Constraints to biofuel feedstock production expansion in Zambia

ABSTRACT

World biofuel production has been increasing to improve energy security and mitigate global warming. Southern Africa’s bioenergy demand could increase with South Africa’s planned fuel blending mandates, triggering increased demand for feedstocks and agricultural land. Ensuring sustained production will require a full understanding of the constraints to production expansion, considering the tradeoffs that may be generated in rural areas, as has been observed for large-scale land acquisitions. We analyse the social and biophysical constraints to biofuel production expansion in Zambia. Previously social constraints have received limited attention even though they may prove more problematic. Results indicate that Zambia is at least moderately suitable for bioenergy investments with biophysically suitable areas largely coinciding with the socially suitable areas. However, existing gaps in compensatory procedures may inhibit large-scale projects’ access to development finance if not aligned with internationally acceptable practices, and generate negative outcomes if safeguards are not in place.

KEYWORDS:

• Biofuels
• social constraints
• biophysical constraints
• Zambia
• Southern Africa

JEL CLASSIFICATION CODES:

• N77
• O13
• Q15
• Q16

1. Introduction

Increasing biofuel blending mandates across the globe have potential to contribute to reduced greenhouse gas emissions, while promoting energy security and generating economic gains. In Southern Africa, South Africa’s planned blending could significantly increase regional demand, given the relative size of its transport sector (Wenberg 2013; REN21 2015). But since meeting bioethanol mandates will require at least 174 000 ha of arable land by 2035¹ (Stone et al. 2015), its land-abundant neighbours will likely remain attractive options as investment destinations, permitting South Africa’s blending rules allow imports. Typically, these countries have been preferred destinations for land acquisitions relating to biofuels and other large-scale agricultural investments (Vermeulen & Cotula 2010; Deininger & Byerlee 2011, 2012).

However, there are tradeoffs that may come with pursuing development of biofuel investments in these countries. And the extent of expansion may be limited or threatened by existing or future constraints. Hence, a full understanding of the local constraints to production expansion is crucial for establishing viable industry – previously, projects failed because of a lack of this understanding (see Locke & Henley 2013). Reaching a scale of production that can sustain an industry requires a set of basic conditions to be in place. On the physical side, there needs to be sufficient arable land, and optimal growing conditions. Production must also be economic, yielding (or promising future) positive financial returns which equal or exceed what producers could hope to get from producing alternative crops. And for reasons explored further below, production also needs to be socially beneficial and not negatively impact the rural communities that engage in – or are affected by – the feedstock production expansion. While in the past the emphasis for making decisions on the siting and scale of feedstocks has focused on the physical-economic feasibility, it is becoming increasingly obvious that there is need to understand the constraints social factors place on aspirations to expand biofuel feedstocks production.

1.1. Why might social constraints impact on biofuel investments?

Both large- and small-scale biofuel feedstock projects have faced substantial challenges since the boom in the early 2000s. Small-scale-led projects faced feedstock-related challenges, with the preferred feedstock proving financially unsustainable. On the other hand, large-scale projects have been more contentious and faced challenges related to the process of land acquisition; several high-profile cases have received attention due to the negative impacts they have had on the livelihoods of neighbouring communities that have lost land and livelihood resources, without seeing the benefits of biofuel projects.

There are several examples of negative impacts from biofuel-related investments, and large-scale land acquisitions in general. Observed impacts on the rural population so far include threatening food insecurity, loss of water rights, and poverty, particularly among vulnerable groups (Locke & Henley 2016). This has been through negative impacts on availability of land, water, and forest resources, given conversion to agriculture use. In other cases, there have been have increased conflicts among rural households and between rural households and the projects (von Braun & Meinzen-Dick 2009; Deininger & Byerlee 2011; Schoneveld et al. 2011; Williams et al. 2012; Azadi et al. 2013; Hunt et al. 2014; Nolte & Subakanya 2016). In other cases, large-scale land acquisitions have only partially used their land (Locke & Henley 2013; Jayne et al. 2014). All this is despite anticipated positive spillovers such as employment, markets, knowledge, and technology from the large-scale agricultural projects (or farms) to their hinterlands.

An important question, then, would be what constraints does this evidence place on the viability of bioenergy expansion projects? Clearly, as the potential negative social impacts of biofuel projects have become recognised, so too have they become important determinants of projects’ economic viability.

First, we consider issues that may possibly impede a project from moving beyond the design phase and into the implementation phase because of the existence of local opposition or outstanding unresolved issues that prevent it from moving forward. These may be hard or soft constraints – financial institutions may withhold funding if projects face unresolved land issues, or projects may also face local opposition from local groups or powerbrokers who refuse to grant a licence. Project viability may be threatened as it has become increasingly clear that the local legitimacy of projects may be questioned further down the line, and pursuing a project in the face of brewing opposition in the hope that this will dissipate over time is often overly optimistic. Many such examples exist across the globe (such as Infinito Gold (Public Citizen 2012) and land conflicts between agribusinesses and locals (Lambin et al. 2013). It is therefore also useful to consider future social constraints where these can be imagined with a reasonable degree of confidence.

Second, we think of social constraints in terms of the impacts that rural transformation projects may have on households. It is expected, and increasingly demanded, that rural development activities – especially those funded by donors and international financial institutions (IFIs) – leave communities in a better position, or at least no worse off, than they were before the intervention.

Social issues can thus constrain production expansion through two channels:

1. Resistance by local landholders to efforts to either encourage them to grow feedstocks (through smallholder/outgrower approaches), or to transfer land to a consolidated land-holding by an investor seeking to control production on a large area of land.

2. Where public finance is involved, funding is increasingly conditional upon project implementers having in place policies and practices to ensure community members provide consent of planned project activities, and where involuntary resettlement is involved they must have in place a resettlement policy that meets international best practices.

Thus, the projected social impacts of any biofuel expansion bear close upfront examination. Against this backdrop, the purpose of this paper is to explore potential severity of social constraints to producing biofuel feedstocks in different areas of rural Zambia under different production models, and identify areas that are likely least constrained by either physical or social factors. It discusses the impact of these constraints for expansion under both smallholder and large-scale models.

The rest of the paper is structured as follows: Section 2 presents the research methods and data sources; Section 3 discusses the study’s key findings; and Section 4 concludes and presents the main policy implications.

2. Methods and data sources

To answer the research questions, we employ a mixed-methods approach consisting of literature review and qualitative and descriptive analysis. While previous initiatives – notably the United Nations Food and Agricultural Organization’s Bioenergy Decision Support Tool – have established screening processes for establishing the viability of proposed biofuel expansion, we are not aware of any attempts that aim to establish the area of land that may be viable based on a combination on physical and social characteristics.

Data for the descriptive analysis is based on nationally representative surveys, namely the Crop Forecast Survey (CFS) and the Rural Agricultural Livelihoods Survey (RALS). The CFS data (2015) covers the 2014/15 season. RALs data covers the 2010/11 and 2012/13 seasons.

The CFS is collected every year by the Zambia Central Statistical Office (CSO) and the Ministry of Agriculture (MoA), with support from the Indaba Agricultural Policy Institute. The CFS collects information on crop production, sales, and land. The dataset is nationally representative of all smallholder and large-scale farmers. It is also representative at the district level, rendering it attractive for identification of socioeconomic constraints at the subnational level.

The RALS is a nationally representative panel dataset collected every three years; it covers 442 standard enumeration areas and 8,840 households. It is the most comprehensive dataset with regards to information on rural livelihoods, crop/livestock production, and sales. Using these data, we identify areas suitable for expansion of biofuels at the subnational level. Specifically, we used data on land-holdings, food availability, and income poverty among rural households. Detailed information on RALS sampling and coverage is given by Chapoto & Zulu-Mbata (2015a).

To assess the willingness and ability of farmers to engage in feedstock production, focus-group discussions were conducted with farmers in areas identified as suitable for feedstock production expansion. A total of 11 districts across four provinces were selected. In each district, two groups of eight farmers were interviewed, with participants drawn from various agricultural camps.

For the analysis of large-scale land acquisitions we looked at different sources of data, including the analysis of government data by Sipangule & Lay (2015)² and the Land Matrix’s database of large-scale land acquisitions. For information on compensation we consulted data from International Finance Corporation client projects in Zambia.

3. Study findings

3.1. Biophysical constraints to biofuel production expansion

3.1.1. Availability of arable land

On paper, Zambia has sufficient arable land per capita to meet biofuel production requirements for both the local and South African market. In 2011, arable land per capita was estimated at 3.1 ha. Although land availability will halve by 2035 due to population growth (Table 1), this will still be higher than most industrialised countries. However, there is a caveat to narrative of availability: inadequate infrastructure limits the extent to which arable land can be utilised and this is one reason why rural people say there is insufficient land in Zambia (Sitko et al. 2015). This is confirmed by the preference expressed by investors seeking agricultural land in Zambia for brownfield sites which are more accessible than greenfield sites (Sipangule & Lay 2015). Nevertheless, efforts to ease this constraint are underway through the government’s farm-block development programme.

Table 1. Estimates of land availability in Zambia up to 2035.

Year	Population	Arable land (ha)	Surface area (ha)	Arable land to person ratio	Land to person ratio
2011	13,100,000	40,000,000	75,261,400	3.1	5.7
2020	17,885,422	40,000,000	75,261,400	2.2	4.2
2025	19,900,000	40,000,000	75,261,400	2.0	3.8
2035	26,923,658	40,000,000	75,261,400	1.5	2.8

There are within-country differences in land availability, but the land constraint is highest in Eastern and Southern Provinces, where 73 per cent of the rural smallholders say there is insufficient land (Chapoto & Zulu-Mbata 2015b).

3.1.2. Climate, soil and water

Climatic and soil conditions may limit what feedstocks can be grown in an area. Zambia has three main AEZs, which are classified based on soil type, temperature, and rainfall (Figure 1). AEZ I receives less than 800 mm of rain per annum, AEZs IIa and IIb receive 800–1,000 mm per annum, and AEZ III receives 1,000–1,500 mm of rain per annum (Chapoto & Zulu-Mbata 2015b) From our analysis, we find most of Zambia receives sufficient rainwater (800–1,200 mm per annum) that can be used to produce feedstocks under rain-fed conditions. This is mostly in agroecological zones (AEZs) II and III, the majority of which is of high agroecological potential (except for AEZ IIb) and generally has low population densities and growth, even by regional standards (Figures 1 and 2; Sitko et al. 2015).

Figure 1. Zambia’s Agro-ecological Zones. Source: Chapoto & Zulu-Mbata (2015b).

Available data estimates renewable annual surface water potential at 100 km². Most surface water is located in Luapula, Northern, and Southern Provinces, accounting for 93.1 per cent of the 12 621 km² available surface water (Figure 2). However, Southern Province is likely to be surface water-constrained, given sugarcane production activities by Zambia Sugar Plc and Kafue Sugar.

Figure 2. Distribution of Surface Water Resources by Province. Source: Author’s calculations based on data from Nyambe & Feiberg (2009).

We find that most areas with adequate rain and surface water also have large farm sizes per capita (Figure 3). Renewable groundwater potential is estimated at 49.6 km³, but little is known about the spatial distribution of groundwater resources. A complete understanding is crucial, especially in the southern-most parts of the country, where irrigation using groundwater would significantly contribute to feedstock production expansion under poor rainfall and high rainfall variability. Moreover, locating projects in this area would cut down transport costs to South Africa. Irrigation potential is estimated at 2.75 million hectares, of which only 156 000 ha of land is presently under irrigation on commercial farms of irrigated cash crops (ZDA 2014; MoA 2016); irrigated crop production by smallholders is almost inexistent. However, at the household level, the 2015 RALS shows that most rural households in Zambia are within 4 km of piped or borehole water, with the lowest distances observed in Eastern Province. However, extremes are observed for Serenje District in Central Province at 8.1 km and Mpika in Muchinga Province at 13 km.

Figure 3. Distribution of Mean Land-holding Size by Agro-ecological zones. Author’s calculations based on RALS 2015.

3.2. Constraints due to features of tenure systems

The majority of land in Zambia is under customary tenure (86.9 per cent of smallholders hold land under customary tenure), and conversion to statutory land is often complex and expensive for smallholders (Sitko et al. 2015).

Findings from focus-group discussions in four provinces of interest for bioenergy investments indicate that there are clear differences in the ease of access to land across space and gender. On inheritance practices, focus group discussions generally suggested that land is passed on within families, especially to male children. Women have use rights to this land, or use land belonging to their husbands when they get married. In previous times, it was more difficult for females to acquire land, but more recently traditional leaders are increasingly allocating land to females in their chiefdoms. Among women, access to land other than that within the family is easier for widows and divorcees. In some cases, traditionally, it is unacceptable for women to separately own land for as long as they are married. For example, in one district, one farmer said: ‘I can let her go look for land if we have children.’ Another said: ‘For the sake of marriage security, I would rather she didn’t look for land because she may be too independent and run away.’

However, drivers of access to land are changing and for the provinces that are likely to be destinations for biofuel investments, we find that finance is a key determinant of whether one is able to access land or not (through rental markets or sales), irrespective of gender or whether they are local or non-local. This offers an explanation of why there has been elite land capture in Zambia. The ease of access that is driven by finance appears more important in densely populated areas near the metropolitan cities.

3.3. Constraints to smallholder-led production expansion

From a social perspective, three indicators are used to determine suitability for a smallholder-led biofuel expansion programme: food insecurity, plot sizes and income poverty. The justifications for selecting these is as follows. First, in areas where people are food insecure, expanding feedstock production may worsen their food insecurity via increases in food prices.³ It may also worsen their nutrition status, given the likelihood that they would also have reduced crop diversity as they expand feedstock production. Small median plot sizes likely indicate a general lack of available land for large-scale feedstock production, and limited ability of smallholders to engage in outgrower schemes. As such, areas with small median land sizes may not be appropriate for expanding feedstock production. A high share of poor households indicates inability to engage in production expansion due to a general lack of capital or equipment, and also indicates a lack of capacity among households to survive external shocks.

We compute land, number of months without food, and poverty terciles across regions within Zambia. Land-constrained regions fall within the bottom tercile; similarly, areas least constrained by poverty fall within the bottom tercile of the share of poor households. Income poverty is used as it is the next best alternative to consumption-based poverty estimates. For these data, we use the US$1.25 poverty line based on 2005 purchasing power parity. Food-insecure areas fall within the top tercile of months without food. Areas satisfying all three constraints are classified as unconstrained; those satisfying two constraints as moderately constrained; and those satisfying one condition or fewer are constrained. Summary statistics for each of these constraints are provided in Table 2.

Table 2. Summary statistics for social constraint variables.

Social constraint	Constraint terciles
	1	2	3
Share of poor households (percentage)	36	51	65
Median farm size (hectares)	0.25	0.45	1
Number of months without food	3	3.6	5

Source: Authors, based on data from CSO (2010, 2013); Sitko et al. (2015); ZDA (2014).

Based on this classification we find that Western Province is largely unsuitable for feedstock expansion; it has a large number of districts that are socially constrained. This is followed by Muchinga and Eastern Provinces. In Southern Province, the southern-most areas are unsuitable – an expected result given the marginal agroclimatic conditions for agricultural production. There are some parts of Eastern Province that are unsuitable, mostly close to Muchinga Province (Figure 4).

Figure 4. Social constraints to biofuel investments: (a) suitability based on food security; (b) suitability based on land-holding sizes; (c) distribution of the share of poor households; (d) Overall suitability based on all social constraints. Source: (a) RALS 2015; (b) CSO/MAL CFS 2014–15; (c) RALS 2015; (d) compiled based on (a)–(c).

Out-migration patterns indicate the general availability for labour for agricultural production in an area and thus a key constraint to expanding production (Lambin et al. 2013). As such, areas with high rural–urban migration may indicate a general lack of labour for agricultural activities both presently and in the future. Rural to urban migration is highest in Luapula and Northern Provinces (44 per cent and 34 per cent, respectively) (Nyambe & Feilberg 2009). Most districts in these areas are also moderately suitable (Figure 4). The highly urbanised Provinces (Lusaka and Copperbelt) experience very limited rural–urban migrations (<10 per cent). In the remaining provinces, rural to urban migration is estimated as 18–30 per cent. However, while there is rural to urban migration in some areas, RALS 2015 shows that, on average, most migration in rural Zambia occurs within rural areas, perhaps because urban areas do not have the employment opportunities that the rural people seek – implying that availability of unskilled labour may not be a problem.

Results from focus-group discussions indicate a general willingness among farmers to participate in biofuel projects; however, most did indicate that there are limitations to how much production expansion can take place among smallholders on their own. Doing so would require increasing their access to inputs, and mechanisation. Presently, household labour is the major source of tillage power, with animal draught and mechanical power used by very few households (Figure 5).

Figure 5. Distribution of Households using Technology. Source: RALS 2015.

There are limits to how much one can cultivate using manual labour, which is the major power source. Similarly, low intensification levels limit productivity and production levels. However, farmers in areas of interest (Luapula, Southern, Central, Northern, and Copperbelt Provinces) did highlight the need for safeguards to be in place, especially given the experiences with past firms, where contracted Jatropha was not bought, leaving farmers without a market. Further improvements to the contractual relations between farmers and firms will be critical to the industry’s success.

3.4. Constraints to a large-scale-led biofuels industry

Zambia is one of the countries that has attracted most interest from investors seeking large tracts of land to develop agriculture projects (Deininger & Byerlee 2011; Oakland Institute 2011).

Using several sources of data,⁴ Sipangule & Lay (2015) detail the number, size, and location of large-scale land investments (LSLIs) across Zambia’s districts. They enumerate 95 foreign LSLIs in 26 districts, covering around 562 312 ha of land. All of these districts are in the Central, Copperbelt, Northern, North-western, Luapula, and Southern Provinces. Their analysis suggests that most investments are clustered along the railway line and major roads, and that investors prefer to take on existing farms rather than invest in greenfield sites.

Recent analysis by the Land Matrix (Land Matrix 2016) covers a smaller number of deals (34)⁵ comprising 390 074 ha. Looking at the intended use of the land, they find that most projects aim to produce food crops (42 per cent), followed by multiple uses, including food and livestock production. Although only making up 15 per cent of deals, investments that target production of biofuel feedstocks cover 53 per cent of the land under contract (Figure 6).

Figure 6. Intended use of land under concluded land deals in Zambia. Source: Authors’ calculations based on Land Matrix (2016).

Unlike in some other countries where the number of deals fell following an initial wave of acquisitions in 2008–10, in Zambia the pace at which deals have been signed has quickened since 2011, with significant deals covering a large area of land (c.200 000 ha) signed in 2012.

A large share of intended deals have progressed to an operational stage. Analysis of their data by the Land Matrix suggests 65 per cent of those deals that have been concluded have progressed to an operational phase; similarly, Sipangule & Lay (2015) find that of the 95 cases they examine, 75 have progressed to an operational phase. However, as is the case elsewhere, the majority of these investments are not farming the full area they have leased: Land Matrix analysis suggests that, on average, ventures farm less than 5 per cent of the area they have under contract. Biofuel feedstock projects, in particular, have faced challenges reaching the production stage: only two of the seven biofuel feedstock projects have reached and remain in production, and it is unclear how much of this land is actually cultivated.

Part of the reason for Zambia’s attractiveness is its legal framework, which provides relatively easy access to land for investors. The 1995 Land Act allows investors who meet a set of conditions⁶ to acquire land. Importantly, the Land Act provides a mechanism for the conversion of customary land to leasehold state land, which removes land from the jurisdiction of chiefs. It also allows traditional owners and existing leaseholders to identify outside investors and recommend them to the Commissioner in the Ministry of Lands (Sipangule & Lay 2015). This mechanism allowing the conversion of customary land to statutory land has been frequently used by Zambia’s growing middle class who, together with central and local government, have led a process of land conversion that has reduced the area of customary land from as much as 94 per cent in the middle of the twentieth century to 54 per cent in 2015 (Sitko et al. 2015).

In addition, in an attempt to stimulate the agricultural sector, the government played an active role in promoting Zambia as a destination for foreign investment in agriculture. The cornerstone of this drive was the establishment in 2008 of a farm-block development programme. Through this programme, the government identified large tracts of land in nine provinces which it has plans to survey, convert from customary to state land, and provided titles for. As of 2014, one of the farm-blocks (Nansanga) was most advanced; as well as providing electricity and dams, the government had conducted a cadastral survey and surveyed and titled the land (GRZ 2014). To attract investment in these locations, the government has made its own investments in key infrastructure, including roads, irrigation, and power (Oakland Institute 2011).

3.4.1. The process of land acquisition for investment

In spite of the shift of land from customary to leasehold, a sizeable proportion of land remains under the custody of chiefs. For large projects, the Lands Acquisition Act of 1996 gives the president the right to acquire any piece of land in the national interest; however, to acquire customary land, three authorities – the district council chair, executive secretary, and the chief – must provide approval for conversion of customary land to leasehold title. The commissioner of lands is also meant to provide consent unless the process causes injustice, or is contrary to national interest or policy. The 1999 Lands Act is not clear regarding what rights the chiefs retains over land following conversion, what other customary rights to use land (other than occupation, which is prohibited) remain, and whether customary laws continue to apply, but Tagliarino (2014) notes that in practice customary claims to land are extinguished, and the chief’s ability to influence actions by the investor are diminished as only the commissioner of lands is party to lease agreements, and leases are only subject to statute and regulations passed by the Ministry of Lands. Chiefs are not given any responsibilities to ensure terms of the contract are enforced.

3.4.2. Legal issues surrounding the processes of acquisition, consultation, resettlement, and compensation

Global experience points to a close association between large-scale land acquisitions and social risks, especially when households are displaced and resettled. In Zambia, there is a mix of evidence of displacement resulting from land investment. Studies from around 2010 found that households were unable to access food as easily after displacement compared to before displacement. Milimo et al. (2011), found that the households forced to relocate following the transfer of church land to an investor suffered from increased hunger as they were unable to farm the new areas as successfully as they could their former land. Similarly, a study by the Zambian Land Alliance of a displaced community, cited by the Oakland Institute (2011), found that following relocation households were only able to produce 25–50 per cent of the crop volume they previously could.

This report also documented cases of households becoming physically disconnected from public services, including schools, and losing access to both food and fuel from forests. Nolte’s (2014) qualitative study found anecdotal evidence of displacement occurring, and notes that only in those cases where forced expropriation occurred is compensation mandatory; in other cases, the amount of compensation that affected households receive depends on the generosity of investors, the bargaining power of the community, or discretionary interventions by local or national government officials. As a result, it is not uncommon for households to receive no or low compensation for lost land. In addition to questions regarding compensation for foregone land, affected households complained that they did not receive compensation for destroyed trees and crops. However, interviews also revealed that a substantial number of households were positive about the impact of LSLIs, especially the employment opportunities they generated and, in a small number of cases, providing infrastructure, access to farm equipment and information on farming techniques.

Key processes to ensure that social issues are minimised are those included in processes of consultation, resettlement, and compensation. There are different approaches to mitigating these risks. In Zambia, Environmental and Social Impact Assessments are the legal mechanism used to determine potential adverse impacts and selection of suitable mitigation mechanisms before resettling people affected by projects or disasters.

However, there are weaknesses associated with each of these processes in Zambia. The 1995 Lands Act contains only weak provisions on consultation: it requires the president to take into consideration customary law and consult those whose interests may be affected. It provides no guidance on how consultation should take place, or whether (and how) affected persons are to be compensated in the event of physical or economic displacement (Tagliarino 2014). Guidance from the ZDA presented above suggests that the lowest level of consultation is with chiefs and village headmen, so consultation with villagers is not always formally required.

There does not appear to be a clear division of responsibilities for leading on issues related to displacement and resettlement between different government departments. Chu et al. (2015) note that while the Zambia Environmental Management Agency is responsible for auditing Environmental Impact Assessments (EIAs) and Resettlement Action Plans, the process of resettlement is overseen by the Department of Resettlement (DoR) and the Disaster Management and Mitigation Unit (DMMU), which sit under the Office of the Vice President. Experiences suggest there is inadequate communication between these offices (Chu et al. 2015). During our research, participants in focus-group discussion suggested compensation was organised through the Ministry of Local Government and Housing, but the valuation exercise was carried out by the MoA.

In addition, while all projects requiring resettlement need to submit an EIA, producing a Resettlement Action Plan does not appear to be a legal requirement for either government agencies or companies whose internal policies do not mandate one. Instead, government agencies often rely on the DMMU and DoR to carry out these tasks; neither department appears to have adequate capacity to monitor and enforce the regulation as they stand.

4. Conclusions and policy implications

Increasing biofuel mandates in Southern Africa offer promise for improving energy security and mitigating global warming, while generating benefits for locals. South Africa’s planned blending creates a large market for produced biofuels in Southern Africa. Given that it is land-constrained, production will have to be from land-abundant countries, such as Zambia. In addition, Zambia’s biophysical suitability, we also analyse social constraints to a small- and large-scale-led biofuels industry expansion. Inclusion of social constraints underscores the importance these have for ensuring a sustainable industry that will not face local opposition or generate negative social costs, and ensuring projects have access to international finance.

We analysed social constraints to small-scale-led production expansion. Using the extent of food insecurity, poverty, and the size of median land-holdings as indicators for social suitability for a smallholder-led biofuel production expansion programme, we find that the biophysically suitable areas largely coincide with the socially suitable areas. Except for Western Province, and the southern-most parts of Southern Province, the rest of Zambia is at least moderately suitable for feedstock production expansion from both a social and physical point of view.

However, while Zambia is biophysically suitable for biofuel investments, social constraints and costs may limit the extent of a large-scale led production expansion. Overall, there is a considerable risk that rural households may be left worse off if large-scale-led investments proceed, as evidenced by studies involving previous land acquisitions. This emanates from inadequacies in the consultation, displacement, and compensation procedures and practices related to land acquisitions for large-scale investments. Without a comprehensive legal framework nor the public institutions with a clear mandate to guide the process of resettlement and compensation, key decisions regarding these issues fall to the individual.

Our findings suggest that a review of existing procedures and its implementation is thus crucial. Otherwise, any investments must be in areas with low population densities in order that risks are minimised, such as on those areas of state land which are not subject to conflict. However, findings from earlier studies including on farm-block sites suggest that these are not necessarily always free from encumbrances and undertaking proper consultation on this land and following best practices for resettlement and compensation is necessary.

Acknowledgements

We are grateful to the United Nations University World Institute for Development Economics Research for funding the publication of this paper. A previous version of this paper containing the same figures and tables was published as a UNU-WIDER working paper under the same title: Constraints to biofuel feedstock production expansion in Zambia. UNU-WIDER Working paper No: 62. Helsinki, Finland Link: https://www.wider.unu.edu/sites/default/files/wp2017-62.pdf

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 While South Africa has around 300,000 hectares of land under sugarcane (which yields most feedstock per unit area), we assume it would not be financially feasible to divert much land from this pool to producing ethanol, and since there is limited other suitable land for biofuels in South Africa, it would be more economically attractive to grow feedstocks in neighbouring countries.

2 This includes the Zambian Development Agency’s list of expressions of interest in land acquisition.

3 Although this channel is commonly identified in conceptual frameworks, the authors are not aware of evidence to support the assertion that food crop displacement at the local level raises local food prices. The one paper that has looked into this in Zambia (Sipangule & Lay 2015) found no effect.

4 This includes data from the Zambian Development Agency and the Land Matrix.

5 The Land Matrix limits its analysis to deals for agricultural purposes (excluding forestry and mining) that entail a transfer of rights to use, control, or ownership of land, cover an area of at least 200 ha, and were initiated after 2000. Data were valid as of 1 February 2016.

6 These conditions include that the company must adhere to conditions of the Investment Act, including that 75 per cent of shareholders are Zambian and are registered in Zambia.