Future2

Scenario Collapse

A devolved order without a capable centre

A world in which, for most people of the Global South cluster, the central government no longer holds effective authority: districts, lineages, faith courts, armed groups and company enclaves rule what they can reach, and no one can tax, protect and deliver across the whole.

A scenario is a possibility that the project has constructed, not a forecast. This page condenses the project's working draft of one scenario. Its sentences describe that world; they do not say what will happen.

Its class in the futures cone

The futures cone sorts futures as they look from today. A class is a label judged today, not a probability, and it can change as the evidence does.

  1. 2040plausible
  2. 2060plausible
  3. 2100plausible
  • plausible: a theory of how societies change leads to it, through mechanisms that have been seen before in other transitions, and at the pace at which they were seen

The idea

This scenario rests on an account of collapse. In it, societies solve problems by adding organisation, each addition returns less, and a society on the falling part of the curve spends its reserves on running costs. A shock it once would have survived breaks it, or its parts withdraw, until organisation falls to what local resources can sustain. That happens only where no rival takes over the land. Here it would happen in the Global South cluster, nearest to this order in the capacity of its states to tax, protect and deliver. Cost would keep the order low, local holders would deter one another, and outside powers would neither absorb nor rebuild the region. In most districts whoever rules would have to bargain with farmers, herders and fighters, while mineral enclaves rule without asking.

A day in 2100

Dawn in a market town of the dry interior. The well opens at the call to prayer, and the lineage council's man counts the jars: water is rationed by household and by standing. Most people farm small fields with hand tools, animals and a few solar pumps, or herd, or carry goods between markets. Young men without land serve a militia captain for grain and a name. Elsewhere in the world machines may do far more than people can; here they are little used, because nothing pays for the power, the parts and the network. In most districts a council of elders and a faith court, ruling by lineage and faith, hold security and speech, while the national government holds money and movement in the capital and the port. Families and traders hold the wealth, and water and pasture are held in common. Taxes to a capital that gave little back are no longer collected. But vaccines and surgery are rare, drought is met without relief, and feuds over water run for years. Strangers who fled other districts work for food without land. Neighbours keep the market truce and rebuild terraces after floods.

  • matches
  • apart
  • a row that defines this scenario
United StatesEuropeanUnionChinaIndiaGlobal SouthclusterPoliticalLocus of authorityBasis of legitimacyMode of decision and contestationCapacity of governanceEconomicProperty formCoordination mechanismMain source of surplusDistribution regimeSocialBasis of status and identityHousehold and community formBelief system and shared narrativeInclusion and exclusion rulesRows apart910966United StatesEuropean UnionChinaIndiaGlobal SouthclusterPoliticalLocus of authorityBasis of legitimacyMode of decision andcontestationCapacity of governanceEconomicProperty formCoordination mechanismMain source of surplusDistribution regimeSocialBasis of status andidentityHousehold and communityformBelief system and sharednarrativeInclusion and exclusionrulesRows apart910966

Where the five cases stand today against this scenario

The twelve rows of an order, case by case

A coding by the project Rows in which a case's state of 2026 is not one that this scenario accepts: United States 9, European Union 10, China 9, India 6, Global South cluster 6.

A row is apart when the case's state of 2026 is not one that this scenario accepts for that row. A case need not match every row to be in the scenario: it must match every row that defines it, and enough of the rest. The count says how many rows differ. It does not say how hard a change would be.

Source: the project's own coding of the five cases in 2026, set against its definition of this scenario. The coding rests on public data and on judgment. Some cells of the coding draw on the Integrated Values Surveys (2024; the European Values Study and the World Values Survey) and on UNU-WIDER's Government Revenue Dataset (2025), both through Our World in Data.

The values as a table
United StatesEuropean UnionChinaIndiaGlobal South cluster
Locus of authority (a row that defines this scenario)apartapartapartapartapart
Basis of legitimacyapartapartapartapartapart
Mode of decision and contestationapartapartapartapartapart
Capacity of governance (a row that defines this scenario)apartapartapartapartapart
Property formmatchesmatchesapartmatchesmatches
Coordination mechanismapartapartapartapartapart
Main source of surplusmatchesmatchesmatchesmatchesmatches
Distribution regimematchesapartmatchesmatchesmatches
Basis of status and identityapartapartapartmatchesmatches
Household and community formapartapartapartmatchesmatches
Belief system and shared narrativeapartapartmatchesmatchesmatches
Inclusion and exclusion rulesapartapartapartapartapart
Rows apart910966

How the world could get there

The project built this pathway backward from 2100 to today, for the Global South cluster. In the latest data 45.2 percent of its people live in countries coded as low in capacity, and for another 49.1 percent the index of government effectiveness falls in a middle band that does not decide. The path turns on an absence: no one would rebuild the centre. No step is judged infeasible.

  1. Until 2040, demands would outgrow what low-tax states can raise, while rescue lending props them up. The path needs the lenders to stop, for which no source was found, against a central work on collapse that expected rescue.
  2. In an exploratory count of the collapses of central authority that began between 1800 and 2018, 46 of the 48 that have ended reached a regular government; the count measures a lost hold on power, and also holds conquest in war. Here several large states would lose their hold before 2060 and not be rebuilt, perhaps as backing for rebuilders lapses and breakdowns come at once, for which no source was found.
  3. By 2075 authority would devolve to councils, captains, faith courts and licensed company enclaves. This works only if pumps, panels and phones can be kept up locally, for which no source was found.
  4. From 2075 to 2100, cost, deterrence and the absence of any rebuilder would have to keep the order. Its physical side was not judged, and that capable states neither absorb nor rescue the region for a generation rests on no source; no modern collapse of central authority that has ended lasted that long.

Who gains, who loses, what strains

In this scenario the winners would hold guns, water or a licence: militia captains, lineage heads, clerics who judge, enclave brokers and the companies behind them. The losers would be city people who lived on salaries and imports, the sick, the old national rich who stayed, strangers in the lower tier and women where custom hardens. The order would carry its own break, since enclave rents could fund the strongest captain, who taxes his neighbours and begins a new centre. It would lock itself in by neglect, as tax rolls, schools and roads decay and the knowledge to rebuild them thins, yet no lock would be complete. Ways out would stay open in 2100: restoration by an armed unifier, absorption by a capable neighbour, or an interregnum, in which the old order has decayed and no new one fits. A federation of councils toward the commons order would stay closed.

What this page does not show

What would break this scenario, the signposts that would show it coming, and its probability are not on this page. A first round of probabilities is in the project's log, and a review that tries to break each scenario and its probabilities is under way.

This page condenses a working draft. It may change as the work goes on.