Intervention
Where the cycle can be broken
The index was always two products. The first is a risk score: how strongly the conditions that enable forced labor hold together in a country. The second is this: a map of where those conditions are most open to being changed. Forced labor is not a single event but a chain of structural conditions, and a chain can be broken at more than one link. This page reads the model for those links, the candidate places a government, a buyer, or an NGO could act.
Three families of lever, three different targets. Only the first two move the published score; the third makes the operation harder to pay for. Domains are equal-weight within a phase, so the side with fewer domains is where a single domain moves the score most.
Two kinds of place in the model
The index distinguishes two roles a structural condition can play. Drivers generate risk: they set how much latent forced-labor risk a country carries. Disruptors are the points where that risk could be made detectable or unprofitable; they do not create the risk, but acting on them can collapse the conditions that let it persist. Every lever below is one or the other, and the model is explicit about which.
Where risk is generated
The vulnerability that exposes people to recruitment, and the conditions that let exploitation run unchecked. Lowering a driver lowers the latent level of risk itself.
Where the cycle can be broken
The points (especially financial) that make exploitation detectable or unprofitable. Acting here does not remove vulnerability, but it can make a coercive operation unable to pay for itself.
What turns risk up or down
Conditions that amplify or attenuate a driver without being a driver themselves: rule of law is the clearest case, where strong institutions blunt a given level of vulnerability and weak ones leave it intact.
Why a second product was built instead of just a score
A risk score answers "where is it worst?" That is useful for triage, but it stops where a policymaker's real question begins: "and what could be done about it?" The model was designed from the start around two axes: the conditions that generate risk (Drivers) and the conditions whose removal could collapse the enterprise (Disruptors). Reporting only the score would throw away half of what the structure already encodes.
The two are kept separate and honest. The score is published and ranked. The intervention map is read off the same structure but framed as candidate levers, not as forecasts, because the index measures structure, and the effect of acting on structure is an empirical question it does not answer.
Where does it concentrate in one country?
Pick a country to see which side of the structure carries its score, and therefore which family of levers below is the most structurally relevant place to look first. This is read directly from the country’s real R and E components.
The levers, by phase
The conditions below are the model’s domains, re-read as places where the structure could change. Each carries its role in the model (a Driver generates risk, a Modulator turns it up or down, a Disruptor is a point where the cycle could be broken) and a plain note on who could act there.
Reduce who is made vulnerable
Recruitment risk is about who, structurally, can be pulled into forced labor, not who already has been. Lowering it means closing the gaps that leave people exposed.
Economic precarity
Poverty and insecure, informal income force people to accept dangerous work. Social protection, income floors, and formalization of informal work are the levers.
Who: government (social protection), NGO/multilateral (income support).
Debt & financialized dependency
Recruitment fees and migration debt convert a job into bondage. Banning worker-paid recruitment fees and regulating labor intermediaries are direct levers.
Who: government (fee bans), buyer (employer-pays sourcing).
Legal non-recognition
Statelessness and absent legal status strip people of recourse. Regularizing status and documenting stateless populations restores the protection that makes exploitation harder.
Who: government (status), NGO/multilateral (documentation).
Constrained mobility
Sponsorship regimes that tie legal status to one employer remove the ability to leave. Reforming or unbundling sponsorship from immigration status is the lever named in the model.
Who: government (immigration / labor law).
Structural disruption
Conflict, displacement, and disaster abruptly enlarge the vulnerable population. Anticipatory protection for displaced people is the lever; the shock itself is rarely preventable.
Who: NGO/multilateral (displacement response), government.
Age & childhood structuring
Child labor, out-of-school children, and child marriage put minors where coercion is hardest to see. Education access and enforcement of minimum-age and marriage law are the levers.
Who: government (education, enforcement), NGO/multilateral.
Ascriptive exclusion & gender structuring
Exclusion by caste, ethnicity, minority status, or gender channels groups into exploitable, hidden work. Anti-discrimination enforcement and targeted protection are the levers.
Who: government, NGO/multilateral.
Stop exploitation running unchecked
Exploitation risk is about whether forced labor, once it exists, can persist without consequence. The levers are the institutions and exit routes that would otherwise check it.
Economic structure & demand
Sectoral demand for cheap, coercible labor: the best-evidenced exploitation condition in the index. Supply-chain due-diligence and sectoral inspection are the levers buyers and regulators hold.
Who: buyer (due diligence), government (inspection).
Weak enforcement / rule of law
Governance enters once, as a protective modulator: strong rule of law blunts a given level of risk, weak rule of law leaves it intact. Building labor-inspection capacity is the lever, and the most consequential one in the model.
Who: government (inspectorate, courts), multilateral (capacity support).
Foreclosed exit
Monopsony — one employer or industry dominating local work, leaving workers nowhere else to sell their labor — and high exit costs make leaving impossible. Portable benefits, the right to change employer, and grievance mechanisms are the levers. Flagged: this condition is under-sourced in the index and carried as a low-confidence stand-in.
Who: government (labor law), buyer (grievance access).
State production of unfreedom
State-imposed and state-tolerated coercion. Here the actor responsible is often the state itself, which makes external leverage (trade conditionality, multilateral pressure) the realistic lever. Flagged: runs on a partial driver set.
Who: multilateral / trade partners, NGO (documentation).
Make it detectable and unprofitable
Forced labor is a money-driven crime: it persists as a durable enterprise only if the proceeds can be retained. These are the classic disruption points: where the cycle can be broken without first removing the underlying vulnerability.
Transnational concealment & laundering
Shell companies, beneficial-ownership opacity, secrecy jurisdictions, and professional gatekeepers let proceeds disappear. Beneficial-ownership transparency and follow-the-money enforcement are the levers.
Who: government / regulator (AML, ownership registries), multilateral.
Cash & informal-economy retention
Heavy cash reliance and large informal sectors let proceeds stay off the books, and absent formal credit makes engineered advances the operation’s working capital. Financial inclusion and informal-sector formalization are the levers.
Who: government (financial inclusion), multilateral.
The Monetization lens is mapped as a separate product and informs the intervention reading; it does not enter the published risk score. Why it sits apart is explained on the framework page.
Why governance appears as a lever but isn’t removed from the score
Weak governance is one of the most established structural drivers of forced labor, so an index that engineered it away would be measuring something other than what it claims. The relationship is disclosed instead of hidden: rule of law explains about two-thirds of the score, and the index says so plainly. Here it appears as a modulator (a protective lever) because that is the policy-relevant way to read it: building enforcement capacity is the single most consequential structural change in the model. The honest caveat is that because governance carries so much of the signal, a country’s position is partly a statement about its institutions, which the simulation lets you test directly.
And where inside a country
National levers still land somewhere specific. Because risk concentrates in particular regions a national average conceals, the most actionable reading is often subnational: the explore page surfaces the corridors (Thailand’s northeastern Isan belt, the Philippine Cordillera) where the conditions are densest. A lever aimed at the country as a whole can be aimed first at the regions the data flag.
Why the index points to regions, not just countries
A country score averages a whole population into one number, and that average erases the places where risk actually concentrates. For programming and enforcement, the regional surface is frequently the more useful view: it tells you where inside a country to look first. The reading is careful, though: the subnational layer exists for 97 countries with comparable household data, and within-country detail is real but modest for these labor signals. It isolates specific corridors rather than rewriting the national picture.
These are candidate intervention points read from a structural model: where risk concentrates and which conditions, in theory, could change it. The index does not forecast the effect of acting on any of them. It is a starting point for inquiry, read alongside on-the-ground knowledge, not a verdict on what will work.
Explore where risk concentrates · Test the structural assumptions · Open a country profile · How it’s built