WokeCorp
Sourcing

Cocoa's 25-Year Broken Promise: The Harkin-Engel Protocol

In 2001, chocolate companies pledged to end child labor in cocoa by 2005. 1.56 million children still work in hazardous conditions in the same two countries named.

By The WokeCorp Team
Cacao pods growing on a tree in West Africa
Côte d'Ivoire and Ghana produce roughly 60% of the world's cocoa. The 2020 NORC survey found 1.56 million children in hazardous labor in those two countries alone. · Photo via Pexels. Pexels License.

The largest chocolate companies in the world signed a document in September 2001 pledging to eliminate child labor from cocoa by 2005. The Harkin-Engel Protocol named Côte d'Ivoire and Ghana specifically. In 2020, a NORC at the University of Chicago survey commissioned by the US Department of Labor found 1.56 million children in hazardous labor in cocoa farming in those two countries. The deadline had been extended four times by then.

That gap, twenty-five years, four deadline extensions, no binding enforcement, is what the Harkin-Engel Protocol actually produced.

Key findings

  • The Harkin-Engel Protocol was signed September 19, 2001 by Nestlé, Mars, Hershey, ADM, Cargill, Barry Callebaut, and others.
  • The original commitment: eliminate the worst forms of child labor in cocoa in Côte d'Ivoire and Ghana by July 1, 2005.
  • The 2005 deadline was extended to 2008, then 2010, then 2020.
  • The 2020 NORC survey found 1.56 million children in hazardous child labor in cocoa in those two countries.
  • The CLMRS had enrolled roughly 110,000 households as of 2020, against a farming population of approximately two million.
  • No binding replacement deadline has been set.
  • Nestlé v. Doe (2021): the Supreme Court dismissed the case 8-1 on jurisdictional grounds, finding the alleged conduct didn't connect to US territory sufficiently.

The protocol's structure, voluntary and unenforced

The protocol's essential feature is that it was voluntary. No government agency enforces it. No certification body can withhold a company's right to source cocoa for non-compliance. The consequences for missing the 2005 deadline were a new deadline. The consequences for missing 2008 were 2010. The consequences for missing 2020 were a statement about continued commitment.

The mechanism the industry built to address the problem, the CLMRS, is a real program with real operations. Participating companies fund household surveys that identify children working in hazardous conditions, enroll those households in remediation programs, track school attendance, and report annually on progress. This is more than most supply-chain ethics programs do.

The problem is scale. Two million smallholder farming households in Côte d'Ivoire and Ghana. The CLMRS enrolled roughly 110,000 of them as of the 2020 NORC assessment. The NORC surveyors found 1.56 million children in hazardous labor anyway.

Farmers working in agricultural fields in West Africa

Cocoa farming in West Africa is predominantly smallholder, with farms typically under five hectares. The CLMRS program's household coverage, at roughly 5–6% of the farming population as of 2020, does not reach the scale of the problem the protocol named. Photo via Pexels. Pexels License.

What the 2020 NORC survey actually measured

The NORC survey's methodology matters here. "Hazardous child labor" under ILO Convention 182 includes tasks that endanger health or safety, using sharp tools (machetes used to open cacao pods), carrying heavy loads, applying pesticides, working at night. These are the conditions common in cocoa harvesting that the protocol targeted. The 1.56 million figure is the number of children doing those specific tasks in the specific geography the 2001 document named.

By contrast, the industry's preferred metric is "children in child labor", a broader category that also includes the larger number of children who help on family farms under conditions not classified as hazardous. The hazardous subset is smaller and represents the protocol's actual stated target. The NORC found that the number of children in hazardous labor in Côte d'Ivoire had actually increased from 2009 to 2018–2019.

| NORC Survey Year | Children in Hazardous Labor (CIV + Ghana) | |---|---| | 2009 | 1.77 million | | 2013–2014 | Approximately 1.48 million | | 2018–2019 | 1.56 million |

The net movement is modest. The direction for Côte d'Ivoire, the larger producer, is upward over the survey period. The Harkin-Engel Protocol was signed in 2001 with a 2005 target. The NORC in 2020 found the metric it was supposed to eliminate to be essentially unchanged from 2009.

The litigation angle. Nestlé v. Doe (2021)

Six Malian citizens filed suit against Nestlé USA and Cargill under the Alien Tort Statute, alleging that the companies knowingly aided and abetted the use of child slavery in cocoa supply chains. The case reached the Supreme Court as Nestlé v. Doe.

The Court's 8-1 decision in June 2021 dismissed the case on narrow grounds: the plaintiffs hadn't sufficiently connected the alleged harms to US territory. The Court held that the domestic conduct alleged, purchasing decisions and supply-chain oversight made at US headquarters, wasn't enough under the ATS without a clearer nexus to the harms occurring in Côte d'Ivoire.

The decision didn't adjudicate whether the companies' conduct was ethical, whether the child slavery alleged actually occurred, or whether companies with knowledge of their supply chains' conditions bear any responsibility. It answered a jurisdictional question about a 1789 statute. The NORC numbers are what the merits look like.

Container ships at a commercial port loaded with cargo containers

The cocoa supply chain runs from smallholder farms through collectors, cooperatives, exporters, processing facilities, and commodity traders before reaching the brands whose names appear in the grocery aisle. The CLMRS operates at the farm level; most of the supply chain sees no CLMRS oversight. Photo via Pexels. Pexels License.

Why hasn't child labor in cocoa been eliminated after 25 years of pledges?

The CLMRS's structural challenge is economic. Cocoa is grown by approximately two million smallholders across Côte d'Ivoire and Ghana. The average farm is under five hectares. The income is low. Child labor in cocoa exists because children on family farms represent an available labor source at a cost (their lost schooling) that the farmgate price doesn't internalize.

Addressing child labor at scale requires either paying farmers enough that they can afford hired adult labor, or funding remediation (school enrollment, alternative livelihoods) at a scale proportionate to the problem. Neither has happened. The CLMRS enrollment numbers confirm that the investment level has not reached the scale of the declared target, twenty-five years in.

The companies that signed the 2001 protocol are worth hundreds of billions of dollars combined. The NORC survey was commissioned by the US Department of Labor, not by the signatories, because the signatories had declined to fund independent progress measurement. That detail is its own data point.

Workers drying cocoa beans on a plantation in West Africa

Cocoa farmers engaged in traditional bean-drying methods, a step central to the supply chains covered by the Harkin-Engel Protocol. Despite two decades of voluntary commitments by major chocolate companies, independent surveys continue to document child and forced labor in these same farming communities. Photo: Zeal Creative Studios via Pexels. Pexels License.

The WokeCorp assessment

The commitment. The Harkin-Engel Protocol is one of the most detailed and publicly visible corporate-responsibility pledges of the past three decades. It was signed by the relevant companies, named the specific problem, and set a specific deadline. That directness makes the gap more measurable than most.

The outcomes. Four deadline extensions and 1.56 million children in hazardous labor in 2020 is the record. The CLMRS is a real mechanism. It covers roughly 5–6% of the farming households in the target countries. The protocol's stated goal has not been achieved on any timeline the signatories proposed.

The core question. Whether this is a failure of ambition, resources, coordination, incentive structure, or something else depends on which uses you pull. What the Harkin-Engel Protocol clarifies is that voluntary commitments with no enforcement mechanism, funded at a fraction of the industry's marketing spend, do not eliminate a structural economic problem in a twenty-five-year window.

Compare with Nike's supply chain commitments, a different industry, the same audit-versus-reality gap. The brand value of being "ethically sourced" is real. The operational cost of actually being so has, in both cases, not been paid.

Sources

  • Harkin-Engel Protocol, September 19, 2001. World Cocoa Foundation. Verified June 2026.
  • NORC at the University of Chicago, Assessing Progress in Reducing Child Labor in Cocoa Production in Cocoa Growing Areas of Côte d'Ivoire and Ghana, October 2020. Commissioned by US Department of Labor ILAB. Verified June 2026.
  • Nestlé v. Doe, 593 U.S. ___ (2021). Supreme Court of the United States. Verified June 2026.
  • US Department of Labor ILAB, 2019 Findings on the Worst Forms of Child Labor: Côte d'Ivoire. Verified June 2026.