Please fill out this short user survey of only 3 questions in order to help us improve the site. We appreciate your feedback!
Evaluating smallholder livelihoods and sustainability in Indonesian coffee value chains
Last registered on December 29, 2020


Trial Information
General Information
Evaluating smallholder livelihoods and sustainability in Indonesian coffee value chains
Initial registration date
February 21, 2017
Last updated
December 29, 2020 9:09 PM EST

This section is unavailable to the public. Use the button below to request access to this information.

Request Information
Primary Investigator
University of Sydney
Other Primary Investigator(s)
PI Affiliation
Universitas Lampung, InterCAFE-LPPM IPB
PI Affiliation
University of Sydney
Additional Trial Information
On going
Start date
End date
Secondary IDs
According to the Indonesian Coffee and Cocoa Research Institute (ICCRI), Indonesia had about 1.2 million hectares in coffee production in 2012, 96% of which is managed by smallholder farmers (ICCRI, 2012). ICCRI estimates that the number of smallholder farmers working in the Indonesian coffee sector reached 1.97 million in 2012, with an average of 0.6 ha of land ownership per farmer. According to the 2014 State of Sustainability Initiatives Report, about 11% of Indonesia’s coffee production is certified organic or to a recognized sustainability standard (notably Starbucks C.A.F.E Practices, UTZ Certified, Fairtrade, Rainforest Alliance, or 4C).

This study is carried out in the Semendo region in South Sumatra. According to local government data (BPS, 2015), 15,440 ha of land in Semendo is currently cultivated with coffee, involving 8,698 households, mostly ethnic Semendo. Coffee farmers from 25 different Semendo villages (desa) are currently involved in the sustainability program that has been implemented by the Indonesian subsidiary of a leading international coffee trading company, since 2012. In 2012, the company also established a local buying station in Semendo and commenced enrolling farmers in a Common Code for the Coffee Community (4C) production unit. Access to the buying station provides farmers with additional returns to quality.

The company pays the cost of obtaining certification, including the cost of training farmers and undertaking required audits. It has established a Farmer Training Centre in the Semendo area and recruited a team of locally-based agronomists who manage the Internal Control Systems (ICS) for the program. This team undertakes farmer training to those groups involved in the program, and includes advice on how to apply fertilizers, composting, pruning, harvesting advice, pest management and marketing. As farmers become compliant with 4C standards and develop a trading relationship with the company through the local buying station, they may be recommended by the company to upgrade to Rainforest Alliance (RFA) certification, which imposes additional requirements, particularly around environmental practices.

The overarching goal of this research is to examine the impact on farmer livelihoods and poverty alleviation within Indonesian coffee-growing communities as a result of processes of verification or certification against different sustainability standards, with additional interest in establishing the role of access to markets in driving impacts.
External Link(s)
Registration Citation
Arifin, Bustanul, Jeffrey Nielson and Russell Toth. 2020. "Evaluating smallholder livelihoods and sustainability in Indonesian coffee value chains." AEA RCT Registry. December 29. https://doi.org/10.1257/rct.1346-2.3000000000000003.
Former Citation
Arifin, Bustanul et al. 2020. "Evaluating smallholder livelihoods and sustainability in Indonesian coffee value chains." AEA RCT Registry. December 29. http://www.socialscienceregistry.org/trials/1346/history/83100.
Sponsors & Partners

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information
Experimental Details
A randomized control trial (RCT) study that will involve a comparison between (i) a treatment group of randomly-selected farmer groups who will receive an upgrade from 4C verification to RFA certification, (ii) a control group of farmer groups who will remain with 4C verification for at least 3 years. This design will allow us to clearly evaluate the impacts of moving from 4C to RA.
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
The measurement strategy is meant to capture each step of a plausible theory of change linking the intervention (essentially, information on improved farm practices) to ultimate socio-economic outcomes. It is plausible that some impacts may not always transmit step-by-step through this channel (e.g., advice to reduce child labour could transmit directly into an outcome like child schooling, without being mediated through farm outcomes), however we believe that this pathway captures most impacts.

Information from the verification/certification intervention (training) leads to:
1. Increased farmer knowledge. We verify a few key aspects of farmer knowledge.
2. Improved farm practices. We focus on key indicators directly tied to key points of the training materials, including land management, environmental protection and management, use of permissible pesticides, fertilizers and other inputs, agricultural labour, and post-harvest processing of coffee. In addition to focusing on practices on the main coffee operation, we also consider changes in time allocation and land investment to other agricultural and non-farm activities (i.e., substitution).
3. Improved farm outcomes. We focus primarily on outcomes on the coffee farm: yields (yields per hectare), productivity, revenue, prices, and profits. In addition to focusing on outcomes on the main coffee operation, we also consider changes performance (yields, earnings) in other agricultural and non-farm activities (i.e., substitution).
4. Improved household earning outcomes. We focus on measure of income, including direct coffee income, but also broader measures of total household income, expenditure and time allocation. We additionally include measures of household subjective well-being.
5. Improved general household outcomes. We include some measures of general household outcomes such as education, health, and female empowerment.
Primary Outcomes (explanation)
-We consider two measures of household income: a general measure of "all household income" and a measure adding up revenues across a number of farm, non-farm and other earning activities.
-We construct indices of farm practices, by giving uniform weight to each element in a group of practices, and aggregating into an index.
-We construct a measure of farm productivity as the residual in a TFP regression, and the productivity residual in a stochastic production frontier estimation of the coffee production function.
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
This field context appears to offer a fair opportunity to evaluate the impacts of this 4C to RA transition through a randomised control trial (RCT). We have therefore proposed that the implementation partner stagger the remaining roll-out of RA certification amongst the remaining 57 farmer groups according to random selection generated by the research team. In February 2015, we held a workshop with the field staff of the implementing partner in Muara Enim to talk through the expectations and advantages of conducting an RCT. The implementing partner has agreed to this proposal and we have now signed a Collaborative Agreement of Understanding to proceeding on this basis.

We have agreed upon the following:

-Randomize the remaining 57 farmer groups (FGs) into 2 groups:
o 29 FGs: treatment (become RA certified immediately, in the 2015-2016 cycle)
o 28 FGs: control (do not become RA certified until the 2018-2019 cycle, at soonest (where the farmers are eventually RA certified is fully at the discretion of the implementing partners))
o The research team reserves the right to stratify the randomization according to certain criteria, e.g., at the village level, or according to livelihood maps

-(As opposed to previous practice) implementation partner will use a direct approach to promote RA certification amongst the randomly selected treatment FGs (i.e., by directly approaching those groups, rather than the previous approach of holding an open socialization campaign in the villages). Implementation partner will commit the necessary time and resources to attain the maximum take-up of RA training and certification amongst the selected treatment group, with a target of 100% take-up in the 2015-2016 training and certification cycle.
Experimental Design Details
Not available
Randomization Method
Randomization done by a computer using STATA
Randomization Unit
Farmer groups
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
59 farmer groups
Sample size: planned number of observations
980 respondents
Sample size (or number of clusters) by treatment arms
29 farmer groups for treatment, 30 farmer groups for control, with approximately 20-30 farmers per farmer group.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Most reasonable power calculations imply that experiment is reasonably-powered to detect a 0.5 standard deviation effect size on main outcomes, on conservative assumptions (i.e., power 0.9). While having two rounds of midline/endline data is helpful, given the high autocorrelation in farming outcomes there is a limit on how much this adds.
IRB Name
University of Sydney Human Research Ethics Committee (HREC)
IRB Approval Date
IRB Approval Number