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.

Recruitment who, structurally, can be pulled in 8 domains · equal weight within the phase × Exploitation whether it can run unchecked 3 domains · each ≈ 2.7× the marginal weight = The published score 0–1 · 184 of 195 countries scored Drivers — reduce what generates the risk (they act in both phases) Governance modulator blunts or amplifies the risk that is already there proceeds Disruptors — starve the proceeds; break the cycle without first removing the vulnerability Monetization whether the proceeds can be moved, hidden, kept 2 domains · tracked as a separate lens never enters the published score DRIVERS Recruitment who, structurally, can be pulled in 8 domains · equal weight within the phase × DRIVERS MODULATOR Exploitation whether it can run unchecked 3 domains · each ≈ 2.7× the marginal weight The published score √( R × E ) · 0–1 · 184 of 195 scored proceeds DISRUPTORS Monetization can the proceeds be moved, hidden, kept 2 domains · tracked as a separate lens never enters the published score

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.

Read this first. This page identifies candidate intervention points: where the index says structural risk concentrates and which conditions, in the model, modulate or defeat it. It does not predict the effect of any specific intervention. It cannot tell you that building a labor inspectorate will cut forced labor by a given amount. It tells you where the structure is sensitive, so that effort can be aimed rather than scattered. The causal step from "this is a lever" to "acting on this lever works here" belongs to evaluation on the ground, not to this index.

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.

Drivers

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.

Disruptors

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.

Modulators

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.

Government / regulator Buyer / business NGO / multilateral
Driver phase · Recruitment8 domains · each 1/8 of R

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.

Driver

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).

Driver

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).

Driver

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).

Driver

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).

Driver

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.

Driver

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.

Driver

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.

Driver phase · Exploitation3 domains · each 1/3 of E ≈ 2.7× the weight

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.

Driver

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).

Modulator

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).

Driver

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).

Driver

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).

Disruptor lens · Monetization2 domains · separate lens — not in the score

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.

Disruptor

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.

Disruptor

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