What gold cannot buy

When Hajooj Kuka explained how he became involved in Sudan’s Emergency Response Rooms, he described a problem of finding a civilian role in a war. In a public discussion published by The New Humanitarian and Start Network on August 27, 2026, the filmmaker and ERR communications officer recalled trying to oppose the fighting, then turning toward mutual aid as a way to keep civic space open. He described communities deciding what to do with resources that began with their own time, skills, cooking equipment and shared spaces. His account gives the work a political purpose as well as a practical one.

That account changes the question I want to ask about Sudan’s economy. How much gold leaves the country is one question. How people retain the ability to make decisions about their lives is another. A family needs food and shelter, but it also needs some authority over whether to leave, where to return, how to earn a living and whose decisions it can challenge. An export receipt cannot establish any of those freedoms.

The distance between the two questions runs through the evidence. There are recorded gold exports, functioning markets and signs of economic recovery. There are also households whose return depends on whether they can pay for water or repair a home. Following those connections requires looking beyond the contest between armed leaders to the institutions, formal and improvised, through which civilian life becomes possible.

UNHCR’s Khartoum returns report, published October 8 and based on monitoring in July, describes decisions shaped by difficult conditions in displacement and perceived improvements at home. It records partial restoration of electricity and public services alongside damaged housing, limited work and unaffordable care. Some respondents wished to remain but said they might move again if essential conditions failed to improve. These findings came from discussions at twelve community centres and five non-random telephone interviews, with no systematic participant count for the centre discussions. They identify pressures, not their prevalence across Khartoum or Sudan.

The report’s water findings show how one broken service becomes another bill. Unreliable power interrupts pumping, and households then purchase water from vendors. Its livelihood findings connect lost productive assets and closed businesses to the difficulty of restarting work. For a household facing those conditions, a return journey can end while the work of returning continues. UNHCR, pages 4–7.

I would judge a recovery partly by whether people can remain where they have chosen to live without having to dismantle another part of their future to do so. Can a parent pay for repairs without withdrawing a child from school? Can someone take a job without losing access to treatment? Those are questions for further reporting, not outcomes that a border crossing or a national growth rate can answer.

Even the counting has to preserve that distinction. UNHCR’s returns monitoring separates people intending to stay in Sudan from those planning to go back to the country they left or onward to a third country. Intentions can change. An internal displacement snapshot, an arrival at a border and a return movement describe different things; the same person may move repeatedly. Adding them together would create an impressive total without a defensible meaning. The underlying regional portal also contains components with different reporting dates.

"Internal displacement: within-country stock; arrivals: cross-border movement; returns: movement toward former homes. None alone establishes a safe durable return."

Who gets to govern

Sudan became independent in 1956; South Sudan became a separate country in July 2011. The historical boundaries matter to today’s economic story, but neither date supplies an explanation of the present war. UNESCO’s historical overview records that sequence. The UN fact-finding mission’s 2024 historical analysis identifies the concentration of power and services around the centre, exclusion elsewhere and persistent impunity as underlying conditions. This is an institutional history of choices and unequal power, not a claim that geography made conflict inevitable.

Darfur belongs inside that history. The same mission traces the RSF to the government’s reorganization of some Janjaweed militia forces in 2013, followed by formal recognition in a 2017 law. Its account links those institutions to the government’s earlier counterinsurgency in Darfur. An ethnic category cannot tell us whether a person supports an armed organization, and a current uniform cannot erase an institution’s history. FFM, paragraphs 92–93.

Mass protests that began in December 2018 preceded the military’s removal of Omar al-Bashir on April 11, 2019. Civilian mobilization continued afterward. The demand for civilian authority did not disappear when the president did. FFM, paragraph 53.

The 2019 constitutional charter proposed a mixed civilian and military Sovereignty Council and a civilian prime minister selected by the Forces of Freedom and Change. It also reserved important roles for military nominees, including nominations for defence and interior ministers. The text makes the compromise visible: civilian institutions were to operate within an arrangement that retained armed power at the top. Its promised bodies and implementation schedule should be read as provisions, rather than proof that every institution was established or every transfer of authority occurred.

On October 25, 2021, the military seized power and dissolved the transitional government. The African Union’s Peace and Security Council condemned the takeover and suspended Sudan’s participation in AU activities the next day. The war between the Sudanese Armed Forces and Rapid Support Forces, with their allied forces, began on April 15, 2023. The parties' subsequent declaration itself dates the fighting to that day.

[
  [
    "2011",
    "South Sudan becomes independent"
  ],
  [
    "2018–19",
    "Uprising and removal of Bashir"
  ],
  [
    "2019",
    "Transitional charter adopted"
  ],
  [
    "2021",
    "Military takeover"
  ],
  [
    "2023",
    "War begins; Jeddah commitments"
  ],
  [
    "2025–26",
    "Court decisions and UN investigations"
  ]
]

Reading these documents together, I see repeated attempts to place public authority under rules that armed institutions could interrupt. That is an interpretation of the sequence, not a sufficient explanation of every decision to fight. It does explain why a settlement measured only by the distribution of offices among armed actors would leave a central civilian question unanswered: what prevents the next interruption?

Nor can two organizational names describe the whole country. The UN Panel’s account of Darfur in 2024 records divided armed movements and shifting alignments. Those descriptions belong to that period; they cannot serve as a current roster. Darfur, Kordofan, Blue Nile and the capital have different histories and experiences of violence. A map of control can help locate events, but its coloured areas cannot tell us who residents recognize as legitimate or how completely any authority administers daily life.

What recovery has to reach

The World Bank’s October 2026 estimates put Sudan’s real GDP contraction at 29.4 percent in 2023 and 14 percent in 2024, followed by growth of 3.1 percent in 2025. Applying those changes successively to a 2022 base of 100 leaves an index of about 62.6 in 2025. The rebound occurred after an enormous loss of output. It did not restore the earlier level.

[
  {
    "2022": "100"
  },
  {
    "2023": "70.600"
  },
  {
    "2024": "60.71600"
  },
  {
    "2025": "62.59819600"
  }
]

GDP measures production, not the distribution of purchasing power or the reliability of a water pump. The Bank describes localized recovery and partial restoration of activity and services, while emphasizing insecurity and institutional fragmentation. A national estimate can accommodate improvement in one place and continued destruction in another. It also cannot assign a value to every unpaid hour neighbours contribute to keeping one another alive. I would read the rebound as a reason to ask where recovery is taking hold and who can use it, rather than as an answer to those questions.

A market can remain open while a household loses access to what it sells. In its September 2026 bulletin, covering August observations, WFP reported that staples remained available in most monitored markets, while supplies were generally below normal. Assistance has to consider both sides of the transaction: whether food can arrive and whether people have the means to purchase it.

Farming brings another set of decisions into view. FAO’s September 24 assessment estimates the 2025 cereal crop at about 5.2 million tonnes, 22 percent below 2024. It separately warns that insecurity, input shortages and dry conditions threaten the 2026 crop, whose harvest was due to begin in November. A completed estimate for one crop and a warning about the next cannot be spliced into a measured decline. The warning does identify what needs to be watched: access to land, fuel, credit and inputs, as well as rainfall.

Land is therefore more than a space that changes colour on a conflict map. To sustain a livelihood, someone must be able to reach it, work it, obtain inputs and bring a product to a buyer. A photograph of cultivation can show a particular landscape; it cannot demonstrate that those conditions hold elsewhere or continue today. The same caution should govern a reconstruction plan that counts hectares while leaving the farmer’s practical access unresolved.

The hunger evidence is grave, and its boundaries are essential. The IPC special report published June 3 estimates nearly 19.5 million people in Crisis or worse acute food insecurity for February through May 2026. Its later projections cover only 56 of 195 localities and displacement settlements, compared with nationwide coverage for that earlier period. Smaller later totals consequently cannot establish national improvement. The report explicitly warns against that comparison.

The October 12 warning, summarized by WHO as an IPC member, identifies growing famine risk in parts of Greater Darfur and Greater Kordofan. A risk warning is not a new national famine classification. Neither caution makes the underlying danger less urgent. It helps keep the response aimed at the places and conditions the evidence actually identifies, while acknowledging areas where a lack of information prevents a fuller assessment.

Oil exposes a different dependence. Sudan and South Sudan have separate production and public finances, yet physical infrastructure links them. The US Energy Information Administration’s March 2024 analysis describes the two countries separately, including production, pipelines and refining. Its old capacity figures cannot establish what a refinery processes today. A facility’s ability to handle a stated volume under design conditions is a different fact from its operation during a war.

The UN Panel on South Sudan’s May 2026 report describes South Sudanese producers and the government selling their shares separately, with exports passing through Port Sudan except oil retained in Sudan. It also explains the role of Heglig facilities in Sudan in processing, blending and transiting Nile-blend crude from South Sudanese fields. Production location, the right to sell a share, use of infrastructure and receipt of a fee are separate parts of the transaction.

"Upper Nile to Dar system to Sudan transit to Port Sudan; Unity/Ruweng to Heglig blending to Sudan pipeline to Port Sudan. This is topology, not a verified present flow."

The Panel records Nile-blend production and export disruption in January and February 2026. It says the allocation of Sudanese infrastructure fees among the parties to Sudan’s conflict was unclear. The record supports a cross-border dependency; it does not supply an October operating log or a verified diagram of payments to combatants. A pipeline schematic should show the relationship while leaving present flow and ultimate recipients unresolved. Panel, paragraphs 122–123 and annex 32.

That distinction is worth carrying into any discussion of rebuilding. Repairing an asset creates an opportunity to move a commodity. The public benefit still depends on who can authorize its use, what obligations attach to the proceeds and whether those decisions can be inspected. Infrastructure can connect economies without providing a trustworthy account of the income it generates.

Gold makes the problem especially visible because the export numbers can look like prosperity. The Central Bank of Sudan’s first-half 2026 trade digest records 8,031.56 kilograms of gold exports valued at about US$1.135 billion. The comparable amended first-half 2025 figures are 8,834.58 kilograms and about US$839.9 million. Recorded quantity fell about 9.1 percent while nominal dollar value rose about 35.2 percent. The 2026 entries are preliminary.

[
  {
    "period": "2025-H1",
    "gold_kg": "8834.58",
    "gold_value_usd_thousand": "839899.75",
    "all_exports_usd_thousand": "1354913.78",
    "status": "amended",
    "source_id": "ECO-S06",
    "source_locator": "p.11 Table4",
    "gold_tonnes": "8.83458",
    "gold_value_usd_million": "839.89975",
    "all_exports_usd_million": "1354.91378"
  },
  {
    "period": "2026-H1",
    "gold_kg": "8031.56",
    "gold_value_usd_thousand": "1135194.72",
    "all_exports_usd_thousand": "1816066.38",
    "status": "preliminary",
    "source_id": "ECO-S06",
    "source_locator": "p.11 Table4",
    "gold_tonnes": "8.03156",
    "gold_value_usd_million": "1135.19472",
    "all_exports_usd_million": "1816.06638"
  }
]

This is a comparison of customs-based export records. It is not a measure of all gold mined, profit earned or revenue available for public spending. The digest also says some cash-crop and non-monetary-gold statistics are updated using UN Comtrade as information becomes available. Comparing its figures with another trade database may therefore involve overlapping inputs. The apparent second source has to be examined before treating it as an independent check. CBOS, introduction and Table 4.

Even a well-matched discrepancy between an exporter’s record and an importing partner’s record would need explanation. UN Comtrade’s methods guidance describes differences in partner attribution, valuation, timing, classification and trade systems. An unexplained gap is a reporting lead; it is not automatically a quantity of smuggled gold. Establishing a particular illicit transaction requires evidence that the aggregate comparison cannot provide by itself.

I would also resist allowing the commodity to swallow the people who work around it. A statistic about gold does not identify an individual miner’s allegiance or tell us who ultimately benefits from a sale. Scrutiny should follow ownership, transactions, intermediaries and the use of proceeds. Calling the whole trade a war-financing mechanism without following those steps would combine livelihoods, legal exports and alleged illicit activity into a single accusation.

There is, however, specific evidence of external support for armed operations. It needs to be read with its dates, methods and responses intact. The UN Panel’s January 2024 report assessed allegations of weapons deliveries through eastern Chad as credible. It recorded the UAE’s denial of involvement in arms shipments and its explanation that flights carried humanitarian assistance. The Panel’s assessment was investigative, not a court judgment.

The 2025 Panel report could not confirm military-materiel transfers from Am Djarass to Darfur in its inquiry into airport activities from May through October 2024. That bounded nonconfirmation belongs beside the earlier assessment. It does not establish that every alleged transfer occurred, or that no transfer occurred through any route in any period. A serious account has to preserve an investigation’s negative findings as carefully as its accusations.

In its September 2026 advance unedited report, the UN fact-finding mission found reasonable grounds to believe that a transnational private-actor network supplied the RSF with weapons, logistics, training and foreign personnel. The UAE told the mission it had remained neutral, supplied no arms and neither recruited nor facilitated foreign fighters; it also disputed links involving entities under its jurisdiction. The mission nevertheless found significant involvement by UAE-linked entities and individuals. It expressly made no finding of Colombian state responsibility. Its separate investigation into external suppliers and responsibility for SAF support remained preliminary, although it found foreign-supplied drones had expanded SAF capabilities.

[
  [
    "2024 UN panel",
    "Credibility assessment",
    "Alleged UAE-linked support via Chad;\nUAE said flights served a field hospital."
  ],
  [
    "2025 UN panel",
    "Bounded nonconfirmation",
    "Specific May–Oct 2024 airport inquiry\ndid not confirm military transfers."
  ],
  [
    "September 2026 FFM",
    "Investigative finding",
    "RSF networks; UAE denies support.\nAdvance unedited version."
  ],
  [
    "May 2025 ICJ",
    "Jurisdiction ruling",
    "Case dismissed without deciding\nthe merits of the allegations."
  ]
]

The United States added another kind of record on June 26, 2026, when Treasury designated eight people and entities it linked to SAF procurement and RSF recruitment networks. An administrative sanctions action is evidence of what the authority decided and alleged. It does not amount to a criminal conviction. Neither commercial contact, a company’s registration nor the manufacturing origin of a weapon independently proves state direction of a particular transfer.

Courts also answer different questions. On May 5, 2025, the International Court of Justice removed Sudan’s case against the UAE because it lacked jurisdiction, given the UAE’s reservation to the relevant treaty clause. The Court expressly took no position on the merits. The decision neither substantiated Sudan’s allegations nor cleared the UAE of the alleged conduct. It is possible to report that legal outcome accurately without using it to settle facts the Court did not adjudicate.

Accountability is also a demand made by survivors. In its March 2026 El Fasher report, the fact-finding mission recorded their repeated wish to know what had happened to missing relatives. The mission found war crimes and crimes against humanity by the RSF around the October 2025 takeover, and concluded that the pattern indicated genocidal intent against Zaghawa and Fur communities. It recorded protective instructions and promised investigations from RSF leadership, while saying it had received no clarification of resulting steps. These are investigative findings and reported responses, not a new criminal conviction.

The International Criminal Court provides a distinct historical example. Its Trial Chamber convicted Ali Muhammad Ali Abd-Al-Rahman on October 6, 2025, of 27 counts of war crimes and crimes against humanity for Darfur crimes in 2003–2004. The official case page records appeals of the judgment and sentence awaiting decisions. That individual trial finding cannot decide responsibility for every later atrocity, but it does show what an adjudicated case involves beyond a declaration or an allegation.

These records make me wary of treating accountability as an optional stage to be reached after the economic work is done. A survivor seeking a missing relative is asking an institution to establish a fact, disclose it and accept responsibility for acting on it. People deciding whether to return also need credible answers about protection and property. Those demands should help define recovery, rather than arrive as an appendix to it.

There is already civilian institutional work to learn from. The ERRs' public description of the Localization Coordination Council explains a process of community priorities, proposals, review, agreements, activity reports and financial reports. Its F-System is a method for deciding how limited resources should be used and documenting what follows. This is the network’s own description of its process, not an audit establishing uniform practice across all rooms.

[
  [
    "1  Community priorities",
    "Identify needs and propose an activity"
  ],
  [
    "2  Costed proposal",
    "Describe the work and its budget"
  ],
  [
    "3  Review and agreement",
    "Assess proposal and agree funding"
  ],
  [
    "4  Delivery and records",
    "Carry out work and document spending"
  ],
  [
    "5  Reporting and learning",
    "Report results; revise next priorities"
  ]
]

In a February 25 interview with Cordaid, ERR communicator Alsanosi Adam described joining as an affected community member and bringing his communication skills. He explained budgets reviewed against local needs, contracts, receipts and programme reporting. He also acknowledged a shortage of specialized medical expertise. Cordaid supports the network, so the interview should be read as attributable testimony from a partner relationship. Its account offers a process to examine, rather than a reason to assume every activity succeeds.

The acknowledgement of limits makes that process more credible to me. Knowing what neighbours need and how to organize around it is valuable. It does not create a supply of specialist clinicians, replace damaged infrastructure or protect volunteers from armed force. An outside partner should ask what it can supply that the community cannot obtain, while respecting who sets priorities. A funding agreement that rewards paperwork while consuming the time needed for delivery can defeat its own purpose.

That is a documented tension, not merely a hypothetical objection. A 2024 SSHAP study, based on twelve interviews conducted from June through August, describes decentralized decisions, local and diaspora support, threats, exhaustion and difficult reporting relationships. Its authors emphasize variation and warn that the accounts do not represent all volunteers. The study gives historical corroboration of the organizing work while resisting a single heroic story about every room.

Formal assistance still has a role. WHO reports supporting operational nutrition stabilization centres and treatment during January through August 2026. Such activity does not establish adequate coverage, but it is evidence against imagining that a volunteer network must replace every clinical and public institution. The practical task is to make support work together without making civilians responsible for overcoming every failure around them.

For an institution deciding how to help, I would begin with a few demanding questions. Who chose the priority? Can the people affected revise it? What happens when a payment is delayed, a supplier fails or a room cannot continue? What record can establish that resources reached the intended activity without exposing the people doing it? An answer should make responsibility clearer and action possible, rather than turn local knowledge into another unpaid reporting obligation.

The signs I would look for in the months ahead are equally concrete: people able to remain after returning; restored services they can afford; safer access to fields and markets; assistance that arrives when promised; and meaningful investigation of harm. None can be inferred from an export total. Each requires listening to people, following a transaction or service to its destination, and leaving room for an unwelcome answer.

Kuka’s account began with people considering what they could contribute and what role remained open to them. That capacity to decide deserves protection in its own right. A kitchen can feed a neighbourhood, a clinic can treat a child, and an accountable public institution should make it possible for both to keep working without depending indefinitely on extraordinary sacrifice. The people doing that work should have a say in the country being rebuilt around them.