Project selection
HRIA of the Green Resources Uchindile plantation in southern United Republic of Tanzania was undertaken to examine common and divergent interests of health impact analysis and the context of HRIA. Uchindile is located on the boundaries of Iringa and Kilombero districts, approximately 100 km from Iringa town, accessible on rough roads. It was selected for its rural location, where impacts could clearly be allocated to the project, not to third-party actors in the area, which did not exist when the project began in December 2008. It also has high poverty and infectious disease rates, low education and employment opportunities, and a growing migrant workforce. In short, the human rights baseline suffered from low state capacity to fulfil rights, and there were many ways in which the project could interact with existing human rights conditions, positively and negatively.
Approach to evaluation
Uchindile plantation, founded in 2000, is owned and operated by Norway’s Green Resources AS. Assessors from NomoGaia, a non-profit think tank that builds and tests corporate human rights due diligence tools, examined likely impacts associated with the plantation’s transition from planting into harvesting. Initial assessment was timed to precede the transition to harvesting because a variety of workforce changes and health risks arise with the use of heavy machinery for tree felling and transport that are not needed during the growing and pruning stages of forest development. Assessment was continued periodically over the ensuing six years because changes in human rights conditions are ongoing. This is partly because one change in human rights conditions can trigger others (e.g. improved access to food can improve health outcomes), but also because harvesting operations occur over a shifting space – once trees are felled in one area, harvesters move to a different area. Furthermore, tree harvesters are semi-skilled workers, while the local area is populated by unskilled workers. Human rights impacts were considered possible as higher-paid workers were brought to the area to carry out paid work on land that was once held by local residents.
Green Resources provided interviews with all major management personnel (14 interviews over the course of three site visits) and a site tour. The assessment was not commissioned by the company and was externally funded by NomoGaia. The company was a willing collaborator in assessment, interested in human rights findings and willing to share data and facilitate interviews.
HRIA was carried out using the NomoGaia methodology, as described above, comprising scoping, cataloguing, scoring mitigation and monitoring [
20]. Scoping entailed a systematic review of all publicly available audits, company financial reports, local and regional health and development reports and existing ethnographic studies in the Mufindi area. Certification reports, EIA, management plans, community questionnaires, annual reports and policy documents were studied as well as Tanzanian laws, Ministry of Health (MoH) reports and data from the national census and two Living Standards Measurement Surveys (LSMS) conducted in 2008 and 2010. A systematic search of all multinational publicly traded companies in Mufindi district, revealed foreign funding for the Mufindi paper mill and the presence of Unilever. Public documents pertaining to these sites were obtained to contribute to context analysis. Additionally, a Google Alert for “Mufindi,” “Iringa,” “Uchindile” and “Green Resources AS” between 2008 and 2014 alerted authors to news stories and activist reports during the assessment and monitoring period. Peer-reviewed literature in the fields of public health, economics, history and anthropology were drawn from a screening of authors’ personal collections as well as a Google Scholar screen for the same terms listed in Google Alerts. Additional national-level data were drawn from international databases, as standardised in the HRIA methodology (
e.g. ILO, UNICEF, UN and WHO data). Data more than 10 years old and not from the Kilombero or Iringa districts were excluded. Data included reports from the grey literature to document both perceptions and misperceptions presented by outside observers and analysts without direct experience in the project area. All data were catalogued alongside sources, and all data were cross-checked during interviews with rightsholders, company personnel and local leaders, clinicians and other relevant authorities.
Cataloguing and monitoring involved primary data gathering and five site visits (March 2009, February 2010, November 2010, November 2013 and March 2014), each lasting between 5 and 10 days, involved engagement with health, education and government personnel (key informants) and rightsholders. Rightsholders are inhabitants of the project area whose human rights are likely to be impacted by project development and operations. Initial site visits represented a baseline from which observations in later visits were benchmarked. Rightsholder interviews were conducted with the most marginalised stakeholders, rather than with a random sample. Key informant interviews helped identify rightsholders experiencing disparate impacts. Semi-structured interviews asked informants to identify “outsiders,” people not considered part of the community and people not involved in community decision making. Particular probes were used to differentiate the power dynamics among men and women, first and second wives, locals and emigrants, and people of various educational attainment and skill levels. Interviews with women, youth, emigrants and other population subgroups enabled deeper exploration of relevant issues through a process of snowball sampling. Rightsholders included full-hire employees, contract labourers (both male and female), former employees, first and second wives of employees, the elderly, children, the ill, disaggregated for Kitete and Uchindile villages and plantation dormitories. Assessors also interviewed workers for job-specific impacts (e.g. fire guard, planters, pruners and nursery workers). Four feedback sessions with rightsholders, health personnel and project staff were held to verify findings. All interviews used semi-structured formats that allowed for digressions (sometimes extensive) onto topics deemed important by rightsholders.
Rights were scored through investigation of over 300 context-, project- and company-related topics, each linked to one of five thematic groupings associated with rights conditions, as shown in Table
1.