Common Indicators for Coffee Sustainability
Guidelines & Key Information
This project displays the initial basic Indicators for farm-level coffee sustainability, as originally determined by the GCP and the SPF Indicator Working Group. COSA--with feedback from the members of the Global Expert Committee--developed and synthesized practical metrics to operationalize the indicators so they can be functional across origins and comparable over time. The approach builds on global experience refined with tens of thousands of surveys and the expertise of the Committee members.
In the framework for this project, no data will be shared. However, it lays the ground for the potential to exchange data and making it easier to do so if needed, be it between business partners, or for sending information to a sustainability standard, etc. The project’s underlying philosophy is that every party has sovereignty over their own data, and is not obligated to share it.
Impact vs. Monitoring Data
The following indicator approaches are built on a Monitoring methodology and not a Full Impact approach. The Monitoring approach generally relies on farmer recall of the most recent production year and reasonable local estimates that can provide good enough information in a simple way. This can facilitate wide adoption and use without the burden of full accounting which can be onerous for some organizations and farmers. Full Impact approaches can be used where desired and is in many cases compatible as it provides more accurate information, but requires more investment and time in detailed record keeping, accounting, and data gathering skills.
Sustainability Monitoring (through farmer surveys) usually relies on a single farmer's response per household--usually the head of household. The head of household can be any one person in the household, but is generally the farm owner or main decision maker. To track activities or other engagements provided to farmers through programs or initiatives, an organization may wish to capture additional information on multiple individuals in a household where relevant (e.g., all training or service recipients). COSA has a separate protocol for this type of producer and household identification and tracking and can provide that to interested organizations.
Producer Sampling Guidelines
Representativeness: While sampling all farmers in a target group (census) is ideal, sampling a portion of farmers can be appropriate if the farmers selected for the survey are representative of the target population as a whole. Being aware of the homogeneity of the farmer population is important as well as individual farmer locations. The ideal approach would be a simple random sample where the appropriate number of farmers are randomly selected from a list and surveyors go to that list of farms to conduct the surveys. COSA has a Sample Size calculator built for Monitoring applications specifically.
Accuracy of farmer recall (memory) diminishes significantly beyond one year, so try only to ask about the last production cycle. It is also optimal to visit farmers soon after the main harvest period (and ideally at approximately the same time each year). It is important to ask questions as close to the end of the last production year as possible to ensure that the full production and harvest cycle is included in the response. The production year refers to the end of the last harvest to the end of the corresponding harvest before that (12 month period).
Try to talk to the head of household for each farm (different people may give you different perspectives but typically the decision-makers will yield the most accurate results).
Quality checks in the first week of a surveyor’s work can also make a big difference; make sure surveyors stick to the specific questions as written.
Certification & Audit Data
Some of the indicator data below may be covered in Audits or through other Compliance inquiries. If an entity wishes to use that data to report on the indicator framework, please be aware of the following:
Compliance and audit data is usually collected on a much smaller sample of farmers than typical Monitoring approaches (audit sampling typically relies on square root sampling instead of a large enough population to ensure statistically sound results). This means that audit data may not be representative of the whole population.
Compliance data typically gives the user a binary result on a single topic, i.e., whether a certain condition was met or not. It does not usually convey the degree to which a certain condition was met, nor can it be used to see incremental change over time. Therefore, to achieve more control over the supply chain and improve the ability to remedy significant issues, it is strongly recommended to use the SMART indicator approaches detailed below (in fact, the approaches below could be built into an organization's Compliance assessment tools).