ILO Forced Labour: Intelligence Source Guide

The ILO is where forced labour is legally defined, statistically estimated and country-by-country supervised. Its indicators, conventions and expert committee observations are the reference framework that trade enforcement, supply-chain regulation and modern slavery litigation all build on.

At a glance

Source	ILO Forced Labour
Category	Conflict, Crime & Human Security › Human Trafficking & Child Protection
Format	`HTML`
Access	Open — no account required
Disciplines	Human Intelligence, Economic Intelligence
Mission domains	Forced Labour & Modern Slavery

Global forced-labour estimates & standards. — as catalogued in the platform’s own source registry.

The International Labour Organization is the United Nations specialised agency for the world of work, founded in 1919 and unique in the UN system for its tripartite structure: governments, employers' organisations and workers' organisations all hold formal seats. Its forced labour output has four distinct parts and confusing them is the commonest mistake analysts make with it. The first is normative: the Forced Labour Convention, 1930 (No. 29), its Protocol of 2014, the Abolition of Forced Labour Convention, 1957 (No. 105), and the associated recommendations, which together constitute the binding international definition of forced labour as work exacted under menace of penalty and not offered voluntarily. The second is supervisory: the Committee of Experts on the Application of Conventions and Recommendations publishes observations and direct requests on individual countries' compliance, drawing on government reports, trade union and employer submissions, and its own research, and the Conference Committee on the Application of Standards examines the most serious cases in public session. The third is statistical: the global estimates of modern slavery, produced jointly with Walk Free and IOM, and the ILOSTAT database of labour statistics. The fourth is operational guidance: the ILO indicators of forced labour, the fair recruitment principles, and a large body of sectoral and thematic research including the periodic analysis of the profits generated by forced labour.

Every other source in this area describes forced labour. This one defines it, and that difference is the whole point. When a customs authority detains a shipment, when a court decides whether a labour arrangement was coerced, when a company's auditor decides whether what they are looking at is bad practice or a crime, the operative test traces back to ILO instruments and the ILO indicators. That makes the ILO the prerequisite source for any SUPPLYINT or ECONINT work on forced labour: it supplies the classification scheme that everything else is expressed in. The second irreplaceable product is the supervisory record. The Committee of Experts publishes detailed, country-specific, evidence-based commentary on named practices in named sectors, sustained across decades, produced by an independent expert body with the formal right to ask governments questions and record their answers. That is a genuine intelligence product in the ordinary sense of the term, hiding in plain sight in a legal database, and it is the closest thing that exists to an authoritative open assessment of state-imposed and state-tolerated forced labour. Analysts who only take the headline global estimate are leaving the best material on the table.

Who publishes it, and why that matters

The ILO is an intergovernmental body whose members are states, and its tripartite structure means employers' and workers' organisations shape its output as well. This produces a characteristic pattern you must read for. The organisation is institutionally cautious in its language, particularly about member states, and diplomatic constraint shapes what appears in official documents. Where a government is implicated in forced labour, the Committee of Experts will typically express deep concern, urge, and note with regret rather than accuse – and that register, once you learn to read it, is highly informative, because the escalation from a direct request to an observation to a special paragraph at the Conference to an Article 26 commission of inquiry is a public, ordinal measure of how serious the situation is judged to be. The Committee of Experts itself is composed of independent jurists rather than government representatives, which is why its findings carry weight that a purely intergovernmental output would not. Funding is assessed contributions plus donor-funded technical cooperation, which gives the standards machinery stability while making the research and field programmes donor-shaped. This is a source with a century of continuity behind it and it is not going anywhere.

Provenance is the first question to ask of any dataset and the one most often skipped. Who collects it, what their incentive is, whether they publish a methodology, and whether they correct the record when they get something wrong all bear directly on how much weight a finding drawn from it can carry.

What a record actually contains

The fields you will be working with, what each one means, and whether it is something you can pivot on. Read the meanings carefully — more analysis is wrecked by misreading a field than by failing to find one, and a field that looks like an observation is often an inference.

Field	Type	What it means	Pivot value
`Convention or instrument`	string	The specific legal text: Convention No. 29, its 2014 Protocol, Convention No. 105, Convention No. 182 on the worst forms of child labour, and the associated recommendations. The instrument determines what obligation is actually in play, and they are not interchangeable.	Ratification status by country, the text itself, and the national implementing legislation the country claims discharges it.
`Ratification status and date`	timestamp	Whether and when a country ratified an instrument, and whether it has denounced any. Ratification triggers reporting obligations and the supervisory cycle, so a non-ratifying state generates far less scrutiny regardless of its actual practice.	Reporting obligations, supervisory comments, and the asymmetry between countries under scrutiny and those not.
`Committee of Experts observation`	string	The published comment on a country's application of a convention. Contains the specific practices at issue, the government's response, the submissions of workers' and employers' organisations, and the Committee's assessment. This is the substantive intelligence.	Named sectors, regions and practices; the trade union or employer body that raised it; the government's own admissions.
`Direct request`	string	The lower-visibility form of supervisory comment, sent to the government rather than published in the Committee's report body but available in the database. Often contains detail that never reaches an observation.	Escalation trajectory – repeated direct requests preceding an observation are a leading indicator of a deteriorating assessment.
`Article 22 government report`	string	The state's own periodic report on how it applies a ratified convention. States describe their own law and practice, which makes these documents useful as admissions and as a record of what a government claims, whether or not it is true.	National legislation cited by the government; the gap between claimed and observed practice; contradictions with worker organisation submissions.
`ILO indicators of forced labour`	array	The eleven operational indicators: abuse of vulnerability, deception, restriction of movement, isolation, physical and sexual violence, intimidation and threats, retention of identity documents, withholding of wages, debt bondage, abusive working and living conditions, and excessive overtime. The working classification scheme for the entire field.	Screening tools, audit protocols, case data such as the CTDC control fields, and the evidentiary standards used by trade enforcement authorities.
`Global estimate figures`	int	The joint ILO, Walk Free and IOM estimates of people in forced labour and forced marriage, reported globally and by region, sector and demographic. Regional and sectoral breakdowns are more informative than the headline and less often quoted.	Regional prioritisation; sectoral risk registers; comparison against case-level and national identification data.
`Sector`	enum	The economic sectors identified in estimates and research: domestic work, construction, manufacturing, agriculture and fishing, mining and quarrying, and commercial sexual exploitation, among others. Sector is the dimension that most directly maps onto supply-chain work.	Harmonized System codes, industrial classifications, supplier categories and the sectoral audit protocols already in use.
`State-imposed forced labour category`	enum	The estimates distinguish forced labour imposed by state authorities from that imposed by private actors. This distinction is politically loaded and analytically essential, because the remedy for one is prosecution and for the other is diplomacy, sanctions or trade measures.	Convention No. 105 obligations, which specifically address state-imposed forced labour, and the supervisory record for the country concerned.
`Supervisory escalation level`	enum	Where a country sits in the machinery: routine reporting, direct request, published observation, examination by the Conference Committee, a special paragraph, or an Article 26 commission of inquiry. An ordinal severity scale produced by an independent body over time.	Trend analysis of a country's compliance posture; the strongest available open indicator of assessed severity for state-linked forced labour.
`Recruitment fee and cost definitions`	string	The ILO's definition of recruitment fees and related costs, which underpins the employer-pays principle now embedded in many corporate standards. A precise technical definition with direct compliance consequences.	Recruitment agency due diligence; migrant worker corridor analysis; the specific fee structures that create debt bondage.
`Profits estimate`	int	The ILO's periodic assessment of the illegal profits generated by forced labour, broken down by region and by form of exploitation. Published in the Profits and Poverty series, most recently updated in 2024 with a figure in the hundreds of billions of dollars annually.	Financial crime and anti-money-laundering work; the economic incentive structure that any intervention has to overcome.
`ILOSTAT indicator`	string	The labour statistics database covering employment, informality, wages, working time, labour inspection capacity and migrant workers. Not forced-labour-specific, but the informality and inspection-capacity indicators are the best available proxies for where forced labour can persist undetected.	National labour inspectorate capacity; informal employment share; migrant worker stocks by corridor.

Coverage — and what is not in it

Coverage differs sharply by product. The normative and supervisory material is close to universal in scope but asymmetric in depth: every member state is covered in principle, but scrutiny concentrates on states that have ratified the relevant conventions, and within those on states where workers' or employers' organisations have made submissions, since a submission is the usual trigger for detailed examination. This means a country with weak or suppressed trade unions generates less supervisory comment than one with strong ones, independent of actual conditions. The supervisory record has genuine historical depth – comments on some countries run back decades, which makes it one of the few sources supporting real long-run compliance analysis. The global estimates are worldwide with regional and sectoral breakdowns, published every few years rather than annually. ILOSTAT is broad but variable, with rich series for countries that have functioning statistical systems and gaps elsewhere. Update rhythm is institutional: supervisory comments follow the reporting cycle, with the Committee of Experts meeting annually and the Conference Committee examining a selected list of cases each June. Nothing here is real time and nothing needs to be.

Known blind spots

Absence of evidence here is not evidence of absence. These are the conditions under which ILO Forced Labour will not show you something that is nevertheless real:

Non-ratifying states generate far less supervisory material. The machinery runs on ratification, so a state that has not ratified a convention faces only limited follow-up under the fundamental principles framework, and the record will be thin regardless of conditions on the ground.
The supervisory system depends on submissions from workers' and employers' organisations to surface problems. Where independent trade unions are banned, co-opted or dangerous to join, the mechanism that would raise the issue does not function, and the country appears quieter than it is.
Diplomatic language obscures severity for readers who do not know the register. A finding expressed as noting with deep concern may describe systematic state-organised coercion, and an analyst reading for plain accusation will miss it entirely.
The global estimates cannot see inside closed states, prisons, military conscription systems and other institutional settings, because the survey methods underlying them are household-based and the alternative sources are fragmentary.
Nothing in the ILO's public output identifies companies. Sectors, regions and practices are named; supply chains and firms are not, which means this source can tell you that a sector in a country is high risk and never that a particular supplier is implicated.
The organisation moves at institutional speed. A rapidly emerging forced labour situation may take years to work through reporting cycles into a published observation, so the supervisory record lags events substantially.
The tripartite structure means the output is a negotiated product in part. Employer and worker constituents contest framing, and some subjects – notably prison labour, conscription and certain state programmes – carry contested interpretations that the documents reflect rather than resolve.
Statistical coverage in ILOSTAT depends on national statistical capacity, so the informality and inspection indicators that would tell you most about where forced labour can hide are weakest precisely where informality is highest.
Forced labour occurring in the digital and platform economy, in scam compounds, and in other rapidly emerging forms is under-represented in a framework built around recognisable employment relationships, though the organisation has begun to address it.

Write the blind spot into the product. A statement that something “was not observed in ILO Forced Labour” is defensible; a statement that it “did not happen” is not, and the difference is what survives cross-examination.

Access, licensing and what you may do with it

Access model: Open — no account required

There is no single API and no unified download. Access splits across several systems and you should learn all of them. The main website carries the topic pages, reports and the global estimates. NORMLEX is the database of international labour standards, ratifications, national legislation and – critically – the full text of Committee of Experts observations and direct requests, searchable by country and by convention. This is the system that matters most for analytical work and it is the one most analysts have never opened. ILOSTAT carries labour statistics with bulk download and an interface for indicator queries. Reports and technical guidance are published as PDFs on the main site. For serious use, plan on scripted harvesting of NORMLEX comments for the countries and conventions in your scope, on a cadence matched to the annual supervisory cycle rather than daily, plus scheduled retrieval of ILOSTAT indicators. The ILO's web estate has been restructured more than once and deep URLs have broken as a result, so anchor your collection on the database roots rather than on saved deep links, and re-verify paths after any site change.

Licence

ILO publications are copyright of the International Labour Organization, and the organisation has moved substantial parts of its output to open licensing with attribution, though terms are not uniform across the whole estate and older material may carry different conditions. Check the notice on the specific document. As UN system material, reproduction for non-commercial research, educational and advocacy purposes is generally permitted with attribution, and commercial reuse or translation typically requires permission. Convention and recommendation texts are legal instruments and are freely reproducible as a practical matter. ILOSTAT data is published for open use with attribution. The practical rule: cite the specific instrument, comment or publication with its year, attribute to the ILO, and seek permission before building a commercial product on bulk-extracted content. Where the ILO reproduces national legislation or third-party submissions, those carry their own rights.

Rate limits and fair use

No published limits, but treat these as institutional research systems rather than data services. NORMLEX in particular is a legal database designed for human browsing, and aggressive crawling will be noticed and is unnecessary – the supervisory record changes on an annual cycle, so a scheduled harvest after the Committee of Experts report is published, plus a periodic check for new ratifications, covers essentially everything. ILOSTAT supports bulk downloads, which you should use rather than iterating queries. Identify your client, cache locally, and back off on errors. If you need sustained programmatic access at scale, contact the organisation; the ILO has a research-cooperation culture and responds better to a request than to a scraper.

Licensing changes, and it changes without warning. A dataset that was free for research this year may not be free for commercial or evidential use next year. Confirm the current terms before you build a dependency on it, and record the terms you relied on alongside the data — the licence in force at the time of collection is part of the provenance.

Collecting it

How ILO Forced Labour is actually pulled, in the order you would set it up. Prefer the bulk or export interface over per-item lookups wherever one exists: it is kinder to the publisher, faster for you, and gives a reproducible snapshot rather than a series of point-in-time answers you cannot reconstruct later.

Method	Format	Cadence	Notes
NORMLEX supervisory comment harvest	HTML	Annual, after the Committee of Experts report	The highest-value collection route. Pull observations and direct requests for the conventions and countries in scope, store full text, and diff against the prior year to detect new or escalating findings.
Ratification status pull	HTML	Quarterly	Track new ratifications and denunciations of Conventions 29, 105, 182 and the 2014 Protocol. A ratification changes a country's obligations and opens it to a supervisory cycle, which is a leading indicator of future material.
Global estimates and thematic reports	HTML	Per publication	The joint estimates, the profits analyses, sectoral studies and guidance documents. Collect the methodology sections, not only the figures, since comparability across editions depends on them.
ILOSTAT bulk indicators	CSV	Quarterly	Informality, labour inspection capacity, migrant worker stocks and working time. Use the bulk download rather than per-query retrieval, and pin the indicator codes since they are stable and the labels are not.
Conference Committee case list	HTML	Annual, June	The list of individual cases examined at the International Labour Conference and their conclusions, including any special paragraph. The clearest annual signal of which countries the supervisory system judges most serious.
Indicator framework and guidance capture	HTML	On revision	The eleven indicators, the fair recruitment principles and the recruitment fee definitions. These are reference data for your own classification scheme rather than observations, and they change rarely but consequentially.

Ingesting it into the platform

Every step below is idempotent and cursor-based: interrupt one and it resumes from where it stopped rather than duplicating rows or losing progress. Collection is recorded per source, so a feed that quietly stops publishing shows up as a stale timestamp instead of silently thinning your coverage.

Register the constituent systems separately — Add NORMLEX, ILOSTAT and the publications estate as distinct collection routes under one source in sources.php, because they have different cadences, formats and reliability and merging them into one feed status hides which part has gone stale.
Load the indicator framework as reference data — Import the eleven forced labour indicators as a controlled vocabulary through import.php, so that case data, audit findings and other sources can be mapped onto a single classification scheme rather than each carrying its own.
Parse supervisory comments into structured records — Extract country, convention, comment type, year and full text from NORMLEX, and store the escalation level as an explicit field. The text is the evidence; the structure is what makes it queryable alongside everything else.
Diff against the previous cycle — On each annual harvest, compare new comments against the prior year for the same country and convention. New observations, escalations from direct request to observation, and newly named sectors are the events worth alerting on through alerts.php.
Map sectors to commodity codes — Resolve named sectors to Harmonized System and industrial classification codes so that a supervisory finding about a sector in a country becomes joinable to trade data and to a client's actual supply chain.
Push country context into the risk layer — Feed ratification status, supervisory escalation level and ILOSTAT informality and inspection indicators into country.php and country-risk.php as documented inputs, keeping them distinct from modelled prevalence estimates from other sources.
Cross-link to case and enforcement data — Use correlate.php to sit ILO supervisory findings beside case-level victim data, trade enforcement actions and national prosecution records for the same country and sector, which is where the legal, statistical and operational pictures either converge or reveal a gap.
Expose through the human-rights and supply-chain surfaces — Publish the resulting country-sector risk view through human-rights.php and the platform's labour and supply-chain views, with the underlying comment text available so an analyst can read the source rather than only the derived flag.

Registered sources and their last-collected state are listed in sources.php, and the scheduled chain that keeps them current is in automation.php.

How it is wrong, and how to tell

Every dataset is wrong in characteristic ways. Knowing which ways is the difference between using a source and being used by one, and it is the part of source evaluation most often skipped because it is the part that takes work.

This is the highest-quality source in this batch and one of the highest-quality sources in the library, provided you take the right products. The conventions are binding law with authoritative interpretation. The indicators are the field's standard classification scheme and have survived two decades of operational use. The Committee of Experts is an independent body of jurists whose findings are evidence-based, sourced to identifiable submissions and government responses, and produced under a process that gives states a formal right of reply, which makes them unusually defensible. The global estimates are weaker – they inherit the methodological problems of any global prevalence exercise, including extrapolation and definitional breadth, and they are co-produced with a partner whose own index is contested – but they are more transparent about method than most alternatives and carry the ILO's statistical discipline. The limitation is not accuracy but reach: the system sees what states report, what workers' organisations can safely raise, and what surveys can capture, and each of those filters is tightest exactly where the abuse is worst. Trust the legal and supervisory material heavily. Trust the statistical material as carefully framed estimates. Do not expect either to be current to the month.

Characteristic false positives

Silence read as compliance. A country with no recent observations may be compliant, may not have ratified, or may have no independent workers' organisation able to make a submission. The supervisory record is triggered by inputs, and absent inputs produce an absent record.
Diplomatic register misread as mildness. Formulations such as noting with deep concern and urging the government to take immediate action are near the top of the escalation scale, and an analyst expecting explicit condemnation will systematically under-rate the severity of published findings.
Global estimate figures quoted without their definitional scope. The headline combines forced labour and forced marriage, and combines state-imposed and privately imposed exploitation, so a figure used to characterise supply-chain risk includes large populations that have nothing to do with supply chains.
Sector-level findings applied to individual firms. The ILO names sectors and practices, never companies, and converting a sectoral finding into an allegation about a supplier is both unsupported and legally risky.
Cross-edition comparison of estimates without checking methodology. The estimates series has changed method and scope between rounds, so apparent global trends may be revisions rather than movements.
Ratification treated as implementation. Many states ratify and then fail to implement, and some of the most serious cases in the supervisory record involve long-ratified conventions. Ratification status tells you what obligations exist, not what happens.
ILOSTAT indicator instability. Indicator definitions, coverage and national methodologies vary and change, and series that look continuous sometimes are not. Pin indicator codes, read the metadata, and check for breaks before drawing a trend.
Timeliness assumed. Supervisory comments reflect reporting cycles that lag events by years, so a country's record can look stable through the period in which the situation actually deteriorated, with the deterioration appearing in the record much later.

None of these make the source unusable. They make it a source that requires corroboration before an assertion built on it goes into a product, which is true of every source and admitted by few.

Ageing

Different components age at completely different rates and this is one of the few sources where that matters operationally. Convention texts and the indicator framework barely age at all – they are the stable reference layer and revisions are rare and well signposted. Ratification status changes occasionally and is easy to keep current. Supervisory comments age on an annual cycle: a comment is current until the next examination of that country and convention, which may be a year or several years away depending on the reporting schedule, and the correct reading of an old comment is that it describes the last time the Committee looked rather than the situation now. Global estimates age across multi-year publication cycles and should be cited with their round. ILOSTAT indicators age by series, with some updated annually and some much less often. A stale use of this source looks like a country risk assessment citing a supervisory observation from several cycles ago as if it were a current finding, or a compliance document quoting a superseded global estimate. The practical rule: always cite the year of the comment or edition, and check whether a more recent examination has occurred before relying on any country-level statement.

What this source feeds

A source is only worth what it lets you conclude. These are the disciplines that collect through it, the mission domains it serves and the data points it yields — every one is a tag, so you can follow any thread from here into the rest of the library.

Collected by these intelligence disciplines

Serves these mission domains

Forced Labour & Modern Slavery

Yields these data points

Location / Coordinates

How each sector uses ILO Forced Labour

The same dataset is worked very differently depending on who you are, what authority you hold, and what you are ultimately producing. A military analyst is supporting a commander’s decision; a journalist is meeting a publication standard; an NGO caseworker is protecting a person. The records are shared — the constraints, thresholds and outputs are not.

🎖 Military and defence

Two concrete applications. First, contracted labour: deployed operations buy construction, catering, cleaning, security and logistics services in host nations, and the ILO indicators are the recognised screening framework for whether that labour is coerced – retained passports, recruitment debt, wage withholding and controlled accommodation are the indicators to write into contract terms and inspection regimes. Second, conflict analysis: forced labour is a feature of many war economies and of state control systems, and the supervisory record plus the state-imposed category in the estimates give an open, citable basis for assessing it in a theatre. Use Convention No. 105 obligations as the reference point when the concern is state-organised coercion rather than private exploitation.

🕵 National intelligence

The Committee of Experts record is an underused open intelligence product on state behaviour. It provides sourced, expert, longitudinal assessments of named practices in named sectors, produced by an independent body with a formal right to question governments and record their answers, and the escalation ladder gives you an ordinal severity measure that is public and defensible. For SUPPLYINT and ECONINT work it supplies the classification framework that lets forced labour findings be combined across sources. Read the register, track the escalation trajectory rather than individual comments, and pay particular attention to cases where a government's own Article 22 report contains admissions that contradict its public position.

👮 Law enforcement

The value is definitional and evidentiary. The eleven indicators are the framework investigators and prosecutors use to distinguish exploitative employment from forced labour, and building them into interview protocols and case theories improves both identification and charging decisions. Convention No. 29's menace of penalty and voluntariness test is the conceptual basis of most national forced labour offences, so the ILO's interpretive material helps in arguing the elements. The supervisory record on a foreign jurisdiction also tells you what cooperation is realistic – a country with repeated observations about non-enforcement is a country where mutual legal assistance on a labour exploitation case will be slow.

🔍 Private investigation and corporate security

For corporate investigators and compliance functions, this is the foundational source rather than a supplementary one. The indicators define what you are looking for, the recruitment fee definitions establish the employer-pays standard now embedded in most credible corporate policies, and the sectoral research tells you which parts of a supply chain carry structural risk. The supervisory record for a sourcing country provides documented, authoritative, citable evidence for a risk assessment file, which is precisely what regulators under modern slavery and forced-labour import regimes expect to see. Remember that it names no companies, so it establishes context and never conclusions about a specific supplier.

📰 Journalism and OSINT media

The Committee of Experts observations are a reporting resource that almost nobody uses, and they are gold: specific, sourced, official findings about forced labour in named countries and sectors, complete with the government's own response. Learning to read the escalation language turns a dry legal database into a running record of which governments are being told, in the most serious terms the system permits, that they are complicit. The global estimates are the headline that everyone else reports; the supervisory record is the story nobody has written. Cite the specific comment and year, and explain the register to your readers rather than paraphrasing it as an accusation.

🌍 NGO, humanitarian and human rights

For labour rights and anti-slavery organisations this is both an evidence base and a mechanism you can use. The supervisory system runs on submissions from workers' and employers' organisations, so a union or a federation with which you work can put evidence directly into a process that obliges a government to respond publicly. That is a lever, not just a source. The indicators support screening and training, the fair recruitment principles support advocacy on recruitment fees, and the country record supports submissions to other bodies. Be aware that in some jurisdictions making such a submission carries real risk for local partners, and that consideration comes before the advocacy value.

🎓 University and research

The supervisory record is a large, structured, longitudinal corpus of expert legal assessment that is substantially under-exploited in empirical research, and it supports work on compliance, on the effect of ratification, and on the political economy of labour standards. Treat it with the same care as any institutional record: the observation-generating process depends on ratification and on civil society capacity to make submissions, so the data is missing non-randomly in a way directly correlated with repression. The global estimates should be cited as estimates with their round and method, and the definitional breadth combining forced labour and forced marriage stated. ILOSTAT is the standard source for labour market covariates but check for series breaks.

Playbook: working ILO Forced Labour end to end

A repeatable sequence from first pull to finished product. Each phase states what you are trying to establish, not merely what to click — the objective is a defensible chain of reasoning, not a completed checklist.

Phase 1 — Fix the legal question before you collect anything

Establish whether you are asking about forced labour under Convention No. 29, state-imposed forced labour under Convention No. 105, child labour under Convention No. 182, or trafficking under the Palermo Protocol. These have different definitions, different obligations and different evidentiary requirements, and an analysis that blurs them will produce conclusions that do not hold up in any of the frameworks.

Phase 2 — Establish ratification status for every country in scope

Pull the ratification record for the relevant instruments with dates, and note any denunciations. This tells you what obligations exist and, just as importantly, how much supervisory material you should expect to find. A thin record for a non-ratifying state is expected and means nothing about conditions.

Phase 3 — Adopt the eleven indicators as your classification scheme

Map your own data – case records, audit findings, worker testimony, media reporting – onto the ILO indicators rather than inventing categories. This is what makes findings comparable across sources and legible to regulators, prosecutors and auditors, and it is free standardisation you should not decline.

Phase 4 — Harvest the supervisory record for your countries and sectors

Go into NORMLEX and pull every observation and direct request on the relevant conventions for your countries, going back at least a decade. Read them in chronological order. The trajectory tells you far more than any single comment, and the sectors and practices named give you targeting detail that no aggregate statistic contains.

Phase 5 — Learn the escalation ladder and score against it

Record where each country sits: routine reporting, direct request, published observation, Conference Committee examination, special paragraph, commission of inquiry. Treat this as an ordinal severity variable produced by an independent expert body over time, and use it in place of subjective severity judgements of your own.

Phase 6 — Read the government's own reports for admissions

Article 22 reports are states describing their own law and practice, and they routinely contain admissions, contradictions and descriptions of programmes that the government characterises benignly and the Committee does not. Comparing what a state says with what the Committee finds is one of the most productive reading exercises available in open source.

Phase 7 — Separate state-imposed from privately imposed exploitation

Split your analysis on this axis and keep it split. The two have different causal structures, different evidence bases and completely different remedies – one calls for prosecution and labour inspection, the other for diplomatic pressure, trade measures and sanctions – and a merged assessment produces recommendations that fit neither.

Phase 8 — Bring in the informality and inspection capacity indicators

Pull ILOSTAT indicators on informal employment and labour inspection resourcing for your countries. Forced labour persists where nobody looks, so these two indicators together are a decent proxy for the size of the space in which exploitation can go undetected, and they are measured rather than modelled.

Phase 9 — Resolve sectors to commodities and to actual supply chains

Convert the named sectors into Harmonized System codes and intersect with real trade flows and, if you are doing corporate work, with the client's supplier base. This is the step that turns an international legal finding into an operational risk register, and it is where most compliance programmes stop short.

Phase 10 — Triangulate with enforcement and case-level sources

Set the ILO picture beside trade enforcement actions, national prosecution statistics, case-level victim data and investigative reporting. Convergence gives you confidence; divergence is diagnostic. A country with severe supervisory findings and no domestic prosecutions has a political problem rather than an evidentiary one, and that is a reportable finding.

Phase 11 — Check currency before you assert

For every supervisory comment you rely on, confirm whether a more recent examination has taken place, and for every estimate confirm the round. The record lags events by design, so an assessment built from comments that are several cycles old should be presented as describing the last examination rather than the present situation.

Phase 12 — Write in the source's own vocabulary

Frame findings in terms of the specific indicators present, the convention obligation engaged and the supervisory status, rather than in the loose language of modern slavery. Products written this way are usable by lawyers, regulators and customs authorities, which is the entire point of grounding the work in the ILO framework in the first place.

The platform ships this as a step-checked workflow in playbooks.php, so progress is recorded against a case rather than held in someone’s head.

What to pair it with

No single source carries a finding. These are the datasets that corroborate, extend or contradict this one — and a source that contradicts is worth more than one that agrees, because it is the only thing that will tell you when you are wrong.

Source	Relationship	What it adds
NORMLEX	prerequisite	The standards and supervisory database itself: ratifications, national legislation, Committee of Experts observations and direct requests. The single most valuable component of this source.
ILOSTAT	extends	Labour statistics including informality, inspection capacity and migrant worker stocks, supplying the measured context in which forced labour findings sit.
Global Slavery Index (Walk Free)	corroborates	Co-partner in the joint global estimates and publisher of a country-level prevalence and government response index built on a different and more contested method.
Counter-Trafficking Data Collaborative	extends	Case-level victim records whose means-of-control fields map almost directly onto the ILO indicators, converting the framework into observed frequencies.
US DOL List of Goods Produced by Child Labor or Forced Labor	extends	Country-commodity pairs with documented evidence, providing the granular sourcing detail the ILO's sectoral findings imply but do not enumerate.
US Customs and Border Protection forced labour enforcement	extends	Where the framework becomes enforceable trade action against specific entities and goods, and therefore where the commercial consequences of a finding materialise.
Trafficking in Persons Report	corroborates	Annual country assessments covering forced labour alongside trafficking, generally more current than the supervisory cycle and written in plainer language.
IOM	corroborates	Co-producer of the global estimates and the largest direct assistance provider, whose operational reporting grounds the statistical picture in casework.

Legal, ethical and operational constraints

There is no privacy or data protection dimension to using this source – it contains institutional and statistical material, not personal data. The legal significance runs the other way: this is the source that determines the legal characterisation of conduct that other sources merely describe. That creates responsibilities. If you assert that a practice constitutes forced labour, you are making a legal claim with the potential to affect trade, employment and reputation, and you should be able to point to which indicators are present and which convention obligation is engaged. Where your product will inform a trade enforcement action, a sanctions designation or a procurement exclusion, the standard of evidence is higher than for an analytical assessment and the ILO framework is the reference against which it will be judged. Attribution obligations under ILO copyright apply to reproduction of publications. Be precise about what supervisory findings say: the Committee of Experts assesses a state's application of a convention, and quoting its findings as a determination that named private entities committed offences misrepresents the mechanism. Finally, in jurisdictions where labour organising is criminalised, circulating supervisory material naming local unions can expose those organisations, and that risk sits with your handling rather than with the publisher.

Operational security

Reading published legal instruments and reports leaks nothing of consequence. Systematic NORMLEX querying reveals which countries, conventions and sectors you are researching, and while the ILO is not an adversarial party, the general principle of separating sensitive research from attributable infrastructure applies. The genuine exposure is on the engagement side. The supervisory mechanism runs on submissions from workers' and employers' organisations, and if your work contributes to a submission, the submitting organisation is identified in the published record and the government concerned will know exactly who raised it. In states where independent labour organising is dangerous, that is a serious risk to real people, and it must be discussed with the partner organisation before anything is filed rather than treated as a procedural step. Similarly, if your analysis names a sector in a country in a way that identifies a specific workplace where informants are located, publishing it can expose them. The ILO framework encourages precision, and precision about location is where the risk concentrates.

Two rules that hold regardless of jurisdiction. Collection that is lawful is not automatically proportionate, and a dataset assembled for one purpose does not carry consent for another. Where the records concern identifiable people, the question is not only whether you may hold the data but whether holding it serves the purpose you are accountable for.

Is it earning its place?

Sources accumulate. Feeds get added during an incident and are never reviewed again, and a decade later the pipeline is carrying dead weight that nobody dares remove. These are the measures that show whether ILO Forced Labour is contributing anything, and they are worth baselining now so the answer is available later.

Proportion of your forced-labour findings expressed in terms of the eleven indicators rather than in loose descriptive language, which measures whether your output is usable by regulators and prosecutors.
Number of countries in your scope for which you have harvested the full supervisory history rather than only the most recent comment, since trajectory carries most of the signal.
Lag between the annual Committee of Experts publication and your ingest of the new comments, which should be days rather than months if you are using this as a monitoring source.
Share of country assessments that cite the escalation level explicitly, which is the check on whether you are using the ordinal severity signal or substituting your own judgement.
Count of sectoral findings you have successfully resolved to commodity codes and actual trade flows, which measures whether the legal layer is reaching operational decisions.
Number of instances where a government's own Article 22 report contradicted its public position and you captured it, which is a good proxy for whether anyone is actually reading the primary documents.
Whether any product has quoted a global estimate without its round and definitional scope, which should be zero and usually is not.

Beware of volume. Indicator counts rise easily and say almost nothing. Unique contribution — findings this source produced that no other source in your stack would have — is the measure that matters, and it is usually far lower than anyone expects.

Tradecraft notes

The distinctions that separate a competent analyst from a fast one:

Learn the escalation register. The difference between the Committee requesting information, expressing concern, expressing deep concern and urging immediate action is the difference between a routine query and the strongest criticism the system issues, and none of it looks dramatic to an untrained reader.
The absence of comment is a function of ratification and of civil society capacity, not of conditions. Treat a thin record as a statement about the mechanism, and go looking for why the mechanism is not producing rather than concluding that nothing is wrong.
Read the government's report and the Committee's observation side by side. The gap between them is the most reliable open indicator of state bad faith available on this subject, and it is a documentary gap rather than an inference.
Convention No. 105 is the instrument for state-imposed forced labour and it is the one to reach for when the concern is prison labour, political coercion or state-organised programmes. Analysts who only know Convention No. 29 miss the framework that actually applies to those cases.
The indicators are cumulative and contextual, not a checklist. A single indicator such as excessive overtime is a labour violation; a cluster including document retention, debt and restriction of movement is forced labour, and the ILO's own guidance is explicit that the assessment is holistic.
Separate the statistical products from the legal ones in your own head and in your citations. The global estimates are contested modelling; the conventions and supervisory findings are authoritative legal material, and conflating their reliability damages the credibility you get from the second.
Recruitment fees are where debt bondage begins in most migrant labour corridors, and the ILO's technical definition of what counts as a fee or related cost is the precise instrument for identifying it. It is far more useful in practice than any prevalence figure.
Use informality and labour inspection capacity as your structural risk proxies. Forced labour survives where nobody inspects, and those two measured indicators tell you more about where to look than most purpose-built risk indices.
Never let the ILO's institutional caution become your own. The organisation writes diplomatically because it must operate with its member states; an analytical product has no such constraint and should state plainly what the findings describe, while citing them accurately.

Questions analysts actually ask

What is the legal definition of forced labour?

Convention No. 29 defines it as all work or service exacted from any person under the menace of any penalty and for which the person has not offered themselves voluntarily. Both elements matter, and the menace of penalty is not limited to physical violence – it covers financial penalties, denunciation to authorities, retention of documents and other forms of coercion. The eleven ILO indicators are the operational tool for identifying it in practice.

What are the eleven indicators?

Abuse of vulnerability, deception, restriction of movement, isolation, physical and sexual violence, intimidation and threats, retention of identity documents, withholding of wages, debt bondage, abusive working and living conditions, and excessive overtime. They are assessed holistically rather than as a checklist – a cluster of indicators establishes forced labour, whereas one alone usually indicates a labour violation.

Where do I find the country-specific material?

In NORMLEX, the ILO's standards database. It holds ratifications, national legislation and the full text of Committee of Experts observations and direct requests, searchable by country and by convention. This is the most valuable and least used part of the ILO's output for analytical work.

Does the ILO name companies?

No. Its public output identifies states, sectors, regions and practices, because its supervisory mandate runs to member states rather than to firms. It establishes that a sector in a country carries documented risk, which is exactly what a risk-based due diligence process needs, and it never establishes anything about a specific supplier.

Is the global estimate reliable?

It is the best available global figure and it is still an estimate with real limitations – extrapolation, definitional breadth combining forced labour with forced marriage, and difficulty reaching closed and institutional populations. Cite it with its round, use the regional and sectoral breakdowns rather than the headline where you can, and do not present it with the confidence appropriate to the legal material.

Why does a country with obvious problems have almost no supervisory comments?

Usually because it has not ratified the relevant convention, or because there is no independent workers' organisation able to make a submission. The supervisory machinery is triggered by ratification and by constituent input, and repression of labour organising suppresses exactly the input that would generate scrutiny.

How current is the supervisory record?

It follows the reporting cycle, with the Committee of Experts meeting annually and countries reporting on a multi-year rotation, so a comment describes the last time the Committee examined the matter rather than the present. For fast-moving situations, use it as authoritative background and get currency from enforcement actions, national reporting and investigative sources.

How does this relate to trade enforcement regimes?

Directly. Forced-labour import bans and due diligence regulations in several jurisdictions use the ILO definition and indicators as their substantive standard, so evidence framed in that vocabulary is the evidence those authorities are set up to receive. Framing your findings any other way makes them harder to act on.

What is an Article 26 commission of inquiry?

The most serious procedure in the ILO supervisory system, established to investigate persistent and grave non-observance of a ratified convention. It has been used rarely, and its use on a forced labour convention is the strongest formal signal the organisation can send about a member state. Its existence for a country is a highly significant analytical fact.

Standards, formats and interoperability

What this source speaks natively, and what it has to be translated into before a partner can consume it. Work that arrives in a recognised format is easier to defend, easier to hand over and easier to automate against:

Forced Labour Convention, 1930 (No. 29) and its Protocol of 2014, the primary international definition and obligation.
Abolition of Forced Labour Convention, 1957 (No. 105), which addresses state-imposed forced labour specifically.
Worst Forms of Child Labour Convention, 1999 (No. 182) and Minimum Age Convention, 1973 (No. 138), for the child dimension.
The eleven ILO indicators of forced labour, the operational classification scheme used across enforcement, audit and research.
ILO general principles and operational guidelines for fair recruitment, including the definition of recruitment fees and related costs.
The International Classification of Status in Employment and related ILOSTAT statistical standards, which govern the labour market indicators.
Sustainable Development Goal target 8.7, the policy framework linking the estimates to international commitments.
The UN Guiding Principles on Business and Human Rights and the OECD due diligence guidance, which import the ILO standards into corporate obligation.

References

Primary documentation and authoritative references for this source. Publishers revise and retire material, so treat the retrieval date as part of the citation and re-check before relying on any of it in a formal product.

Forced labour and modern slavery — International Labour Organization. The topic entry point: definitions, the indicator framework, the global estimates and the current programme of work.
NORMLEX — International Labour Organization. Standards, ratifications, national legislation and the full text of supervisory comments. The single highest-value component of this source for analytical work.
ILOSTAT — International Labour Organization. Labour statistics including informality and labour inspection capacity, the measured context in which forced labour risk should be assessed.
International Labour Organization — ILO. The organisation's main site: reports, the Profits and Poverty series, sectoral research and the Conference documentation.
Walk Free — Walk Free. Co-producer of the joint global estimates, whose own index applies a different and more contested method to the same subject.
International Organization for Migration — IOM. The third partner in the global estimates and the largest provider of victim assistance, linking the statistics to operational casework.
List of Goods Produced by Child Labor or Forced Labor — US Department of Labor, ILAB. Country-commodity pairs with documented evidence, the practical bridge from ILO sectoral findings to sourcing decisions.
Forced labour trade enforcement — US Customs and Border Protection. Where the ILO standard becomes an enforceable import restriction, and the clearest demonstration of why framing findings in this vocabulary matters.
Trafficking in Persons Report — US Department of State. Country assessments covering forced labour and trafficking, more current than the supervisory cycle and useful as a cross-check.
Counter-Trafficking Data Collaborative — IOM and partners. Case-level data whose control fields map onto the ILO indicators, turning the framework into observed frequencies across real cases.

Link integrity: every reference above was verified with a live request when this page was generated. Where a publisher had moved or withdrawn a document, the link was repointed at a preserved copy in the Internet Archive and marked as archived. Anything with no reachable copy anywhere had its link removed rather than left to rot — the source is still credited, it simply cannot be linked.

Put it into practice

The Quantus Intel threat intelligence platform operationalises this source: it loads the eleven indicators as a controlled vocabulary, harvests and diffs the supervisory record so escalations surface as alerts, resolves named sectors to commodity codes and real trade flows, and keeps the authoritative legal material clearly separated from modelled estimates.. Browse the full source catalogue, or follow any tag above into the rest of the library.