Smallholder Vanilla Production

The Mananara Nord region of northeastern Madagascar is situated along the Bay of Antongil, looking outward toward the Indian Ocean. The bay has long been a site of global trade. The Dutch East India Trading Company established trading posts in the region during the early 1600s, and the area was a pirate hub throughout the seventeenth and eighteenth centuries (Ellis Reference Ellis2007; Pearson Reference Pearson1997). Throughout its history, the region has produced a variety of goods for international markets, including timber, honey, sugar cane, black pepper, and cacao. Today, the three main cash crops in the area are cloves, vanilla, and coffee. Cloves and vanilla have been in the region for nearly 100 years and gained prominence as an export commodity during the French colonial era. During colonial times, the French tightly controlled the vanilla market, forcefully restricting smallholder farmers in Mananara from cultivating vanilla beans in favor of supporting plantation production systems further north (Osterhoudt Reference Osterhoudt2017).

Today, the commercial center of the region is the town of Mananara Nord, which is surrounded by many smaller rural districts and villages. One of these villages is Behazo – the site of the following case study on vanilla organic certification.Footnote ² As with other villages in the region, the majority of Behazo households engage in both subsistence and cash-crop agriculture, organized into diversified and well-tended agroforestry fields. The region is known for producing high-quality vanilla beans – an identity that farmers are understandably quite proud of, given the difficulty of cultivating and curing vanilla.Footnote ³ In Behazo, the scale of vanilla cultivation varies considerably from person to person, ranging from a few kilograms of harvest to several tons. In this area, vanilla cultivation is not primarily a wage-labor enterprise, as famers tend to their own fields (although larger-scale farmers may hire help at certain points in the vanilla harvest cycle).

In Behazo, and across many regions of Madagascar, farmers often form local associations called fikambananas, usually comprised of between twelve and fifteen men and women. In the Mananara region, many of these associations focus on vanilla cultivation, with members sharing labor, equipment, and information on vanilla markets and cultivation techniques. Fikambananas also play a social function within communities, organizing events and workshops. For a fikambanana to be officially recognized by the Malagasy government, members must register paperwork both with the administrative offices in Mananara and in the provincial capital city of Toamasina. Beyond associations, the next level of farmer organization is a farmer cooperative, comprised of a group of twenty or so farmer associations. While cooperatives today are increasing in number and strength in the Mananara region, this is a relatively new development within the past ten to fifteen years.

Vanilla Trade in the Global Market

The average vanilla bean travels a long and highly mediated path from Behazo farmers to American and European kitchens. After harvest, some producers sell their “raw” (uncured) green vanilla beans to other families or associations for curing. Other farmers will cure their own vanilla beans – a months-long process that involves carefully blanching, sweating, sun-drying, and sorting the beans. Households then sell these “finished” beans to small-scale collectors who travel to villages to buy vanilla to resell to larger collectors based in the town of Mananara.Footnote ⁴ Mananara operators then sell this vanilla to export societies, often based in the larger port cities, which then export vanilla to international import and wholesale companies, which then sell to distributors in Europe and the United States. Distributors sell vanilla within domestic markets, including to retail outlets and food manufacturers.

Madagascar produces about two-thirds of the world’s vanilla exports, over half of which are sold to the United States (Odoux Reference Odoux2003). The primary region of vanilla production is the Sambava region in northeastern Madagascar, known as the “vanilla triangle,” to the north of Mananara (Andriamparany et al. Reference Andriamparany, Hänke and Schlecht2021). Insiders within the international vanilla market describe Madagascar’s vanilla market as “in no way standard” and “highly speculative and very sensitive to rumors” (Odoux Reference Odoux2003: 4). The volatility of the vanilla market has been especially pronounced since the 1990s, when the Madagascar vanilla market was deregulated in response to pressure from international trade organizations (Salas Reference Salas2007). The uncertainty within Madagascar’s market reverberates throughout the international vanilla industry, as other vanilla-producing countries adjust their vanilla prices in response to Madagascar’s prices.

Market deregulation, combined with a cyclone that destroyed much of Madagascar’s vanilla crop, contributed to a dramatic price spike in 2003, with export prices climbing from approximately $70 per kilogram to $500 per kilogram (Salas Reference Salas2007). This dramatic boom was followed by an equally dramatic crash in 2004; by 2006, international vanilla prices had lost about 75 percent of their 2003 value (Salas Reference Salas2007). The market downturn hit rural Mananara farmers particularly hard. With such low prices, many vanilla collectors simply stopped coming to Mananara. The collectors who did continue to work in Mananara offered growers extremely low prices for their gourmet vanilla. The depressed vanilla market stretched on for more than a decade, finally recovering around 2016.Footnote ⁵

During the prolonged years of the vanilla bust, farmers struggled to earn enough income to support their basic needs; many families dedicated less time to their vanilla crops, concentrating on the clove market instead. Other farmers became curious about new forms of trade relationships including organic certification, which in 2004 was beginning to gain momentum in international markets. In Behazo, one such group was the local fikambanana VOKATRA, whose members contacted the Peace Corps Madagascar office to request a volunteer to help them pursue new market opportunities. Peace Corps agreed to the request, and, in 2005, at the lowest point in the vanilla market, my partner and I arrived at the village of Behazo as Peace Corps volunteers charged with the task of helping to develop an organic certification program in partnership with local vanilla farmers.

Trade Agreements: The Written and the Spoken

The dramatic altercation between farmer association members had been foreshadowed since the beginning of the certification project, which had consistently encountered a range of suspicions, misunderstandings, and frustrations. These stemmed in part from the fact that the ECOFIN project, as with most projects of organic certification, primarily equated traceable knowledge with written knowledge. For ECOFIN, legible, transparent traces were written traces, as exemplified by the requirements for voluminous paperwork, numbered harvest batches, and written contracts (Hetherington Reference Hetherington2012; Mutersbaugh Reference Mutersbaugh2002; Randrianarivelo et al. Reference Randrianarivelo, Randrianarivelo, Abhukara, Fitzgerald, Pargee, Vent, Auerbach, Rundgren and Scialabba2013; Skaria Reference Skaria, Amin and Chakrabarty1996). A written trail theoretically allows knowledge to cross scales between vanilla farmers, trade intermediaries, and consumers, allowing actors separated by space to keep track of the movement of materials through supply chains. Such legible paperwork is assumed to support more ethical relationships of trade and to eliminate dishonest practices.

Privileging written forms of knowledge, however, overrides the oral forms of traceability that Malagasy farmers carefully establish within their own trade relationships. In many regions, spoken agreements are a component of speech-making and public pledges, and bind together people and ancestors in social contracts (Bloch Reference Bloch1998). In many of the fikambanana certification meetings I observed, for example, one person would present information and ideas in an oral format, then another member would repeat back the information to confirm what was said. Often, the person responding would elaborate on key points or raise follow-up questions. These interactions followed familiar oral templates for public rhetoric common in this region of Madagascar (Osterhoudt Reference Osterhoudt2017). This format leaves space for people to raise questions, confirm plans, and establish leadership. Such spoken interactions, done before a group, also create a community of listeners who can hold parties accountable for promises made.

The certification programs, however, largely ignored such spoken and dialogic methods for establishing transparency. As such, they overlooked the important social, economic, and cultural weight given to verbal contracts and oral performance to cement relationships, including relationships of trade. Instead, the certification organization insisted that, to be transparent, agreements must be written down; including a spoken component to trade agreements was not regarded as necessary to the process.

Yet written agreements between certification agencies and farmer associations are often written in a language and in terms that farmers are not familiar with. Farmers are not given space to negotiate the terms of these contracts, which are drawn up by certification agents in consultation with the private sector partners who plan to buy the certified products. Written agreements are therefore usually negotiated, written, and safeguarded by individuals considered to fall outside of local networks of trust (Skaria Reference Skaria, Amin and Chakrabarty1996).Footnote ⁷ This produces trade agreements that are more transparent for certification agencies and other extra-local actors, but less transparent for farmers.

The exclusive reliance on written knowledge also has implications for accountability – an essential component of transparency projects. In the case of any potential dispute within certification projects, the written contracts will carry much weight in the resolution process. For many Malagasy farmers, however, written contracts are more difficult to enforce than their oral counterparts (Bloch Reference Bloch1998; Osterhoudt Reference Osterhoudt2017). For example, if farmers feel that a written contract has been broken, they will likely be required to find recourse through the courts – a system highly skewed against smallholder farmers. The formal court system contrasts with more local ways to settle grievances – for example, through a council of elders or analogous community governance structures such as dina agreements for local natural resource management (Rakotondrabe and Girard Reference Rakotondrabe and Girard2021). A council of elders, however, would likely not feel confident to comment on the intricacies of an international trade contract drawn up in French or English if it came into dispute, nor would the nonlocal partners in the contract be likely to agree to settle any claims in this way.

Written contracts therefore set up processes of accountability that favor certification agencies and vanilla buyers, who can more easily enforce contracts if the farmers are seen to break agreements. Farmers, on the other hand, do not have the same available resources to enforce written contracts if the certifiers do not follow through on the agreed terms, or if they engage in corrupt or dishonest practices. This protection from accountability confers much power on certification agencies (see also Moberg Reference Moberg2014).

Of course, many local processes of accountability are also skewed in favor of farmers with more political influence and economic resources. Yet, while both processes have their strengths and weaknesses, one is more fully situated at the local level, which brings transparency in its own right: farmers better understand what is stacked against them within more local systems, and have established strategies for navigating them.

Field Inventories: How to Count Vanilla

The emphasis on written records privileges knowing things through counting and inventories – popular tools of many development and conservation initiatives (Li Reference Li2007; Muehlmann Reference Muehlmann2012; West Reference West2006). Yet, counting is not without cultural nuance and political maneuvering. For example, organic certification programs require farmers to carefully estimate their anticipated vanilla harvest for the season. These estimates are recorded in individual field inventory sheets and then compiled across all farmers in the association. Harvest estimates are used to determine the quantity of vanilla that becomes organically certified; this number also helps to determine the cost of annual organic certification.

Most often, organic certifying agents are worried that farmers overestimate their production for the year. If the group is certified for an anticipated ten tons but produces only seven tons, farmers could potentially bring in three tons of “counterfeit” vanilla and claim it as part of the ten tons under certification. Thus, the certifier who came to the village for Behazo’s pilot project was keenly interested in verifying that the estimates farmers gave for their vanilla production matched what she saw in the fields. To accomplish this task, we visited several sample fields to ground-truth data on the farmers’ projected yields.

Once in the fields, however, the certifier noticed the opposite trend to what she had anticipated: instead of exaggerating the extent of their vanilla vines under cultivation, the farmers had consistently underestimated the number of vines – often to a significant degree. Observing this trend, the certifier remarked dismissively that the farmers did not seem to know what was growing in their own fields. Calling an impromptu meeting, she then took time to explain to farmers how to count their own vanilla vines. I was surprised by her conclusion that farmers lacked skills to estimate their own vanilla production – in my experience, vanilla farmers had a good knowledge of how many vines they had under cultivation.

When I spoke with farmers about this discrepancy a few weeks after the inspection, they explained to me that people did not miscount their vines; rather, they deliberately underestimated their harvest numbers. They gave several reasons for this decision. One was that the younger farmers did not want to publicly announce that they had more vines than the elder farmers, which would be a sign of disrespect that could disrupt social relationships within the fikambanana. People with a potentially large vanilla harvest did not want to become a target for theft, nor for people asking them for money or favors. Many vanilla farmers did not want to commit to specific production numbers, knowing the many challenges that could prevent a good harvest, including cyclones, pests, or diseases. Finally, being a new project, farmers were hesitant to pledge all of their vanilla to certification, preferring to follow the risk-hedging strategy of diversifying their trade relationships across various buyers and markets (Osterhoudt Reference Osterhoudt2017).

Thus, it was not that the farmers needed to be educated on how to count their vines, but rather that they had specific reasons for miscounting them in the first place. In the end, it was the “numerically correct” counting desired by certification agents, and not the “socially correct” forms of counting useful to farmers, that prevailed. Before ECOFIN agreed to certify the vanilla as organic, farmers were required to go back and “correct” the anticipated yield numbers on their inspection sheets.

Representing Transparency: The Spreadsheet

The organic certification project not only privileged certain forms of writing and counting, but also how this information was represented (see also Galvin Reference Galvin2018). Most notably, the process encouraged mastering the Excel sheet format, as became evident during a follow-up visit I made to the organic certification project in 2016. Nine years after the first disastrous attempt at organic certification, I again sat in on a meeting between Behazo farmers and an ECOFIN agent. This time, the farmers were part of a large cooperative, with professional leadership and a membership of hundreds of farmers. The meeting was not held outside a village home, but in a fancy hotel in Mananara. Instead of scattered papers pinned under rocks, the cooperative representatives worked on laptops, with the field inventory data neatly presented in the form of Excel spreadsheets. The data were beautifully legible, arranged in neat columns. Looking at the screen, the ECOFIN agent complimented the group on its work. There was a cursory visit to a few vanilla fields, but the spreadsheet inspired trust, and the ground-truthing of reported harvest numbers no longer played as much of a role in the proceedings.

I left the meeting impressed by how well the farmers had figured out what the murky system of transparency was supposed to look like, through the eyes of the certifier. They realized the connection that certification programs make between the quality of information presented and the presumed quality of the thing that the information represents (Galvin Reference Galvin2018; Hetherington Reference Hetherington2012; Tracy Reference Tracy2016). With the presentation of an Excel spreadsheet that inspired trust, the materiality of the vanilla vines faded into the background.

Economic Transparency, to a Point

A common goal of transparency projects is to bring to view problematic economic relationships, such as corruption, predatory lending, paying workers nonliving wages, and extreme profit capture. Organic and fair trade programs, for example, often circulate narratives that criticize “middlemen” involvement in commodity trade networks, connecting the activity of trade intermediaries to decreased profits for farmers. Certification agencies promote their projects as an alternative to this model, noting that participating farmers receive higher prices for their products, in part by forging more direct linkages between farmers and consumers (Osterhoudt et al. Reference Osterhoudt2020).

This baseline narrative, however, does not disclose how certification agencies themselves act as “middlemen” in the trade process. These organizations charge significant costs for certifications. Although farmers and businesses may receive welcome price premiums for organically certified vanilla, these premiums must be balanced against the considerable direct and indirect costs to farmers and businesses to obtain certification. Organic certification in Madagascar is conducted by accredited for-profit entities, and they charge a good deal for their services – services that must be performed each year. For businesses engaged in certification programs, they must often hire teams of people to manage certification paperwork and reporting requirements, representing another significant expense.

In addition to the direct costs of certification fees, the organic certification process includes significant indirect costs (Randrianarivelo et al. Reference Randrianarivelo, Randrianarivelo, Abhukara, Fitzgerald, Pargee, Vent, Auerbach, Rundgren and Scialabba2013). For farmers, there are the indirect costs of lost work time and the need to purchase new materials required to pass certification, such as vanilla bean drying racks, wooden storage boxes, or forms of organic fertilizer. There are also the indirect costs for both farmers and businesses that come from hosting certifying agents during their site visits. These agents stay in villages for several days, and often expect to be shown a “good time” by farmers. After nine years of hosting ECOFIN project representatives, for example, farmers noted that they feel pressured to buy the agent lavish food and drinks and put them up at the most expensive hotel in the region. A happy agent, they say, is more apt to give them a better inspection.

In their project narratives, however, certification agencies do not make readily visible the direct and indirect costs they charge to farmer associations and vanilla buyers. Organic certification agencies often present their work as purely a social mission, instead of as a socially minded business, and many consumers do not realize that they rely on extracting profits from the commodity chains they certify. Overall, while making some forms of economic costs transparent, certification programs obscure their own role as economic actors within the vanilla commodity chain.

Ultimately, certification agencies act in a role analogous to the “middlemen” they criticize, as they capture significant profit and value from agricultural supply chains. These outside agencies often replace local Malagasy individuals involved in the vanilla supply network – usually younger Malagasy men and women from local villages who transport vanilla from place to place, connecting farmers with buyers. Organic certification projects, instead of empowering local Malagasy individuals to perform the paid work of certification, bring in people from outside the region to profit from the vanilla trade. Certification projects also open up space for people like me – an American with no experience in vanilla farming – to capture forms of value within the vanilla supply chain (Corson Reference Corson2016; Guthman Reference Guthman1998). In the end, certification programs do not eliminate trade intermediaries as much as they replace local intermediaries with extra-local ones (Osterhoudt et al. Reference Osterhoudt2020).