: the post-colonial dismantling of the very systems that might have begun correcting the colonial measurement deficit.
The 1980s did not just impose austerity on public services. They imposed austerity on institutional memory itself.
The result was a compounding gap. By the time many African economies were classified as the poorest in the world — and treated accordingly in international economic policy — that classification rested on data produced by institutions that were under-resourced, working with frameworks that had never been designed for these economies, and that had recently been further weakened by the conditionalities of the very institutions doing the classifying.8
- Post-independence governments began building the statistical infrastructure needed to actually see their own economies.
- The commodity price collapse of the 1970s created fiscal pressure across the continent.
- SAPs introduced from the 1980s onward required cuts to public-sector institutions as a condition of lending — including statistical offices.
- The institutions responsible for economic memory were weakened at precisely the moment when correcting the measurement deficit mattered most.
- International classifications of these economies — used to set policy, allocate aid, and determine credit — were built on data produced by the weakened institutions.
III. The Digital Economy and the Third Encoding
The arrival of digital platforms in African markets was widely narrated as a correction to this history — a leapfrog moment in which communities that had been excluded from formal economic systems would now be brought into visibility. The evidence shows something more complicated.
Digitization has unquestionably created new trails of economic data where none existed before. Mobile money, QR payments, and digital logistics platforms are generating transaction records for millions of micro and small enterprises.9 In this narrow sense, the informal economy is becoming more legible. But legible to whom, and in what form, and owned by whom?
The data trails being generated by platform activity belong, in almost every case, to the platform. The market trader whose creditworthiness is now being assessed by a fintech lender is being assessed on the basis of behavioral signals extracted from her transactions — signals she did not choose to generate as records, did not consent to interpret as creditworthiness indicators, and cannot access, correct, or carry to another lender. The community whose supply chain is being traced for EU compliance purposes is generating provenance data that flows to the buyer, not back to the producer. The informal network of trust relationships — who pays, who delivers, who holds reputation across a market — is absorbed into platform logic and re-presented as a score.
A 2026 World Economic Forum analysis observed what is perhaps the clearest articulation of the gap: "A market trader knows which regulars pay on time because she has watched them settle small debts at the day's end for years."10 That knowledge is market memory. It is not informal. It is not unsophisticated. It is precise, longitudinal, and socially embedded. What the digital economy has not yet built is infrastructure for that memory to be held by the people who generated it.
Digitization has made the informal economy more legible. It has not made it more sovereign.
This is the third encoding of structural amnesia. The first was colonial — what the ledger chose to count. The second was structural adjustment — the dismantling of institutions that might have built better ledgers. The third is architectural — the design of digital systems that extract value from community knowledge without creating durable, transferable, community-owned memory of that knowledge.
- Digital platforms enter markets where formal economic memory is weak, providing infrastructure where none existed.
- Platform activity generates economic data — transaction histories, reputational signals, behavioral patterns — that constitute the first legible record of this economic activity.
- This data is held by the platform, not the participant. It is not portable, not correctable, and not inheritable across institutional boundaries.
- Community-level knowledge — trust networks, long-term payment behavior, contextual relationships — is absorbed into platform scoring systems, losing the provenance of who generated it and under what conditions.
- The "visibility" created by digitization therefore replicates the colonial pattern: economic activity becomes visible to external systems of classification, without that visibility returning to the communities that generated the evidence.
IV. Amnesia as Architecture
The word developing carries within it an assumption about time — that the condition it names is temporary, a phase on the way to somewhere else. What the evidence above describes is not a phase. It describes a structure: a series of design choices, made at different moments, by different actors, that produced a consistent outcome — the systematic exclusion of certain communities from the systems that hold economic memory.
Colonial measurement excluded informal economies from the ledger. Structural adjustment weakened the institutions that might have corrected that ledger. Digital platforms have created new ledgers — but kept them private.
Each of these is a design choice. None of them is inevitable. The colonial administrator chose what to count. The SAP architects chose which institutions to defund. The platform engineers chose where data sovereignty sits. These are not natural features of markets. They are features of the architecture within which markets operate.
Structural amnesia is therefore not a data gap to be filled by more data collection, better surveys, or improved satellite imagery. It is a design flaw in the systems that determine whose economic history is remembered, where that memory lives, and who controls access to it. Correcting it requires different architecture — infrastructure that originates memory at the community and producer level, that survives as provenance moves through institutional boundaries, and that remains accessible to the people who generated it.
The word developing is not a description of these communities. It is a description of the systems that have failed to see them. Those systems are the ones that need to develop.
This article is part of DALE's Market Memory Series — a body of work examining the conditions under which provenance and value survive (or do not survive) as economic knowledge moves through systems. DALE builds infrastructure for communities to originate, hold, and transfer market memory across institutional boundaries.