'Opposites attract' described how globalisation is leading many NGOs to turn their attention away from national governments and towards business, and either confronting companies directly or seeking opportunities to collaborate. One result of this is the growth of NGO-operated certification schemes that focus on social or environmental aspects of business.

Some individuals in the NGO sector have expressed concern about working with corporations on policy issues, believing that there is a danger of becoming co-opted. Collaboration between businesses and NGOs can be controversial, and pose significant challenges to the participants.

One such initiative was developed by the US-based Rainforest Alliance and the US-based corporation Chiquita Brands International. The Rainforest Alliance is dedicated to 'the conservation of tropical forests for the benefit of the global community'. Meanwhile Chiquita, one of the largest agricultural firms in the world, was Costa Rica's third largest banana exporter in 1999, accounting for more than 18% of the total $623 million worth of exports.

The relationship between the two organizations has its origins in an expansion of banana production in Costa Rica in the late 1980s. The area of land area under cultivation increased from 20,000 hectares to 50,000 hectares in just five years. This had the effect of heightening concerns about the environmental and social impacts of banana plantations. Already Chiquita had faced legal action in the US by Central American employees who claimed that handling certain pesticides had caused them long-term health problems.

In 1992, two NGOs - the Rainforest Alliance and the local Fundacion Ambio - and one banana producer, Chiquita, agreed to work together. The result of this collaboration was the 'Better Banana' project (BBP). Eight years later this certification scheme, now run by a network of NGOs coordinated by the Alliance, had become one of the largest eco-labeling initiatives in the world, certifying coffee, citrus fruits and bananas. More than 160 banana farms were certified by the Better Banana scheme, covering 120,000 acres in Ecuador, Colombia, Panama, Guatemala, Honduras and Costa Rica.

The report Growing Pain? provides an in-depth analysis of the history of this initiative, considering how it was managed, what it achieved, and the arguments deployed by its critics. Conclusions are made about the management of certification schemes. Below are the key findings from that study:

Firstly, the collaboration is controversial. Rainforest Alliance and Chiquita contend that since 1993 they have begun changing the face of tropical agriculture.

However, critics have claimed that the scheme was an exercise in 'greenwash', and that it did not tackle the most important issues in banana production.

Rainforest Alliance's and Chiquita's motivations for working in partnership mirror those identified in previous studies of business-NGO relations. The specific case of banana production in Costa Rica was felt to be a good area in which to advance the NGO's aims, namely protecting forests and reducing the environmental impacts of plantations.

Chiquita recognized that the market was changing and becoming more sensitive to environmental issues. However, it did not wish to develop itself as a player in a 'niche' market (such as organically grown produce). It therefore sought an 'intermediate' definition of 'environmentally friendly'.

Chiquita recognized that it needed to consult outside the company and gain credibility for its initiative. Building alliances with NGOs was seen as a way to do this.

Moreover, the company's management believes that the information revolution means there is no alternative to transparency in demonstrating corporate social responsibility.

The Better Banana certification system was based on earlier experience with forest certification. It proved that transferring a model from one sector to another raises some fundamental questions, such as:

Is certification the best way of dealing with this issue?

How do the industries compare: is certification practical?

How does the context compare: is there enough interest to make it happen - or to stop it from happening?

What was it that made a particular initiative successful?

Managing the alliance between a northern NGO and its southern counterpart posed some difficulties of its own. In particular, the division of work between the two organizations was not clearly defined at the outset, leading to tension and complicating relations with certified companies. For example, on occasion Chiquita received conflicting communications from the two NGOs.

The Better Banana project is now operated by the Conservation Agriculture Network - a group of Latin American NGOs - with Rainforest Alliance acting as the secretariat. The CAN is intended to expand the role of southern NGOs, and pluralise the NGO relationship with Chiquita.

Although intended as a certification scheme that would be open to any company, the involvement of Chiquita had the effect of overshadowing the participation of other banana companies. This 'closeness' was a mixed blessing. On the one hand, trust between the participants led to rapid learning about how to improve environmental aspects of banana plantations, and how to operate a certification scheme. On the other hand, this trust to some extent compromised the NGO's role as auditor.

The close relationship also meant that the Rainforest Alliance did not require as much evidence of social and environmental improvements as it might, leading to disputes about what BBP had actually achieved.Although incidents of non-compliance with the standard (such as aerial spraying while workers were in the fields) were reported, no 'de-certifications' were made.

Some critics believed that the Rainforest Alliance became financially dependent on Chiquit, and was thus beholden to follow the company's directions. In reality, the relationship was more complex, with the company making donations to the Alliance at the start but then stopping. BBP now charges fees for certification assessments, but remains dependent on additional charitable funding from the Alliance.

Given growing concerns about the social impacts of banana production, the participants increasingly considered issues beyond the original interest, expertise and mandate of the participating NGOs. Although Better Banana addressed the issue of employees' health and safety in a detailed fashion, it did not effectively deal with workers' rights issues, such as freedom of association, and freedom from harassment and sexual discrimination.

This was because the auditors were environmental experts, using the concept of sustainable development as a framework for understanding social issues.

As a result, the project had the unwanted effect of alienating certain groups in civil society, such as trades unions. By 2000 it recognised this shortcoming, engaged with those groups, and began improving the social criteria of the standard.

As the Better Banana project became increasingly influential in Latin America, the Rainforest Alliance's accountability to a range of stakeholders became more important.

Although representatives from trade union, environmental, producer and community groups were involved in the early stages, once the certification system was developed there was no system whereby they could participate in reviewing the standards or the monitoring system.

Moreover, even if such groups had been invited to participate on an ongoing basis, there would still be the problem of NGOs, such as the Alliance, controlling the process, despite not being primary stakeholders. Thus multi-stakeholder ownership of any system for endorsing corporate practice is essential.

Despite the setbacks, the experience of Better Banana suggests that independent certification offers a mechanism for delivering and communicating better corporate social and environmental performance.

(By Jem Bendell)