Villagization and Plantation Development in the Lower Omo: The Planners’ View

In 2012, the South Omo Zone Agriculture Bureau produced a “Villagization Plan” which described how, in the course of the coming year, the majority of inhabitants of the zone—some forty-five thousand people—would be moved into new villages (FDRE 2012). According to official documents, moving these people, referred to collectively as pastoralists, into new villages and introducing them to new farming methods would improve their lives. In the Plan the benefits of villagization are described as follows: (1) growing new crops, such as sugarcane, on irrigated fields would improve food security; (2) providing people with schooling and medical services would improve their health; (3) settling in planned villages would reduce conflicts that result from “mobility in search of water and pasture” (FDRE 2012:5).

As laid out in the Plan, the campaign would begin with community meetings at which the benefits of villagization would be explained, whereupon people would be assigned plots on which to build their new houses. At the same time as the sugar-cane plantations, processing factories, and new towns and villages were being built on land that was owned by the local communities, people would be issued with formal titles to both residential land and to farmland: “In the case of land that relies on rain [for cultivation] 2 hectares would be apportioned to each settler pastoralist, whereas those settling on irrigable land would be apportioned 0.5 hectares each” (ibid., p.8). They would also be provided with food aid “for about eight months, until the crop they plant in the first year reach for harvest [sic] and they could feed themselves” (p.18).

The expansion of artificial irrigation in the lower Omo Valley was made possible by the construction of the Gibe III dam upstream, which began in 2008. In conjunction with the dam, a system of canals was built to distribute water to the Omo-Kuraz Sugar Plantation, within which selected plots of land were set aside for use by “settler pastoralists.” The Omo-Kuraz Plantation is part of the Kuraz Sugar Development Project, a state-owned enterprise projected to cover 175,000 hectares (Ethiopian Sugar Corporation 2015). At full capacity, it was predicted, Kuraz would be responsible for fully a quarter of the country’s sugar and ethanol production, more than any other plantation in the country (Ethiopian Sugar Corporation 2014). In line with this vision, the Villagization Plan describes the overarching goal of the villagization project as to “change the economic and social condition of the pastoralists and make them out-growers for the sugar factory” (FDRE 2012:4).

Investigating the Bodi Experience of Villagization

It is important to note that although the people who were targeted by the South Omo villagization program are frequently described as “pastoralists” by government officials (e.g., Meles 2011; Walta 2012), their diets were traditionally based mainly on farming of sorghum and maize, with dairy products providing an important supplementary component.^{Footnote 2} In the more southerly parts of the lower Omo valley, there is ample floodplain on which to cultivate these crops; further north, people rely on a triad of livelihood strategies: herding, flood-retreat farming, and rain-fed agriculture (Turton Reference Turton1985). This misconception about the livelihoods of the target population—the idea that, being pastoralists, they are unacquainted with agriculture—is widespread, and has had significant repercussions for the ways development is imagined by planners.

Our study focused on the Me’en living on the eastern side of the Omo River, a segment of the larger Me’en ethno-linguistic group known to the Ethiopian authorities and outsiders as the Bodi, who on account of their location close to the Kuraz Sugar Development Project were the first people in the lower Omo to be targeted for villagization.^{Footnote 3} Numerically among the smaller ethno-linguistic groups of the region, the Bodi were estimated in the most recent census to number approximately 8,000 people (CSA 2008). According to the Villagization Plan, between 2012 and 2013, 1,430 households were to be resettled in three “new villages” in Bodi territory. (Assuming an average household size of five, this would represent more than seven thousand people.) Infrastructure for the villagization sites was first installed in 2012—schools, clinics, veterinary centers, and mill houses.

It was not until planning was well under way that the government arranged public meetings to announce the villagization plan, and little effort was made to accommodate local ideas regarding the layout of the sites or the kinds of livelihoods that might be possible there. As Yidneckachew, who studied the public consultation process, observed, “the process was a nominal participation of pastoralists [sic]. . . . Discussions at the public forums were top-down. . . . [The] facilitators determined and controlled the agenda, the alternative solutions, and the process of the consultation” (2015:296, 300). The forums were orchestrated so as to permit the airing of “personal concerns about the project,” but did not afford the opportunity for any major challenge to the premises of the scheme. Rather than being open to all comers, participants in the forum were “called” up by the organizers based on their “proximity to the project command area.” According to a member of the Regional Steering Committee interviewed by Yidneckachew, “participants were not to represent any [particular] group or the pastoralists’ concern” (2015:296).

Settlers began to arrive in May of 2012, and official data suggest that by September of that year a few dozen households were living in each of the villages. The principal “benefit” they received was an allocation of 30kg of grain per household per month as food aid. Land for irrigated farming was unavailable for approximately the first six months, but by early 2013, agricultural extension agents (known locally as development agents) had begun directing the settlers to grow maize using an irrigation system provided by the Sugar Corporation. The promise of the planners was that by using new seed, fertilizer, and artificial irrigation, two or three harvests would be possible each year, with additional income available from selling cane to the Sugar Corporation for processing in the new factories.^{Footnote 4} In practice, delays in the installation of infrastructure—both irrigation systems and the sugar factories—meant that these harvests were elusive. This was only one of the ways in which the experience of villagization failed to meet official expectations. Although people were at first given plots of 0.5 ha to cultivate maize for subsistence on irrigated plots, later the size of their plots was reduced to 0.25 ha. Once the processing factory was completed, the planners suggested, people would be entitled to an extra 0.75 ha of land for sugar-cane cultivation, in addition to the 0.25 ha they were already cultivating.^{Footnote 5}

We carried out our research both in villagization sites near Hana town (the administrative center of Salamago woreda) and in a comparison community in the division of Bodi territory called Gura.^{Footnote 6} In the following section, we describe our research methods and some of the challenges faced in the process of carrying out the research.

Comparing Ethnographic and Survey Data on Food Security

During the period of our research, Hana town was undergoing a transformation from a mere police post into an increasingly bustling town, and camps for full-time plantation workers, who numbered in the thousands, were springing up alongside the plantation. Both through informal labor migration and through planned resettlement, Salamago woreda as a whole was becoming increasingly ethnically diverse. Resettlement from Konso, a highland area of southern Ethiopia, had begun in the mid-2000s; and from 2012 increasing numbers of spontaneous migrants were also arriving from other parts of Ethiopia’s Southern region in search of work on the new plantations.^{Footnote 10} These newly-arrived migrants were especially visible in Hana town: in the mornings they lined the main street to await transportation by lorry to the plantation sites, and in the evenings they filled the restaurants and bars—businesses that themselves had been established only within the previous year. In the space of a decade, not only the Bodi but also the neighboring Mursi and other indigenous people of Salamago woreda had become minorities in their own lands.

Our survey provides a snapshot of food access conditions in the villagization sites and in Gura in August 2013 (Table 1). The food insecurity scale ranges from 0 to 27, with higher numbers representing more experiences of food insecurity (hunger, skipped meals, reduced dietary diversity, and worry over food) within the past month.

Table 1. Food Access in Villagization Sites Near the Kuraz Sugar Plantation and in the Cattle Camps of Gura, in August 2013

* Mean values are significantly different between the two sites (p <.05)

Our initial hypothesis was that food security would be lower in villagization sites: that is to say that compared to Bodi communities not subject to villagization, households in the villagization sites would experience greater food insecurity. This prediction was based on a reading of the literature on the social impacts of resettlement, which demonstrates that relocation has more often deepened poverty than ameliorated it (Cernea Reference Cernea, Cernea and McDowell2000; Scudder Reference Scudder and Tortajada2012).^{Footnote 11}

The household survey data, however, do not support this hypothesis. Food insecurity was high in both of the study areas—85 percent of households in the villagization sites and 97 percent of households in the Gura cattle camps reported experiences in the preceding month that reflect moderate or severe food insecurity according to conventional criteria (Coates et al. Reference Coates, Swindale and Bilinsky2007)—but responses to the survey suggest that people in Gura experienced a degree of food insecurity significantly higher than those in the villagization sites.

In interpreting these results, it is important to remember that at the time of the research, the government was distributing food aid only to people in the new villages. Our ethnographic data also allow us to check the validity of the food security indicator against other aspects of context. How did the communities differ in terms of food access, beyond food aid? How did the people feel about their situation?

Bodi Views of Food Security and Well-Being under Villagization

As this was the first attempt at smallholder irrigated cultivation within the Kuraz Sugar Development Project, there was significant pressure on development agents to make it into a showcase for the success of the whole project. Once irrigated plots were made available in early 2013, development agents monitored the participants’ fields closely and held the Bodi accountable for any failure to bring the maize to maturity. This monitoring and nagging was resented by the Bodi, because crucial aspects of the farming—the size and location of plots, the type of seed, and the time of planting—were beyond their control. Settlers perceived themselves as occupying a lowly position in a command hierarchy, which grated with Bodi ideals of autonomy. As one settler told us:

The supply of water from the new irrigation system was also controlled by the Sugar Corporation. Some settlers found themselves in a position of advantage—those whose fields were closest to the canal received the water first when the canal gate was open, and could then block it so that the water entered their fields only. But crop yields from the irrigated fields were generally disappointing. Even when planting occurred during the rainy season and water was available from the canal—and the crops therefore received both rainwater and artificial irrigation—the resulting harvest was not enough to feed the settlers, let alone to produce a surplus for sale.

In addition to simple availability of food, a sense of well-being for the Bodi also includes the opportunity to consume—and share—food and drink in familiar ways. The physical arrangement of the new villages presented a key challenge here. The villagization sites, once they were established, bore little resemblance either to the clusters of houses found in cultivation sites (olman) or to the cattle camps (ori) that we described earlier. Although the new villages were supposed to gather several hundred households into more nucleated settlements, within these sites each household was positioned quite distant from the next—each on its own 0.5 hectare grid square—and in the allocation of plots to settlers, the local administration gave people little choice over either the location of their houses or whom they had as neighbors. These features of the new villages implied a very different set of routines and relationships than the ones the Bodi were accustomed to. When asked whether he shared his morning coffee with neighbors, a resident of a new village answered:

Bodily discomfort and social alienation emerged as key idioms for describing the disruptions and dislocations associated with villagization and plantation work. When one of us (LB) visited an irrigated field where women were weeding and asked them how they felt about the new conditions, at first some young women answered shyly that the government was arranging things well. But then an older woman spoke up: “Here, we are forced to live like serategna [Amharic: day-laborers]!” In their own fields, she went on to say, they had shade to rest in, and could make coffee; they enjoyed the comfort of their homes, and they mixed farm work with breaks drinking or socializing. But in the irrigated fields set aside for the settlers, which were approximately a half hour’s walk away from new villages, all the shade trees had been cut down. In order to get a little shelter from the sun, they had to spread clothes over the immature maize plants and then crowd into what little shade it afforded, to drink the beer of the working-party. From being a place to live and work, the fields had now become mainly a place of (hard) work.

The heat, the sense of powerlessness, the meager harvests, and the harassment of the development agents made life hard for people in the villagization sites. Food aid provided by the government was the principal attraction. The following excerpt from an interview with a man from a cattle camp a few kilometers from Hana town provides insight into the motivations of many of the first people who registered as residents of the new villages.

This strategy of building a makeshift house at the villagization site in order to obtain rations seems to have been tried by many other families. Rather than representing a vote of confidence in the development plan, the principal reason for participating in the villagization campaign was in many cases to obtain food aid. Many of those who opted to settle before the irrigated fields were made available were struggling to support themselves—some, like the man quoted above, had few cattle, and could not therefore afford to sell any to buy grain; some had few able-bodied children, or elderly husbands. Others were relatively well-off, but wanted to take advantage of the food being distributed in the villagization sites.^{Footnote 13} Similar to the case of villagization in central Ethiopia studied by Helen Pankhurst (Reference Pankhurst1992), this initiative appears to have attracted a variety of people for a variety of reasons—some of them among the poorest and most desperate, and others who were more prosperous but saw an opportunity to profit from the situation.

When we compare the experiences of the Bodi with other documented cases of resettlement, it becomes apparent that certain crucial aspects are due not so much to villagization as to the political and ecological changes that accompanied it: the fact that it was carried out at the same time as (1) a massive influx of labor migrants, which added to the almost decade-long presence of people deliberately resettled from Konso and (2) the ecological reengineering of the landscape to make it fit for plantation agriculture. In these respects the situation in the lower Omo closely resembles developments in the Awash Valley in the 1960s and ’70s (Kloos 1982), and the combination of villagization and plantation development that has been carried out more recently in Ethiopia’s western province of Gambella (Fana Reference Fana2016a, Reference Fana2016b). In all three of these cases, a variety of forces undermined the livelihoods and food security of indigenous people at the same time as some of them were attempting to make a living in the resettlement sites. In the next section, we place our ethnographic findings in relationship to the work of other scholars of forced displacement and well-being, and we briefly consider the kinds of resistance the Bodi mounted to the changes that were forced upon them. Finally, we reflect on the implications of different research methods and interpretive frameworks for shedding light on these problems.

Do Our Bodies Know Their Ways?

We have highlighted in this study certain misconceptions on which villagization in the lower Omo has been based—(1) that the target population for resettlement were pastoralists unacquainted with agriculture, when in fact they were already reliant on farming for the majority of their diets; (2) that these individuals would enthusiastically embrace the opportunity to “settle” as soon as they were made aware of it, whereas in practice they viewed it skeptically; (3) that the resettlement sites would be more nucleated than places like Gura, whereas in important respects they were more spread out; and (4) that food security and well-being would inevitably be improved by these changes. We have shown that villagization was actually experienced by the Bodi settlers as undermining food security and diminishing well-being. It was experienced as heat, as disruption to routines, as a loss of control over livelihoods, as being treated like children rather than adults, and as abstracting work from ordinary sociality (see Carrier Reference Carrier2001). The question asked by one of our informants, “Do our bodies know their ways?” is emblematic of the profound disruption that many Bodi felt in the context of the upheaval that was forced upon them.

Unpacking the words our informant used helps clarify some implications that are otherwise hazy. A literal translation of the phrase would be “Does the body/self know? It doesn’t.” Rɛɛ, which we have translated here as “body,” can also be understood as “self,” a double meaning that challenges the mind / body dichotomy in Western philosophy.^{Footnote 14} Students of the Mursi, culturally one of the Bodi’s closest neighbors, have analyzed the Mursi word for “body” and its relationship to local notions of selfhood. Eczet explains that the Mursi word ree denotes a general state of being, rather than the physical body as such—hence the phrases “the body is good” or “the body is bad” refer to personal well-being or ill-being.^{Footnote 15} Furthermore, the “body” or “self” is intensified or strengthened through positive social interactions, when it is perceived by others (Eczet Reference Eczet2013:86; see also Lienhardt Reference Lienhardt, Carrithers, Collins and Lukes1985:155). Mursi and Bodi acknowledge that persons are made up of social relations; social interactions are therefore actively sought out (Fayers-Kerr Reference Fayers-Kerr2013).

The opportunity to drink coffee with one’s friends in the morning—something that people missed in the new villages—may seem a banal thing, occurring as it does against the backdrop of land alienation and the upending of a political and ecological order that for the Bodi has deep cosmological significance (Buffavand Reference Buffavand2016). But this practice is indicative of a set of meaningful routines that allow people to experience what is good in life, including connection with others. There’s a naturalness about these routines (“God made us this way”); when people are forced to abandon them, they suffer (“our bodies feel bad”) (see Weisner Reference Weisner, Mathews and Izquierdo2009). In urban settings in Ethiopia, price spikes following the global financial crisis of 2008 impacted on well-being in part because they made it hard for the poor to afford coffee, and precluded the social gatherings associated with coffee drinking (Hadley et al. 2012).^{Footnote 16} Freeman describes social gatherings for coffee drinking among the Gamo of the Ethiopian south-west highlands as a “technique of happiness” which allows people to “subsume [themselves] in the social fabric” (2015:162, 171). The villagization scheme’s effective prohibition of such “techniques of happiness” can be seen as part of a continuum of violence inflicted on the Bodi and others in the name of development.^{Footnote 17} Focusing only on the most obvious manifestations of this violence, such as the bulldozing of one’s backyard, can serve as a barrier to empathy for those of us who have little in our life experience to compare with this. Focusing on the denial of small freedoms, on the other hand, can help make larger injustices more comprehensible.

“To be resettled,” Oliver-Smith has written, “is one of the most acute expressions of powerlessness because it constitutes a loss of control over one’s physical space. The only thing left is . . . the body” (2010:14). The hunger and insecurity that our informants spoke of underline the fact that the body in question is vulnerable. If the set of routines and practices we rely on to make a living and to give life meaning—in short, our culture—is thought of as a second skin, then this is a body that has been stripped of that skin.

Local Resistance, and Its Limits

The sense of indignity and loss of control that we witnessed in the lower Omo are compatible with what researchers of forced displacement have documented in many other places (Oliver-Smith Reference Oliver-Smith2010). Another commonality is the ubiquity of resistance. One of the resources on which the Bodi have drawn in order to preserve a sense of dignity is their long experience of resistance to the projects of powerful outsiders (something they share with many other peoples of the Ethiopian periphery; see James Reference James1979; González-Ruibal Reference González-Ruibal2014). In the case of the resettlement campaign, this resistance took the form both of overt conflict and covert subversion. Open conflict occurred principally between the Bodi and immigrants from Konso, and mostly took the form of tit-for-tat killings. These conflicts began almost immediately upon the arrival of Konso settlers in the government resettlement drive in 2004. The government intervened in these cases by attempting to broker peace meetings. But with the region becoming newly critical to national development plans, the government installed a larger federal military presence in the woreda from 2012 onward. A denouement occurred in early 2014, when, after a new round of fighting had broken out between the Bodi and settlers from Konso, the Ethiopian military intervened on the side of Konso settlers, firing on Bodi who approached Konso settlements; according to Ethiopian Satellite Television the soldiers injured seventeen women, children, and youth (ESAT 2014). In the midst of this conflict, school teachers and other government workers left the new villages, as did most of the settlers who had been living there.

Resistance to the government’s designs was more commonly of a covert kind: (1) refusing to show up to meetings, or to endorse the plans presented at them; (2) partial compliance, in the form of building new houses, laying claim to irrigated land, and receiving (and sometimes re-selling) food aid, and (3) adopting a bet-spreading strategy by trying out life in the new villages, while keeping one foot planted in their former homes. It also manifested in the porous nature of the resettlement sites, and the ways in which people moved back and forth between the new villages and cattle camps and rain-fed cultivation areas—treating the villagization sites not as their only homes, but as another node in their network.

These resistance tactics work, however, only as long as other livelihood options remain open. As Scott has noted in his work on the “hill people” of Southeast Asia, grabbing of land and other resources by outsiders has diminished the power of techniques of resistance that for centuries helped people on the periphery maintain a degree of autonomy. Resettlement—the “transplantation” of “presumptively loyal and land-hungry valley populations” to the hills—has been a key strategy in the outsiders’ playbook (Scott Reference Scott2009:xii). Another and complementary strategy has been to redesign the landscape, using engineering technologies that effectively turn the “hills” into “valleys”—or, in the Ethiopian case, technologies that make the lowlands useful for highlanders. We see all three of these strategies employed by the Ethiopian state and its corporate partners in the lower Omo, to the detriment of the indigenous people of the region.

In some ways our research may have captured a moment when things were at their best for the Bodi in terms of engagement with the government’s designs: they were able to access food aid, but also to live without it. By 2016 these opportunities had narrowed. Filling of the reservoir behind the Gibe III dam, and the suspension of flows from the upper basin, led to the end of the seasonal flood in the lower Omo. The dam, and the diversion of water along newly-built canals, enabled the establishment of the Kuraz Sugar Development Project and private cotton estates, but effectively ruled out river-bank cultivation—a livelihood strategy that was crucial for an estimated ninety thousand people living along the banks of the Omo (Turton Reference Turton2010). The option of opening up new fields by the riverside on which to grow sorghum, or clearing bush on which to grow maize, has been largely foreclosed to the Bodi and their neighbors, through processes that the planners did not undertake to explain.

In 2016, Bodi families who for several years had been growing maize on plots set aside for settlers continued to farm there, but they recognized that this alone would not meet their needs, and many had fallen back on cultivation in rain-fed areas, which due to the end of river-bank cultivation had become newly crucial to their survival. Should the rains fail—not an unlikely scenario in this semi-arid region—famine would ensue, just as it did in the Awash in the 1970s (Kloos 1982). Moreover, since the Omo flood is vital not only for the livelihood systems of most of the people of southwest Ethiopia, but also for Lake Turkana and its fisheries, the dam and plantation schemes plausibly threaten the food security of all indigenous peoples of the lower Turkana basin, on both sides of the Ethiopia-Kenya border (Anonymous 2013; Avery Reference Avery2013).^{Footnote 18}

Bearing Witness to Development Forced Displacement

In drawing attention to the potential of household survey research to hide, rather than expose, the essential features of this predicament, we are sounding a conclusion rather different from the conventional “More research is needed. . . .” The potential of research to represent faithfully the experience of those who are paying the costs of development cannot be taken for granted. In focusing only on the villagization piece of what is in fact a much wider set of interventions, our food access survey risks presenting a misleading picture. As we have shown, the survey results are best interpreted as showing that in both the villagization sites and the community of Gura, a majority of people were struggling to obtain enough food at the time of our research. Having carried out this survey, and having invested considerable effort in conforming as closely as possible to the conventions of quantitative social science research, we can make this claim with particular force. The weakness of the survey—and of cross-cultural surveys in general—is that, in pursuit of data that are widely comparable, it excludes important aspects of local context. As the designers of USAID’s household food insecurity questionnaires noted, “To develop a tool that is culturally invariant, some cultural specificity must be lost” (Deitchler et al. 2010:v). Or, as a cultural anthropologist might put it: the further your instruments travel, the less useful they are (Shweder et al. 2007).

In the time that has elapsed since we carried out the research for this paper, multiple delegations from donor country agencies such as USAID and DFID have paid “fact-finding” visits to the lower Omo, only for their findings to be misreported or buried (Hurd Reference Hurd2013; Turton Reference Turton2014). The testimony of the Bodi that we have placed at the center of this article is a very small contribution towards redressing the debt owed by foreigners to some of the people who confided in them. It may be that the delegates from these agencies failed to report what they heard because it did not make sense to them: it did not square either with the official narrative of infrastructure development and improved service provision (the Ethiopian government’s narrative) or with the views of those critics who have focused on specific abuses, such as assaults and rapes committed by soldiers (e.g., Human Rights Watch 2012). The systematic forced displacement that the Bodi and Mursi have experienced has in fact occurred alongside the infrastructural development of which the government and the Sugar Corporation boast, and it raises more difficult political questions than does any isolated case of abuse. Household surveys, the standard means of generating evidence about human problems in the social sciences, are not well suited for investigating processes such as these. Unless surveys are accompanied by in-depth exploratory research, they risk overlooking the problem and merely reaffirming the preconceptions of the surveyors.

Whether villagization is improving food security and well-being, it turns out, is not a very good question. It assumes certain things that, in the course of our research, we have come to recognize as problematic. Foremost among them is the assumption that villagization is a potentially positive alternative to a livelihood system that remains viable, although possibly less conducive to food security and well-being. In fact, the hydro-engineering schemes that make irrigated agriculture possible for settlers simultaneously undermine the possibility of making a living in the “traditional” way. From being expert stewards of their landscape, the Bodi and other peoples of the Omo-Turkana basin are being forced into the position of ignorant and vulnerable laborers, out of touch with the world around them (see Hobart Reference Hobart and Hobart1993). A survey that compares villagization sites with other communities that have not been villagized, but does not take into account the broader context of resource alienation and massive manmade ecological change, will miss this crucial fact.

By using ethnographic methods—by listening to the people on the front lines of this unfolding crisis, who oriented us towards the relationships between the villagization scheme and broader changes underway in the region—we managed to avoid giving too much credence to the results of the household survey. Spending extended periods of time in the lower Omo, and getting a feel for the dynamics of a culture that has quite successfully weathered past ecological and political changes through a combination of herding, flood-retreat farming, and rain-fed agriculture, also helped us to see the limitations of the food insecurity questionnaire we employed. Food access at a given point in time (which, in spite of the scale’s name, is what the Household Food Insecurity Access Scale really measures) should not be taken as synonymous with the broader construct of food security, which implies risks to food access over the longer term. In actuality, there are few situations in which people are less food secure than when they are dependent on rations from a government or an aid agency, or than when the social and ecological systems on which they have relied for generations are being redesigned around them.

On the basis of the principle, “You broke it, you fix it,” the moral responsibility for repairing the damage done to the livelihoods of the people of the lower Omo now rests with the Ethiopian government, its backers and collaborators—including Salini (the firm that designed the Gibe III dam), the Ethiopian Sugar Corporation, and the lease-holders who are growing cotton for export on land appropriated, without compensation, from local communities. This will require something more than staking out irrigated plots of 0.25 or 0.5 hectares per household, along with a promise that one day they may become outgrowers for the plantation. If analyses of the relative economic productivity of pastoralism and sugar production in the Awash serve as a guide (Behnke & Kerven Reference Behnke and Kerven2013), a more equitable arrangement would be for the Sugar Corporation to hand over all profits from the plantation to the people from whom it has taken the land and water.

As we write, Ethiopia’s Prime Minister Hailemariam Dessalegn is inaugurating the Gibe III dam (Salini 2016). Downstream, the Bodi and their neighbors are hungrier and more desperate than they were before this EUR1.5 billion project began. In the years it would take for further research to be carried out and published, another dam might be built on the Omo, and thousands more people dispossessed and impoverished. Gibe IV is already under construction. For this reason, we urge our colleagues not only to consider the in-built distortions of some kinds of data collection, but also to make existing knowledge more widely available, and to assist in communicating what they know beyond academic circles.^{Footnote 19} One need only look to the extensive literature on “development-forced displacement” to see the parallels with the experience of the Afar and Oromo in the Awash, as well as with dozens of other cases where planned development has forced people into states of dependency. Such initiatives make possible increased security and well-being for some only at the cost of increased insecurity and suffering for others (Roy Reference Roy1999). It is hard to escape the conclusion that in cases like these, famine is not the result of technical failures, but is rather the logical end result of projects based on socially unjust premises.