In 2023, the International Labour Organisation released its Global Employment Trends for Youth report and recorded a global youth unemployment rate of 13.3 per cent. The figure was noted in development economics circles, cited in a handful of policy briefs, and largely absent from the front pages of newspapers covering the same year's wave of protests across nine countries. This is not incidental. The data on youth unemployment and the data on social unrest are rarely presented in the same room, by the same institution, in the same year. They should be.

The global average conceals what the country-level data reveals. In Tunisia, youth unemployment stood at 36.8 per cent. In South Africa, under the expanded definition that includes discouraged workers, the rate for those aged 15 to 34 exceeded 60 per cent. In Spain, one in four young people was unemployed. In Greece, the rate remained above 25 per cent a decade after the Eurozone crisis that produced it. These are not countries with comparable economies, political systems or cultural contexts. They share one variable: a generation that cannot find work and a recent history of mass protest.

The ILO's NEET measure, which captures young people not in employment, education or training, placed the global figure at 289 million in 2023. This is not a statistical abstraction. It is 289 million people between the ages of 15 and 24 who have no structured claim on the economy's daily output, no institutional anchor, and no financial stake in the continuation of the current order. The number is larger than the population of Brazil. It is published annually. It produces no emergency legislative session.

Time, Proximity, and the Absence of a Stake

The economics literature on protest participation consistently identifies three preconditions that predict who joins a demonstration and, when a demonstration becomes a riot, who remains in the street. The first is time. The second is proximity. The third is the absence of a stake in what might be destroyed.

Unemployment delivers all three simultaneously.

A person in full-time employment has constrained hours, a financial relationship with the formal economy, and some asset or income stream they stand to lose if social order breaks down. Their opportunity cost of participation is high. A young person who has been unemployed for eighteen months has unconstrained hours, no financial relationship with the formal economy worth protecting, and no asset likely to depreciate in a riot. Their opportunity cost approaches zero.

This is not a moral observation. It is an economic one. The decision to participate in a mass protest, or in its more disruptive form, a riot, follows from a rational calculation of expected costs and benefits. Research in this tradition, including Thiemo Fetzer's 2019 study in the Quarterly Journal of Economics on UK austerity and the Brexit vote, and a substantial body of IMF working papers on youth unemployment and social instability, has demonstrated that this calculation is sensitive to labour market conditions. When those conditions deteriorate, particularly for the young, the cost side of the calculation falls and participation rises. This is not a metaphor. It is a finding with a standard error.

The second factor, proximity, is amplified by urban youth unemployment. The protests and riots of 2025 and 2026 did not begin in rural areas. They began in city centres, peripheral housing estates, and urban zones of concentrated youth unemployment. These are places where unemployed young people live in close physical proximity to each other, to symbols of the state, and to commercial infrastructure that makes a protest economically disruptive. The organisational capacity for mass mobilisation is a function of density. Urban youth unemployment concentrates precisely the people most likely to mobilise in the places most structurally suited to it.

The Education Trap: When a Degree Guarantees Nothing

The relationship between education and unemployment runs in a counterintuitive direction across a significant number of the countries that experienced unrest in 2025 and 2026. Across much of North Africa and parts of Sub-Saharan Africa, the youth unemployment rate for university graduates exceeds the rate for those without formal qualifications. The formal sector has not expanded at a pace sufficient to absorb the graduates that secondary and tertiary education produces. The investment in human capital has been made. The return has not arrived.

This creates a particular social formation: an educated, urbanised, politically aware generation that has done what was asked of it. It attended school, passed examinations, completed a degree, and arrived at the formal economy to find the door closed. The grievance this produces is of a specific and concentrated type. It is not the diffuse frustration of generalised poverty. It is the precise, legible anger of a contract that has been broken by the party that wrote it.

Tunisia produced the original template for the Arab Spring of 2010 and 2011, and the dynamics are instructive. University graduate unemployment was running at rates that inverted the usual educational premium. Young Tunisians with degrees were, in measurable terms, more likely to be unemployed than peers who had left school earlier and found informal work. The revolution of 2011 was not led by the uneducated poor. It was led by this formation: educated, urban, unemployed, and acutely aware of what they had been promised and what they had received.

This matters for how governments misread the protests they face. The standard diagnosis of social unrest attributes it to failure, to dysfunction, to the pathology of communities that have not integrated into the economic mainstream. The data suggest the opposite. The people standing in the street are frequently those who tried hardest to integrate, completed the most education, and were most systematically refused the outcome that integration was supposed to guarantee.

The Global South Carries the Youth Unemployment Premium

The Arithmetic of Mobilisation

The IMF has estimated, across a series of working papers, that each percentage point increase in youth unemployment is associated with a statistically significant increase in the probability of social unrest in the following twelve months, controlling for other variables. This finding does not appear in the Fund's programme conditionality. The structural adjustment programmes that increase youth unemployment as a side effect of fiscal consolidation do not carry a riot-risk disclaimer. They should.

The arithmetic of mobilisation connects the three articles in this layer. A generation that cannot afford to eat, because food prices have risen faster than wages (Article 1). A generation whose wages, where they exist, have been falling in real terms for a decade (Article 2). And a generation, covered here, that cannot find work at all. Each of these conditions is sufficient on its own to produce grievance. Together, they produce a specific result: a large cohort of young people for whom the formal economy has, measurably, failed at every point of contact.

When the protest begins, these people are already outside. They did not need to leave a job to attend. They did not need to arrange childcare. They did not need to weigh the risk to a mortgage or a savings account. The riot was not a departure from their daily life. It was a continuation of it by different means.

Governments that respond to this with additional policing are solving the wrong problem. They are addressing the symptom that became visible on the streets while leaving intact the conditions that produced it. The data on youth unemployment has been available for years. The governments of the countries where protests broke out had access to it. The ILO published it. The World Bank published it. The African Development Bank published it. The people standing in the street already knew it. It was only the people with the power to change it who appear to have treated it as someone else's problem.