Hypothetical U.S. Mass Class Uprising Escalating to Sustained Armed Conflict

Full report

On this page

Executive summary

This report stress-tests a hypothetical, class-based mass uprising in the United States in which a broad “99%/Occupy”-style movement begins with demonstrations, strikes, occupations, and civil disobedience but develops an armed wing and eventually produces sustained lethal conflict with government forces. It is not an assessment that such an uprising is currently likely, and “French Revolution” is treated as an analogy for mass mobilization, institutional rupture, radicalization, factional struggle, and political transformation—not as a prediction that the events of 1789–1799 would literally repeat.

The central finding is that scale and state cohesion matter more than ideology. Under the baseline assumption that the federal government, most state governments, the armed forces, and the large majority of law-enforcement institutions remain organizationally cohesive, a violent revolutionary movement would have a poor prospect of overthrowing the U.S. constitutional order at the city or regional level. Even nationwide mobilization would not make regime change the modal outcome unless accompanied by major elite fractures, security-force defections, loss of administrative control, or a parallel crisis of state legitimacy. Historical revolutions become genuinely regime-threatening when mass mobilization is coupled with breakdown inside the state, not merely when demonstrations become very large.

The casualty estimates below are stress-test ranges rather than predictions. There is no empirically validated formula that converts revolutionary participation into deaths. The model therefore uses very broad intensity bands and separates armed-rebel deaths, government/security deaths, direct civilian deaths, indirect civilian mortality, wounds, detention, and displacement. Historical U.S. urban unrest demonstrates that even short local crises can produce dozens of deaths, thousands of injuries or arrests, and very large property losses; genuinely prolonged civil wars produce orders of magnitude more harm. The 1992 Los Angeles unrest, for example, killed 63 people and caused roughly $1 billion in contemporary property damage, while the Insurrection Act was invoked at California's request.

Under the modeled central-to-severe sustained-conflict envelopes:

These ranges are intentionally wide. They should not be read as “the number of people who would have to die” or as a success threshold. The highest-uncertainty quantities are civilian mortality, displacement, duration, and political outcome. The most robust qualitative prediction is simpler: once a mass movement becomes a sustained armed conflict, a zero-casualty or low-casualty revolutionary transition ceases to be a realistic baseline.

The U.S. already has considerable emergency-response capacity, but it is designed principally to surge around disasters and localized emergencies, not to absorb simultaneous mass-casualty events across many metropolitan areas indefinitely. HHS explicitly defines medical surge as sudden demand on beds, personnel, and other resources that can overextend health systems; the federal Hospital Preparedness Program and health-care coalitions exist partly to mitigate that problem.

Economically, the distinction between a violent week in one city and a genuine nationwide civil conflict is enormous. IMF research on conflicts finds persistent losses in output, investment, consumption, and trade, while recent World Bank analysis finds that high-intensity conflicts can leave GDP per capita roughly 20% below its previously projected path after five years. Those international results are not direct forecasts for the United States, whose institutions, capital markets, reserve-currency role, and economic structure differ substantially, but they provide an order-of-magnitude warning against treating prolonged internal warfare as simply a larger riot.

The report keeps weapons and logistics at a systems level—pre-existing small arms, generic illicit diversion, ordinary civilian transportation, food, fuel, communications, and medical supply—and does not map acquisition methods, vulnerable facilities, routes, or tactical targets.

Assumptions and scenario design

The figures that follow are generated from a scenario model, not from a claim that contemporary U.S. conditions are on a revolutionary trajectory. For scale, the model uses a rounded 345 million U.S. population baseline. That is deliberately rounded rather than pretending to forecast population to the person; Census Vintage 2025 is the most recent completed official estimates vintage.

The economic baseline is likewise contemporary but not treated as static. BEA's second estimate for the second quarter of 2026 reported annualized real-GDP growth of 1.5%, following 2.1% in the first quarter, illustrating that the hypothetical shock would strike a large functioning economy rather than an already collapsed wartime economy.

The principal assumptions are:

The three scales contain their own low, medium and high participation bands:

Using the rounded population assumption, the nationwide bands correspond to roughly 1.5%–8.7% of the total U.S. population participating at least once. The crucial distinction is between these cumulative participants and the much smaller number simultaneously engaged in violence.

The scenario should therefore not be visualized as “30 million people fighting soldiers.” A more realistic pattern is millions engaging politically, economically, or socially; hundreds of thousands participating repeatedly in confrontational actions; and a smaller armed minority generating a disproportionate share of direct lethal violence.

A plausible generic progression is:

timeline
    title Illustrative progression from mass movement to armed conflict
    Initial trigger : Economic or political shock
                    : Large demonstrations and strikes
    Days to weeks   : Occupations and civil disobedience expand
                    : Police confrontations and mass arrests
    Weeks to months : Lethal incidents create martyrs and retaliation narratives
                    : Armed splinters emerge
                    : Movement begins to factionalize
    Months          : Recurrent armed clashes
                    : National Guard and expanded federal response
                    : Supply and service disruptions intensify
    Months to years : Either negotiation and demobilization
                    : Or regional insurgency / national civil conflict
                    : Outcome increasingly depends on state cohesion

The uncertainty is asymmetric. It is fairly easy to say that sustained gunfire would produce deaths and injuries. It is much harder to predict whether 100 deaths trigger demobilization, radicalization, elite compromise, or a million-person expansion. Political feedback dominates mechanical casualty arithmetic.

Casualty model and scenario estimates

There is no scientifically defensible universal “revolution casualty rate.” Participation denominators are especially problematic because an occasional marcher, an armed militant, a paramedic, and a full-time fighter face radically different exposure.

The model therefore uses stress-test intensity bands per 1,000 cumulative movement participants, not a claim that history obeys these rates:

The two lines below are the lower and upper bounds of those analyst-defined bands:

xychart-beta
    title "Modeled direct-death burden per 1,000 movement participants"
    x-axis ["Low intensity", "Mixed", "Heavy"]
    y-axis "Direct deaths per 1,000" 0 --> 30
    line [0.05, 0.5, 5]
    line [0.5, 5, 30]

These rates are not thresholds for success. They are bookkeeping assumptions for harm analysis.

The reason to allow the upper bands to become large is that historical progression from “uprising” to “civil war” changes the source of mortality. Death no longer comes only from clashes at demonstrations. It comes from repeated shootings, fires, accidents, revenge violence, criminal opportunism, disrupted emergency medicine, interrupted utilities, shortages, and displacement. The ICRC's research on protracted urban conflict emphasizes precisely these secondary effects: damage to electricity, water and sanitation systems can propagate through dependent services and impose long-lived civilian consequences well beyond an initial strike or firefight.

Localized metropolitan case

A one-metro conflict is the scenario for which historical U.S. unrest offers the most useful—though still imperfect—comparisons. The 1967 Detroit upheaval produced 43 deaths, 1,189 injuries and more than 7,000 arrests in commonly cited Kerner Commission-era figures; the 1992 Los Angeles crisis killed 63, injured roughly 2,400 and resulted in more than 12,000 arrests. Neither was a sustained revolutionary war, which is why this report's severe city case extends substantially above them.

At a central sustained-conflict level, 75,000–200,000 cumulative participants with perhaps 3–8% becoming armed produces a modeled envelope of roughly 50–600 armed-movement deaths, 15–200 security deaths, and 50–700 direct civilian deaths. A severe high-participation branch can exceed those values; the table is intended as a central-to-severe envelope, not a mathematical maximum.

Indirect mortality remains comparatively contained if hospitals, power, water and transport continue operating. It rises sharply if neighborhoods become inaccessible to ambulances or if hospitals experience simultaneous trauma surges. HHS warns that medical surges strain beds, workforce, supplies and specialty-care capacity even in ordinary disasters.

Regional case

At several metropolitan areas across adjoining states, network effects become decisive. Mutual-aid policing, state Guard resources, hospital transfers, court capacity, insurance markets, freight movement and media narratives all become interconnected.

The modeled regional envelope—roughly 1–4 million cumulative participants and 5–12% armed during the sustained phase—produces direct deaths in the several-thousand to several-tens-of-thousands range, plus potentially thousands to tens of thousands of indirect civilian deaths.

The widening indirect range is intentional. A region can survive large demonstrations with surprisingly little indirect mortality if electricity, water, food distribution and hospitals remain functional; alternatively, months of repeated disruptions can make an otherwise survivable chronic illness or injury fatal.

Nationwide case

A true nationwide armed movement is qualitatively different rather than merely a larger version of a city riot.

At 10–25 million cumulative participants, with perhaps 7–15% participating in armed activity at some point during the severe phase, the central-to-severe envelope produces roughly 105,000–700,000 direct deaths across rebels, security personnel and unarmed civilians, with an additional 20,000–500,000 indirect civilian deaths in the modeled range.

Those figures have extremely low forecasting confidence. A nationwide conflict that collapses after six months could fall far below them. A multi-year conflict accompanied by state fragmentation, famine-like logistics failures, epidemics, systematic reprisals, or heavy military defections could exceed them substantially.

The most important causal sequence is therefore not “participation → casualties” but:

flowchart TD
    A[Mass class grievance or triggering shock] --> B[Demonstrations, strikes and occupations]
    B --> C{Movement discipline and state response}
    C -->|Violence remains exceptional| D[Negotiation, elections, reforms or peaceful stalemate]
    C -->|Repeated lethal confrontations| E[Armed splinters expand]
    E --> F[Retaliation, arrests and radicalization]
    F --> G{Government and security institutions remain cohesive?}
    G -->|Mostly yes| H[Containment, attrition, prosecutions and negotiated off-ramps]
    G -->|Major fractures| I[Competing authorities and territorial political rupture]
    I --> J{Settlement reached?}
    J -->|Yes| K[Constitutional or regime transition]
    J -->|No| L[Protracted civil conflict]

In other words, security-force cohesion is more politically consequential than raw protest size. A movement involving millions but facing cohesive institutions can still fail politically. A movement of smaller numerical size can become regime-threatening if the state itself fractures.

Escalation dynamics, recruitment and geography

A broad class movement would probably begin with a much wider social coalition than the armed faction that later emerges. Its recruitment narrative could unite people around wages, housing costs, debt, corporate or elite power, corruption, unemployment, perceived political exclusion, or a triggering episode of state violence. These are scenario assumptions, not a claim about a current revolutionary constituency.

The progression to violence would likely depend on several tipping mechanisms.

First is repression-radicalization feedback. A lethal confrontation can reduce participation by frightening moderates, but it can also create martyrs, delegitimize institutions among sympathizers, and strengthen factions arguing that peaceful tactics have failed. Which effect dominates cannot be inferred from the casualty count alone.

Second is elite and institutional defection. Refusal by elected officials to cooperate with one another, police or Guard units declining orders, resignations by senior officials, or visible military fragmentation would be far more significant than another hundred thousand demonstrators. Under the baseline assumptions here, such fragmentation remains limited; relaxing that assumption is the main pathway toward a French- or Russian-style regime rupture.

Third is social-media amplification. Research by Enikolopov, Makarin and Petrova found evidence that social-network penetration increased protest participation in Russia, supporting the broader proposition that social platforms can reduce coordination costs for collective action. In a violent U.S. crisis, the same infrastructure could simultaneously distribute legitimate information, facilitate mobilization, circulate unverified casualty reports and accelerate rumor-driven retaliation.

That would create a persistent information problem: videos of genuine abuses, old footage mislabeled as current, fabricated atrocities, mistaken identities, and selective editing could circulate faster than verification. The operational consequence would not necessarily be centralized coordination; a more plausible result is fragmentation into different realities, with local groups responding to incompatible accounts of what happened.

Geographically, the highest risk of mass mobilization would generally be in dense metropolitan areas and administrative centers because people can assemble in large numbers and political institutions are concentrated there. Rural areas would generally have less mass density but could experience more geographically dispersed confrontations, road closures, service isolation, and small armed-group activity. This does not imply a list of tactical “targets”; the model deliberately does not identify specific facilities, routes, or vulnerabilities.

Urban conflict is disproportionately dangerous to civilians because housing, hospitals, utilities, businesses, public transportation and conflict participants occupy the same physical environment. ICRC work on urban warfare emphasizes that damage to one service can cascade into water, sanitation, electricity and medical systems and can continue harming civilians after the immediate violence ends.

Weapons supply in this model remains deliberately generic. It assumes that armed splinters rely principally on small arms already circulating domestically, illicit diversion or theft, and whatever equipment individuals bring with them. It does not assume systematic access to armored vehicles, artillery, military aircraft or strategic weapons. A large-scale military defection transferring heavy weapons to a revolutionary faction would be a regime-changing variable outside the baseline model.

Nonweapon logistics could ultimately matter more than the firearm count. A mass movement of millions needs food, transportation, communications, medicine, shelter and money; government responders require many of the same systems. Consequently, prolonged conflict becomes a contest over whether ordinary social and commercial systems continue working, not merely an exchange of gunfire.

Factionalization would be extremely likely in the nationwide scenario. A slogan broad enough to encompass “the 99%” does not provide a common answer to questions such as property ownership, taxation, policing, constitutional structure, elections, federalism or post-revolutionary governance. A coalition could therefore split among nonviolent labor/civic organizations, electoral reformers, revolutionary socialists, anarchist or autonomist factions, local armed groups, ideological extremists, and ordinary criminal actors exploiting disorder. Historical revolutionary coalitions frequently become more divided after the old political consensus weakens; the French case is a particularly dramatic illustration of revolutionary factions turning coercion on one another during the Terror. Estimates for deaths associated with the French Terror run into the tens of thousands and remain historiographically contested.

This fragmentation creates one of the largest uncertainties in the entire analysis: “the revolution” would probably stop being a unitary actor well before a nationwide civil conflict reached maturity.

Government response, legal authorities and detention capacity

The most realistic government response is layered escalation, not immediate nationwide military occupation.

In an initial city-level crisis, local police, county agencies and state police would remain the first-line institutions, followed by mutual aid and gubernatorial activation of National Guard capabilities. Federal law-enforcement agencies would become more important as conduct crossed state lines, involved federal offenses, or affected federal responsibilities.

For the National Guard and active-duty military, several distinct legal authorities matter.

The Posse Comitatus Act, 18 U.S.C. §1385, generally prohibits willful use of covered federal armed forces to execute domestic law unless the Constitution or an Act of Congress expressly authorizes it. That means ordinary domestic disorder does not itself convert the active-duty military into a national police force.

The Insurrection Act supplies one of the principal statutory exceptions. Under 10 U.S.C. §252, the president may use the militia or armed forces when unlawful obstructions, combinations, assemblages or rebellion make enforcement of federal law through the ordinary course of judicial proceedings impracticable. Related provisions permit federal assistance after a state's request and in certain circumstances involving obstruction of federal law or constitutional rights.

There is also 10 U.S.C. §12406, under which the president may call National Guard units into federal service in specified circumstances including rebellion or when federal law cannot be executed with regular forces. It is legally distinct from the Insurrection Act even though their subject matter can overlap.

The important implication is that federal troops do not automatically appear simply because a protest becomes violent. A city riot can be handled under state authority; escalation toward direct federal military law enforcement becomes more plausible as ordinary civilian enforcement becomes unable to execute federal law or as the statutory rebellion/insurrection conditions become relevant. The Insurrection Act also does not simply erase constitutional protections or automatically create martial law; courts can continue to review the legality of government conduct.

A stylized response ladder is:

Rules governing military force in domestic operations would not simply be copied from overseas combat. Department of Defense issuances maintain standing rules for use of force and arming requirements, and domestic missions would operate under mission-specific legal orders, including self-defense and restrictions on force.

The expected pattern would therefore be graduated force where feasible, detention rather than lethal force against people who can be safely arrested, and deadly force associated with immediate lethal threats or applicable mission authorities, rather than a general license to fire on crowds. The actual rules would depend on legal status, mission and command orders.

Mass arrest is a separate constraint from battlefield control.

The Bureau of Justice Statistics reported that U.S. local jails held about 657,500 people at midyear 2024 and processed roughly 7.9 million admissions over the preceding year. Earlier detailed data showed approximately 915,800 jail beds in 2023 with about 73% occupied, although nominal unoccupied beds do not equal immediately usable emergency capacity because facilities, personnel and detainees are unevenly distributed.

That distinction matters enormously. A city can process thousands of arrests by drawing on surrounding counties and releasing lower-risk defendants. A regional event involving 100,000 additional simultaneous detainees would impose severe pressure. A nationwide scenario involving several hundred thousand to millions of arrests would create acute bottlenecks in booking, transport, pretrial hearings, defense counsel, prosecutors, judges and correctional staffing even though the normal U.S. correctional system already processes millions of admissions annually. The federal judiciary publishes extensive criminal caseload data, but its normal capacity is not designed for a synchronized revolutionary prosecution surge.

Thus “arrest everyone” would not be an administratively simple government strategy. Prosecutorial prioritization, pretrial release, plea bargaining, transfers between jurisdictions and court triage would become increasingly important as scale rose.

Criminal liability would also vary by conduct. Federal law separately criminalizes rebellion or insurrection under 18 U.S.C. §2383, but participation in a political demonstration does not by itself make a person guilty of that offense; individual conduct, intent, evidence and applicable jurisdiction would matter.

Humanitarian, infrastructure and economic consequences

The humanitarian consequences are best understood through interdependent lifelines rather than casualty counts alone.

FEMA's Community Lifelines framework groups functions such as safety and security, food/hydration/shelter, health and medical services, energy and communications because emergency recovery depends on maintaining several of them simultaneously.

A violent uprising would create a cascading sequence such as:

flowchart LR
    A[Violence and access restrictions] --> B[Transport delays]
    B --> C[Reduced food, fuel and medical deliveries]
    A --> D[EMS and hospital surge]
    D --> E[Delayed routine care]
    A --> F[Local power / communications interruptions]
    F --> G[Water, payments and communications degraded]
    G --> B
    E --> H[Indirect civilian mortality]
    C --> H
    G --> H
    H --> I[Displacement and public-health stress]

Healthcare

The immediate medical burden would include gunshot injuries, blunt trauma, burns, vehicle injuries, crowd-crush incidents and acute psychological trauma. The more serious long-duration problem is competition between emergency trauma care and normal medicine.

A hospital receiving dozens of major trauma patients at once may defer elective procedures. A network receiving sustained casualties for weeks starts consuming blood, operating-room time, ICU beds, anesthesiology, surgical staff, pharmaceuticals and ambulance capacity. HHS explicitly identifies beds, workforce and resources as medical-surge constraints, while GAO found that even COVID-era surges left some hospitals without enough staff, supplies or space.

The indirect death count then begins to include people who were never near a confrontation: dialysis patients missing treatment, heart-attack or stroke victims experiencing delayed transport, diabetics unable to obtain medication, high-risk pregnancies, people with infections, and nursing-home residents affected by utility or staffing failures.

Infrastructure and public services

The United States has substantial redundancy, but redundancy is uneven. The American Society of Civil Engineers' 2025 national report card graded U.S. infrastructure C overall, with weaker ratings in several categories including energy, transit, roads, wastewater and drinking-water systems. That does not mean those systems are about to fail; it means a conflict shock would land on infrastructure with varying levels of reserve and resilience.

The localized scenario would probably produce patchy outages and physical destruction rather than metropolitan collapse.

The regional scenario introduces mutually reinforcing failures: repair crews cannot reach damaged areas, fuel deliveries are delayed, employees do not report to work, communications congestion interferes with coordination, and hospitals or shelters become dependent on backup systems.

The nationwide severe branch is where indirect damage can overtake visible physical destruction. The ICRC's work on protracted urban conflict emphasizes that deterioration of water, electricity, sanitation and health systems can continue long after direct combat and can substantially increase public-health risks and displacement.

Food, fuel and displacement

The United States would not immediately “run out of food.” The more realistic danger is regional distribution failure: stores still exist and farms still produce, but transport interruptions, absenteeism, payment failures, curfews, insecurity and localized fuel shortages prevent normal replenishment.

At city scale this would more often mean days of shortages, closed stores and emergency distribution. At national scale, repeated disruptions could produce persistent regional scarcity and large internal migration.

The displacement estimates in this report—up to 150,000 locally, several million regionally, and potentially 2–20 million nationally—are particularly uncertain. They assume people predominantly flee from violent urban districts to safer areas elsewhere in the United States rather than becoming international refugees. No foreign military intervention is modeled.

Economic impact

Direct property destruction is only one component of economic cost. Lost work, closed businesses, impaired transportation, canceled investment, insurance losses, higher security expenditure, financing stress and capital flight can exceed the replacement cost of burned structures.

For a first-year deviation from the no-conflict GDP baseline, the model uses:

The lower and upper lines below visualize the modeled first-year output bands:

xychart-beta
    title "Modeled first-year real GDP deviation below no-conflict baseline"
    x-axis ["One metro", "Regional", "Nationwide"]
    y-axis "Percent below baseline" 0 --> 15
    line [0.05, 0.5, 4]
    line [0.4, 3, 15]

These are analytical ranges, not econometric forecasts. Cross-national conflict research is useful for calibration but cannot simply be pasted onto the United States. The IMF's 2026 analysis finds that conflict sites experience substantial and persistent output losses, while World Bank work finds around a 20% cumulative GDP-per-capita shortfall after five years in high-intensity conflict cases relative to earlier projections.

The sectoral pattern would probably be uneven:

Finance would react fastest. Equity and credit risk would reprice abruptly, bank funding and municipal credit in affected areas would deteriorate, and investors would seek liquidity. The response of U.S. Treasuries and the dollar is less obvious than in a foreign conflict because the source of risk would be the U.S. sovereign itself; traditional safe-haven behavior could therefore compete with domestic political-risk concerns. This is an inference from conflict economics rather than a firm forecast.

Transportation and logistics would face route closures, staffing problems, damaged assets, insurance costs and unpredictable delivery times. Even without systematic attacks on infrastructure, uncertainty alone reduces throughput.

Energy would face localized interruptions and difficulties accessing equipment for repair. The largest national risk would be cascading service failure rather than the physical destruction of the entire grid.

Retail, hospitality and commercial real estate would suffer rapidly in conflict zones, while healthcare, security, emergency logistics and repair industries would experience extreme demand and workforce strain.

Insurance losses would include property, business interruption, auto and liability claims, followed by extensive disputes over scope and coverage. Those losses are included only indirectly in the ranges above because policy language and legal classification of particular events would vary substantially.

Historical precedents, political implications and prioritized evidence

The most useful historical lesson is that “revolution” covers radically different phenomena. A week of urban violence, the collapse of an old monarchy, a military mutiny, a national civil war and a peaceful democratic transition cannot be compared simply by counting dead bodies.

The comparison substantially weakens the notion that a successful American uprising would have to “echo the French Revolution” in bloodshed. The historical variable most associated with revolutionary transformation is not a minimum casualty count. It is the loss of the incumbent state's ability or willingness to govern, which can happen peacefully, violently, or not at all.

For the hypothetical United States, the legal and ethical stakes would remain enormous even if the government ultimately prevailed.

Civil liberties would be placed under extraordinary pressure. Constitutional protections do not disappear merely because the Insurrection Act is invoked; legal analysis of the Act emphasizes that military conduct remains constrained by the Constitution and other federal law and remains subject to judicial review in important respects.

At the same time, sustained armed violence would create genuine government interests in public safety, investigation and prosecution. The hardest legal disputes would concern the boundary between protected political activity and criminal conduct, proportionality of force, searches and seizures, curfews, surveillance, detention, emergency orders and whether measures adopted temporarily become permanent.

The ethical risk is symmetrical in an important sense. An armed revolutionary faction can erode civilian protection by blurring itself into a much larger peaceful movement; government authorities can compound the damage if they treat an entire ideological or economic constituency as equivalent to the armed faction. Collective attribution would both violate liberal-democratic norms and potentially increase radicalization.

Long-term governance outcomes therefore span a wide spectrum:

Under strong government cohesion and relatively short conflict, the likely legacy is expanded security law, criminal prosecutions, political polarization, perhaps economic concessions or institutional reforms, and years of civil-liberties litigation.

Under a long regional conflict, a negotiated package of reforms becomes more plausible alongside stronger policing and surveillance institutions. The country may emerge formally constitutional but substantially changed politically.

Under nationwide conflict with major security-force fracture, constitutional replacement or regime transition becomes conceivable—but so do military government, competing governments, authoritarian consolidation, territorial fragmentation, or a negotiated settlement acceptable to neither original camp.

There is therefore no defensible assumption that a successful revolution would produce the social order its original “99%” coalition intended. The French and Russian cases are warnings precisely because revolutionary coalitions can become less unified after gaining power or weakening the old government, not more unified.

Prioritized evidence base

Overall confidence: high that sustained nationwide armed conflict would produce large direct and indirect human, economic, legal and institutional costs; medium that government escalation would broadly follow the local → state/Guard → federal law-enforcement → exceptional military-authority ladder; low-to-medium on duration and displacement; and low on any precise casualty count or final political outcome.

The decisive fork is not whether a hypothetical “99%” movement initially attracts millions. It is whether escalation destroys or preserves institutional cohesion. Under the baseline in this report—no foreign intervention, no heavy rebel arsenal, and no wholesale fracture of the armed forces or state governments—the city and regional revolts are most likely to end in suppression, negotiation, political concessions or some combination of the three. A genuinely nationwide French-style revolutionary rupture requires relaxing the baseline's most important assumption: the continued cohesion of the state. Doing so also moves the scenario into the range where hundreds of thousands of deaths, millions of injuries or detentions, millions of displaced civilians, severe recession or depression, and long-lived damage to constitutional government become plausible rather than remote tail risks.