A closer look at displacement for hydropower projects
Relatively abundant rainfall in Uganda not only provides the basis for hydropower production but also widespread rainfed subsistence agriculture and a high population density with small land plots in most of the project areas. Consequently, displacement of people is one of the main issues that projects need to carefully manage to minimise impacts on local people and to reduce project risks associated with sensitive issues regarding land, compensation and livelihood restoration. Some experiences with handling of displacement in the small hydropower projects are summarised below but is limited to physical and economic displacement and not associated issues such as potential food insecurity, landlessness and marginalisation. As will be seen below, there is great variation in the degree to which developers have been able to avoid displacement.
Highly variable extent of displacement
There are two forms of displacement. Physical displacement refers to relocation or loss of shelter. Economic displacement refers to loss of assets or access to assets that leads to loss of income sources or other means of livelihood. Project-related land acquisition as well as restrictions on land use can cause displacement.
Based on data available at the time of GET FiT project approvals, the small hydropower projects were estimated to cause physical displacement of about 150 households and economic displacement of about 1,400 households. It is important to note that most of the economically displaced households only lost a small proportion of their land.
GET FiT has observed great variation in the degree to which projects have managed to avoid and minimise land take and displacement. There is a trend of more displacement when land take grows (see Figure 3 below). There is an exception to these trends as some projects have very limited or no displacement despite a relatively large land take. The latter group of projects are located in drier areas without significant agricultural production, with very low population density and therefore a limited incentive to minimise land take. About 87% of the projects (13 out of 15) have a land take below 1 hectare per GWh of estimated annual power production. Of the two projects with a larger land take than 1 ha/GWh, one has a reservoir with considerable inundation and the other is in a dry area without agriculture and no strong incentive to minimise land take.
A group of four projects (27%) cause very limited displacement (physical and economic), less than 0.5 households per GWh of estimated annual power production. The rest of the projects impact more than 2.3 households per GWh. The main difference between these two groups of projects is level of rainfall in the project areas and consequently extent of agriculture and population density. Within the second group of projects (those in areas with considerable rainfed agriculture), there is large variation in impact levels. Here, the ability of developers to adjust project design to avoid displacement is an important factor influencing levels of displacement and associated risks.
Figure 3 - Land take (in hectare) per GWh (estimated annual power production) plotted against total number of displaced households (physical and economic displacement) per GWh for 15 small hydropower projects approved for GET FiT support.3
If we split total number of displaced households into physical and economic displacement, Figure 4 shows a general trend of physical displacement increasing with the extent of economic displacement. This figure also shows two main groups of projects in terms of physical displacement. Nine projects (60 %) have avoided all or virtually all physical displacement, four of which have also fully or nearly fully avoided economic displacement. These projects are in areas generally not suited for agriculture and with very few people, typically further downstream in catchments where climatic conditions are drier. The other five of the nine projects in this group have managed fully or largely to avoid physical displacement even if located in areas with rainfed agriculture and higher population density. On the other side, a group of four projects causes the highest levels of physical displacement, all have recorded more than 0.74 households physically displaced per GWh of estimated annual power production.
Figure 4: Number of physically displaced households per GWh (estimated annual power production) plotted against number of economically displaced households per GWh for 15 small hydropower projects approved for GET FiT support.
Some GET FiT experiences from engagement with developers
The above shows the need for GET FiT’s to focus on management of displacement. Project developers were expected show that they had avoided displacement as far as possible, and when avoidance was not possible, minimise displacement by exploring alternative project designs. The degree of success in avoidance varied considerably, partly due to the developers’ planning capacities and partly due to the topographical and climatic conditions at project sites.
For those displaced, the livelihoods and standards of living should as a minimum be restored, preferably improved. Appropriate compensation and livelihood restoration plans were therefore essential and a focal issue for GET FiT. Developers overwhelmingly relied on cash compensation, which was the preference over land-for-land and house-for-house compensation among communities, despite the risks in terms of long-term livelihood security when cash compensation is used. Cash is also a simple form of compensation for the developers, at least in the short-term. GET FiT requested developers to do more to raise awareness and incentivise compensation in-kind rather than cash, particularly for vulnerable persons, and to monitor whether livelihoods are restored. The lack of developers’ experience from long-term engagement with project-affected persons, particularly vulnerable people, contributed to the bias towards cash compensation.
Categorising project-affected people and clearly distinguishing between those affected only marginally on the one side (most PAPs) and those who are substantially affected and left in a vulnerable position on the other (few PAPs), was another challenge. For the first group, cash compensation may not jeopardise living standards as most of the livelihood base remain unaffected. The second group is potentially much more vulnerable as they may have lost a large proportion of their land, remain with land of poor quality or their ability to sustain themselves has otherwise been reduced. GET FiT requested most developers to revise resettlement action plans and livelihood restoration plans to better differentiate livelihood restoration approaches according to PAPs’ vulnerability.
Some developers make commendable efforts to work with local civil society organisations (CSOs) that have experience in community work as part of livelihood restoration. Experiences are mixed as abilities of the CSOs have proved variable, underlining the importance of good selection processes for partners in livelihood restoration.
Implementation of resettlement and livelihood restoration plans is ongoing and the degree to which projects manage to restore, or improve, livelihoods in a sustainable manner will be reported on later.
3 Data are taken from studies done prior to GET FiT approval. For some projects, the extent of land take and displacement have increased during detailed design and construction. These increases will be reported on later.