Pastoralism in Kenya
Pastoralism in Kenya
Pastoral regions of Africa are witnessing a sharp increase in armed civil conflict. Research has identified ‘mismatched’ agricultural development projects as one major factor driving this violence by displacing local pastoral groups. However, the evidence also points towards some potential solutions that may bring about peace.
Local context matters for the success of development projects. For example, in Indonesia, a nationwide school construction programme boosted girls’ educational attainment only among ethnic groups who practice ‘bride price’ customs (Ashraf et al. 2020). Similarly, in central Africa, contemporary public health programmes are less successful in villages where colonial-era governments implemented harmful medical campaigns (Lowes and Montero 2021). In these cases, cultural and historical characteristics strongly influenced the degree to which development interventions were well-matched or ‘mismatched’ to the local populations they are designed to support.
The role of ‘mismatched’ development projects in aggravating civil conflict
The 2008 World Development Report called for increased investment in crop agriculture projects across Africa. However, for much of the continent’s population (around 20%), and even more of its land mass (40%), the dominant mode of subsistence is pastoralism rather than crop agriculture. According to reports, pastoral groups are sometimes forcibly removed from communal grazing lands to make way for these development projects, which, in turn, may disrupt the peaceful equilibrium between them and neighbouring agricultural groups.
In McGuirk and Nunn (2024), we examine the impact of agricultural development projects that are implemented in traditionally pastoral areas of Africa (mismatched projects) and compare them to those implemented in traditionally agricultural areas (matched projects). Specifically, we study the potential role of mismatched development interventions in aggravating civil conflict in these areas.
Two sharply contrasting perspectives emerge upon examining the existing evidence. On the one hand, states (and their donors) maintain that commercial crop agricultural projects will bring peace and prosperity to pastoral Africa. By increasing the productivity of labour-intensive agriculture, these projects are expected to increase the opportunity cost of armed conflict, yielding peace (Dal Bo and Dal Bo 2011, Dube and Vargas 2013, McGuirk and Burke 2020). An additional motive arises from the belief among certain state actors that pastoralists ought to transition to sedentary livelihoods. By this logic, agricultural projects will not only raise output, they will also “civilise” mobile pastoralists who otherwise represent a threat to settled communities (Catley et al. 2013).
On the other hand, case-study evidence suggests that development projects often make pastoralists worse off. Pastoralists’ dependence on highly flexible customary land-use arrangements leaves them particularly vulnerable to expropriation. While one might expect a Coasian solution to the problem—in which displaced pastoralists are compensated using project surplus—in practice these transfers tend to be scarce, leaving pastoralists to resort to violence.
Using spatial panel data to study the impact of agricultural projects
To determine which of these accounts is more systematically consistent with the data, we build a panel dataset at the level of a 0.5 degree cell covering the African continent over a 20-year period. To observe conflict, we use data from the Uppsala Conflict Data Program and Armed Conflict Location and Event Data projects to generate a measure of conflict incidence that is equal to one when there is at least one conflict event in a cell-year and zero otherwise.
To observe the traditional location of transhumant pastoralist ethnic groups across Africa, we combine data from Murdock (1959), which is a map of African ethnic societies at the eve of colonialism, with data from the Ethnographic Atlas (Murdock 1967), which contains information on the modes of subsistence for these groups. We generate a 0-1 index of reliance on transhumant pastoralism, which captures the share of consumption from animal husbandry for groups who are mobile and whose primary livestock animal is suited to herding. The spatial distribution of this measure is presented in Figure 1.
Figure 1: Spatial distribution of transhumant pastoralism index
Spatial distribution of transhumant pastoralism index
Spatial distribution of transhumant pastoralism index
Finally, to observe the location of development projects, we turn to AidData (2017), which organises data on World Bank development projects from 1995 to 2014. This dataset contains information on the time, location, and sector for nearly 15,000 projects across Africa during this period. Of these, 22% are in agriculture only and a further 4% combine both agriculture and ‘animal production’, or pastoral, interventions. We use the residual 74% of non-agricultural projects as placebo checks in our analysis.
To estimate the effects of these development projects, we compare the change in conflict in cells that receive a project to the change in nearby cells that do not. The main findings in our paper are as follows. We first examine the impact of well-matched agricultural projects, i.e. those located in traditionally agricultural societies. We find that these projects lower the risk of armed conflict by 1.6 percentage points, or around 50% of the mean. This is consistent with the hope that commercial agricultural projects can generate peace by raising the productivity of farmers.
Agricultural projects in traditionally pastoral societies lead to armed conflict
We see a starkly different picture for mismatched agricultural projects, i.e. those located in traditionally pastoral societies. These projects raise the risk of armed conflict by around 6 percentage points—almost double the mean—in the median pastoral region. These contrasting effects are illustrated in Figures 2 and 3, where we plot the dynamic impact of projects on conflict in the traditionally agricultural and pastoral samples, respectively.
Despite these large and persistent effects on conflict in pastoral areas, we do find evidence of a potential solution. In cells where crop agricultural projects were paired with animal production projects, the effect on conflict is fully muted. This implies that a joint allocation of projects that allows both sides to share the benefits is sufficient to avoid conflict altogether.
Interestingly, we find that this joint allocation is more likely to occur when pastoral groups have a greater share of power in national politics. This increase appears to come at the expense of non-agricultural projects, suggesting that pastoral groups are willing to trade these projects for alternatives better suited to their environment. The net effect of this reallocation is reduced conflict in pastoral areas. We interpret this as evidence of a mechanism linking political power-sharing to peace.
To better understand specific mechanisms, we study the effect of development projects on land-use changes using data from the Copernicus Climate Change Service. We find that agricultural projects result in the conversion of pastureland to cropland, but the effect is again muted when animal production projects are part of the project bundle. This suggests that conflict is due to forced changes in land use, and that by pairing agricultural and pastoral projects together, groups can reap the benefits of greater productivity without experiencing the rise of violence.
Figure 2: The effect of agricultural projects on conflict in agricultural cells (THP=0)
The effect of agricultural projects on conflict in agricultural cells (THP=0)
Figure 3: The effect of agricultural projects on conflict in pastoral cells (THP>0)
The effect of agricultural projects on conflict in pastoral cells (THP0)" data-entity-type="file" data-entity-uuid="5c7b65ce-a284-453d-bf97-8b8bc36f9063" src="https://voxdev.org/sites/default/files/inline-images/figure3\_8.png">
The effect of agricultural projects on conflict in pastoral cells (THP>0)
Agricultural projects leave pastoral households worse off
Finally, examining development outcomes, we find that agricultural projects tend to increase economic activity (as measured by nightlights) in both agricultural and pastoral areas. Thus, while projects deliver both prosperity and peace in agricultural areas, the picture in pastoral areas is mixed, with greater economic activity but more conflict. One explanation for this pattern is that, within pastoral areas, agricultural projects make pastoral households worse off than non-pastoral households. We confirm this interpretation using data from the Demographic and Health Surveys. In the presence of mismatched agricultural development projects, households from pastoral ethnic groups are one standard deviation poorer than households from non-pastoral ethnic groups in the same cells.
Pathways to peace and prosperity for pastoral communities
Our findings indicate that, in general, pastoral households do not share in the prosperity generated by agricultural development projects in pastoral territories, increasing the risk of conflict. However, peace and prosperity are more likely to prevail when pastoral groups have sufficient power to influence the type of development projects implemented on their land. Our findings illustrate the importance of having policies that match the needs of the local population.
References
AidData (2017), WorldBank_GeocodedResearchRelease_Level1_v1.4.2 geocoded dataset, World Bank.
Ashraf, N, N Bau, N Nunn, and A Voena (2020), “Bride price and female education,” Journal of Political Economy, 128(2): 591–641.
Catley, A, J Lind, and I Scoones (2013), “Development at the margins: Pastoralism in the Horn of Africa,” in A Catley, J Lind, and I Scoones (eds.), Pastoralism and development in Africa: Dynamic change at the margins, Routledge.
Dal Bó, E, and P Dal Bó (2011), “Workers, warriors, and criminals: Social conflict in general equilibrium,” Journal of the European Economic Association, 9(4): 646–677.
Dube, O, and J F Vargas (2013), “Commodity price shocks and civil conflict: Evidence from Colombia,” The Review of Economic Studies, 80(4): 1384–1421.
Lowes, S, and E Montero (2021), “The legacy of colonial medicine in Central Africa,” American Economic Review, 111(4): 1284–1314.
McGuirk, E, and M Burke (2020), “The economic origins of conflict in Africa,” Journal of Political Economy, 128(10): 3940–3997.
McGuirk, E F, and N Nunn (2024), “Development mismatch? Evidence from agricultural projects in pastoral Africa,” Unpublished manuscript.
Murdock, G P (1959), Africa: Its peoples and their culture history, McGraw-Hill.
Murdock, G P (1967), Ethnographic atlas, University of Pittsburgh Press.
Africa civil conflict Agricultural development pastoralism Conflict Agriculture