To influence commodity market actors, it is necessary to identify which companies or entities are handling the commodities from specific territories with supply chain information. How do actors perform in the supply chain? Which companies are exposed to risk?
Agricultural supply chains generally involve a myriad of actors at both ends of the chains, i.e. producers and consumers. In the middle, though, far fewer actors handle, transport, process and export the commodities. In Côte d’Ivoire for instance, just five corporations export about half of the country’s cocoa produced by almost a million of smallholder farmers, through a few thousands of intermediaries, local traders and cooperatives.
Starting with the middle segment of supply chains
Tracking the middle segment of supply chains is an excellent starting point for supply-chain transparency. Tracing commodities from their point of export up to their first point of aggregation (mills, silos, cooperatives, etc.) enables linking supply chain actors to the subnational jurisdictions where commodity deforestation is monitored.
The middle portions of these internationally-traded commodity supply chains are generally regulated and monitored, if only for fiscal purposes, even if this information is rarely made available to the public.
That data is generally spread across different and disconnected reporting channels – for instance, the ownership of industrial assets might be reported to the fiscal administration, the monthly volumes of production per asset are reported to the Ministry of Agriculture, per-shipment export data is with the customs administration, etc.
- Best situation. In some cases, a national institution centralises all the data, facilitating supply chain governance. This is notably the case for organisations in charge of coordinating or monitoring the implementation of national certification schemes (e.g. the Malaysian Palm Oil Certification Council)
- Second-best situation. Relevant data to track supply chain connections is scattered across different administrations, and some of it may even only be held by private actors. For this type of situation, the methods used by the Trase initiative provide a solution adaptable to each context. Learn more about the method.
Step 4 milestones
- Database with best available traceability information to first-level supply chain assets or aggregators (cooperatives, mills, silos, etc.).
- In the absence of farm-level data, a stakeholder committee decision on a calculation method to allocate supply chain volumes per asset/aggregator to each subnational jurisdiction.
Tracking the middle segment of supply chains as starting point for supply-chain transparency. Source: EFI
Towards farm-level data in high-risk places
Regarding the upstream portion of the supply chains, the availability of farm-level sourcing information heavily depends on the supply chain and country. Numerous traceability efforts have been made by agribusinesses to trace sourcing back to individual farms. This more detailed traceability, from public and private endeavours, and its possible disclosure, can be the focus in high-risk places.
It should be noted that companies’ traceability investments generally end with their direct suppliers. Indirect sourcing, through other suppliers, traders, cooperatives, is rarely monitored, although it can represent a very significant proportion of agribusinesses’ supply chains. In Côte d’Ivoire for instance, the six biggest cocoa traders (marketing 74% of the production) source 60% from indirect sourcing and none of this is traced.
Reconstructing a supply chain map from independent or scattered datasets: the Trase method
The Trase initiative is the first transparency initiative that unlocked access to subnational supply chain information at scale, by using previously untapped datasets for the purpose of monitoring sustainability risks in supply chains. For a given supply chain, in the absence of an existing national information system, the Trase method allows for reconstructing the most likely connections at all stages in the chain through the matching of different datasets, including:
- Per-shipment export data, from customs authorities or vendors of bills of lading databases
- Fiscal data on the ownership and capacity of supply chain assets (plantations, storing and processing facilities, etc.)
- Traceability reports
- Production data
- Transportation data
The starting point for the reconstruction is to have quantitative information at the two boundaries that make up the middle part of the supply chain: production data per local jurisdiction and per shipment export data (per company).
The second stage is to produce an ‘asset map’ to locate all the facilities that make up the backbone of the supply chain: depending on the supply chain, cooperatives, warehouses, silos, slaughterhouses, mills, refineries, bulking facilities – all the installations through which the commodity is handled before export.
The third stage is to establish the links between all of the supply chain assets taking into account their throughput capacity and ownership. Sometimes, companies’ fiscal registration numbers – or other identifiers – appearing in production, fiscal and export datasets can be used to reasonably infer the links between certain industrial plantations or supply chain assets and specific export transactions. Other sources of information may be used, such as self-disclosure from companies on their sourcing locations, or independent surveys conducted on critical supply chain nodes with a high concentration of assets.
Finally, when no better information is available to infer the links between supply chain assets and production areas, spatially-explicit data on transport costs is used to optimise the estimations. Effectively, transport costs constrain supply chain connections more or less heavily depending on the commodity context.
The result is an approximation of the reality, knowing that the assumptions made at the different stages of the reconstruction introduce different levels of uncertainty. However, the method enables and spurs constant improvement of a given supply chain map when new data becomes available.
