Policy options to support the Agriculture Sector Growth and Transformation Strategy in Kenya
This report provides scientific evidences supporting policy options for the new Agriculture Sector Growth and Transformation Strategy in Kenya. A Computable General Equilibrium (CGE) model specifically modified for the context of Kenya is used to address the impacts of six policy options. For the purpose of the study, a disaggregated version of a 2014 Social Accounting Matrix (SAM) has been developed for Kenya. Multi-sectoral analytical tools are used to describe the Kenyan economy and inform about which agri-food value chains present the greatest impact in terms of output, employment and value added. Modelling results of policy reforms are then presented, discussed and used to draw policy implications.
The study
The Ministry of Agriculture, Livestock and Fisheries (MoALF) of the Kenyan government is thoroughly reviewing its Agriculture Sector Growth and Transformation Strategy (ASGTS) to update it to fit the new global dynamics of the agricultural sector. A new ASGTS and a new National Agricultural Investment Plan will be established as guidelines for the period 2018-2030. MoALF is performing a consultative process to receive contributions from relevant stakeholders, following an approach that proved effective in preparing agricultural growth and transformation strategies under similar frameworks. A modelling exercise to explore ex-ante socioeconomic impacts of alternative agricultural growth and development options was recommended to support the final decision of MoALF.
Policy context
Agricultural policies, in particular extension services, irrigation investments, rural infrastructure and input subsidies in Kenya have been developed around the main objectives of growing productivity and increasing farmers' income. Within this framework, several policies have been formulated and implemented to introduce stability in agricultural output, to commercialise and intensify production, and to promote appropriate and participatory policy formulation and environmental sustainability. Nevertheless, the Kenyan government is still failing in its commitment to the Comprehensive Africa Agricultural Development Programme with regard to the proportion of public budget to be devoted to agriculture.
Main findings
Scenarios project significant impacts on the production of agricultural commodities. A scenario that subsidises input has the most positive effect on food staples, whereas one that supports irrigation is seen to benefit marketed and exported crops. Agricultural production reacts differently to different policy changes across different regions, but the variability is rather limited.
The value added in agriculture has the greatest increases under the irrigation scenario, and still significantly in the combined scenarios where the policy mix is less biased towards large agricultural farms, to the detriment of smallholders. Kenya is traditionally an agri-food importing country, with export income relying mostly on tea. Under most scenarios, there is an increase of net imports that is considered in line with the projected increase in wealth.
Simulated scenarios show the capacity to create rural jobs and absorb a significant number of workers coming from urban areas. This could be interpreted as an indication that the policy measures simulated have the potential to slow down the urbanisation process.
Irrigation – showed the highest magnitude on agri-food production (especially towards cash crops) and exports. Investment in irrigation has also shown the highest impact on boosting absolute and per capita GDP growth.
Input subsidies – showed the greatest effect on production increases for subsistence farmers and food staples.
Extension – marginally benefits agricultural productivity and contributes significantly to poverty reduction. The analysis showed that investments in extension boost food crops production and incomes of semi-arid and high rainfall areas of Kenya.
Rural roads – boost agriculture value added, mainly of cash crops. Unlike most other policy interventions, investment in rural roads noticeably benefit both agricultural and non-agricultural sectors by helping reduce transaction costs.
Rural health and education – showed only marginal effects on agricultural production, but – not surprisingly – had the highest positive impact on employment generation and wage increases for skilled workers.
Trade liberalisation – benefits the whole agricultural sector, but showed slightly more positive effects on export crops than on food staples.
Policy recommendations
Based on an assessment of the social and economic impacts of the various policy interventions, this report suggests a number of recommendations to guide the new ASGTS and National Agricultural Investment Plan (NAIP).
1. To envisage the allocation of a greater share of total public resources to the agriculture sector, as part of a broad structural transformation agenda. The level of expenditures in support of food and agriculture still falls short of the 10% target. The increase in GDP obtained from devoting 10% of government budget to agriculture is estimated to stand at around 1.5% of overall GDP by 2030. The effect on agricultural GDP is even more substantial as over sixty-percent of the generated value added remains within agriculture.
2. To accelerate Government investment in the livestock sector. The analysis showed that among the value chains analysed for this study, livestock commodities (incl. fisheries) have the greatest positive impact on employment generation, output and value added.
3. To provide targeted public support to maize, fruits and vegetables. Among the food crops analysed, all three commodities appeared as those where highest gains in terms of job creation and income of rural households were observed.
4. To extend public investments in tea and coffee. Of the cash crops analysed for this study, tea and coffee stood out vis-à-vis their effect on boosting output, value added and exports.
5. To stimulate agricultural productivity in the medium and long run, following a balanced public investment strategy that couples spending in agricultural production with rural infrastructure expenditures. This study showed that a combined scenario in which the government invests in irrigation, roads and subsidies to fertilizers and seeds is the most beneficial in terms of value added. This policy mix is also the least biased towards a certain beneficiary, i.e. both larger farms and smallholders benefit.
A social accounting matrix for Kenya for 2014
Access the SAM dashboard
A new SAM for Kenya for 2014 was estimated from the latest re-based national accounts for Kenya. The new SAM deviates from other classical SAMs in terms of structural assumptions. The Home Production for Home Consumption (HPHC) concept is introduced in the SAM by assuming that each household has a corresponding ‘productive activity’. The regional agricultural breakdown in the 2014 Kenya SAM is based on agro-ecological characteristics. The country has been divided into seven agro-ecological zones (AEZs), in addition to the two major metropolises, Nairobi and Mombasa. The Households have been further disaggregated into rural and urban, according to the area of residence. Moreover, the two metropolises, Nairobi and Mombasa, have been broken down into income quintiles. According to the classification of work by education, there are three types of labour in the SAM: skilled, semi-skilled and unskilled labour. Each labour factor is also regionalised. The SAM is fully documented in Mainar Causape et al., (2018).
The computable general equilibrium model
The model used for this study is a recursive dynamic single-country computable general equilibrium (CGE) model. This version is an extended version of the model documented in Aragie et al. (2017) which several enhanced features: recursive dynamic version, handling of migration flows, impacts of investments on physical and human capital, demographics, modelling of nested production and international trade in the context of the developing countries. The model incorporates additional behavioural relationships that better account for economic relationships in developing countries, in particular in the least developed and Sub-Saharan African countries. To properly model agriculture and food security issues in Kenya, the model depicts key structural characteristics of the economy and of the agricultural sector as the dual role of semi-subsistent agricultural households, which play the non-separable double role of producers and consumers. Other additions to the model, including nested consumption function, endogeneity of the functional distribution of income, domestic migration and factor market segmentation, are fully documented in Aragie et al. (2017).
Policy scenarios
Extension
Government expenditure on extension services is gradually increased up to 10% of the government budget, starting from the initial level of current Kenyan government budget spending. The change in government spending on extension services triggers a productivity increase in agricultural production through labour, fertilisers and seeds. It is assumed that spending on extension services will allow farmers to use fertiliser and seeds in a more efficient way and also increase their farms’ productivity.
Input subsidy
The government expenditures in input subsidies (fertiliser and commercial seeds) are gradually increased up to 10% of the government budget, starting from an initial level of 200 million KSh of current Kenyan government budget spending and gradually going up to 10% of the total government budget.
Output subsidy
Subsidies to small-holder producers which provide the farmers a mark-up on the average price per unit of composite output they produce, are increased up to 10% of the government budget starting from zero in the base year.
Irrigation
The investments in irrigation (which mainly uses the construction sector) are gradually increased up to 10% of the government budget. The investment is financed by an equivalent increase in government savings. We also compare the irrigation development for small holder farmers and commercial farms (i.e. large farms) in two sub-scenarios presented in Annex 4. The scenario does not account for environmental issue such us water availability which is taken as granted or other non-economic related issues.
Rural roads
The investments in road (which mainly uses the construction sector) are gradually increased up to 10% of the government budget. The investment is financed by an equal increase in government savings. Investment in rural roads lowers the cost of bringing agricultural and food products (among others) to the market for all agri-food sectors.
Rural health and education
Government expenditures in health and education are gradually increased up to 10% of the government budget. Labour productivity, death rates and birth rates are a function of health and education spending. Specifically, higher education expenditures are linked to an increase in labour productivity and a decrease in birth rate, while higher health expenditures are linked to a decrease in the death rate.
Combined scenarios
Agriculture-focused scenario
10% of the national budget is allocated among policy options, with a higher share allocated to those aimed at improving agricultural production. A third of the budget is allocated to irrigation investments, a third to input subsidies to fertilisers and seeds, and a third to extension services.
Rural-economy-focused scenario
10% of the national budget is allocated among policy options, with a higher share allocated to those aimed at developing the rural economy. A third of the budget is allocated to rural roads, a third to public expenditures on health and a third to public expenditures on education.
Trade liberalisation with African countries
A hypothetical liberalisation of agricultural trade between Kenya and other countries is simulated. A 20% decrease in the tariffs imposed by Kenya on imports from other African countries for agricultural and food commodities is simulated. Different specifications for exports to African countries and exports to the rest of the world apply. We assume that reducing tariffs on Kenyan exports would cause a 20% decrease in the price of Kenyan export commodities.
Economic growth
Absolute and per capita GDP, against the baseline, show positive and very similar patterns under all scenarios. The irrigation and road scenarios are those whose increase result more pronounced.
The agricultural value added (summing all agricultural crops), given the exogenous assumptions of the baseline, is growing by more than 5% per year, on average. Output subsidies and investments in irrigation and road have the highest impact in boosting agricultural value added.
Combined scenariosPoverty reduction
The model is not set to answer precise questions about poverty reduction. A poverty reduction analysis would rely on microsimulations which could have showed detailed results on poverty reduction effects of simulated policies. Unfortunately, the last available household budget survey, on which microsimulations have to be based, is more than 10 years old. Nevertheless, the model can give a few limited hints on the issue of poverty reduction looking at the income, expenditures and welfare changes of different representative household groups within the country.
Looking at household income, expenditure and welfare, a clear pattern can be observed. Extension has limited impacts on income, irrigation and outputs subsidies benefit mainly rural households (and partially urban outside the large cities). Road investments have a more spread benefit to urban and large cities households. Education is more beneficial for households outside the large cities, while trade liberalisation brings balanced welfare and income gains. Again, investments in irrigation and road are the scenarios with the highest overall benefit.
Combined scenariosJob and migration
Regarding the capacity to generate new jobs, employment multiplier analysis shows that agricultural and livestock sectors show, in general, multipliers values above the average of the Kenyan economy. In particular, the values of cash crops (maize, other cereals, roots, fruits and vegetables) and most livestock sectors should be highlighted.
Policy scenarios show the capacity to create rural jobs and absorb a significant number of workers coming from the urban areas. In a way, this can be interpreted as the possibility of the policy measures simulated to slow down the urbanization process.
As highly expected the education scenario is the one with the highest impact in terms of increase wage rate and increased employment opportunity for skilled workers.
The results suggests that agricultural policies can help reducing the migration in most of the regions except Coastal and Arid North, which offer rather limited possibilities for expansion of agricultural production through the simulated policies. Hence in these regions, the policy makers should consider further measures to improve both the living conditions of the rural households and technical conditions of agricultural production.
The model results suggest that the simulated policies will boost the employment in agricultural sectors. The policies that link to the other sectors of the rural economy are more likely to help create jobs in non-agricultural sectors, especially in food production sectors. This puts emphasis on the need for a holistic approach in the job creation impacts of the scenarios.
Improving the road infrastructure is the most effective policy to create skilled jobs both in agriculture and food production sectors. Further its adverse effect on non-agricultural sectors is lower compared to the other scenarios.
Combined scenariosExport competitiveness
Kenya is traditionally an agri-food importing country, with export relying on few commodities, especially tea. Under all scenarios but extension and input subsidies, there is an increase of imports that should be considered on line with the increase in wealth.
Under all scenarios, exports of agri-food commodities increase, even with some variability in the magnitude of changes. The extension and more evidently the input subsidy scenario, increase the competitiveness of staple food exports. This looks beneficial for the food industry sector which expands its exports possibility.
A complete different effect is achieved under the irrigation (and partially road) scenarios where traditional export crops become much more competitive than staple and expand their global market significantly.
The trade scenario shows that a deeper African integration represents a great opportunity to expand exports of agri-food commodities.
Combined scenariosFood security
Food availability
A first element of food security is food availability. Under all scenarios, domestic production of agriculture and in particular staple food increase. Subsiding input will have the most positive effect on food staples and HPHC production. Some scenarios (improving extension, road infrastructure or education) will mainly scatter positive effects, especially those labelled as HPHC. The agricultural production reacts differently to policy changes in regions but the variability across regions belonging to the same category (small holder, export or market oriented) is rather limited.
The combination of the above mentioned effects, contribute to an increase in agri-food consumption under all scenarios.
Food affordability
In terms of food affordability, scenarios subsidising extension or inputs bring most of food staple prices down. On the other hand, investments in irrigation and road has the effects of reducing significantly the prices of exported corps (which are almost not consumed domestically) at the expense of staple crops and meat commodities which in general become more expensive.
Irrigation
Irrigation stands as one of the most effective policy intervention to improve the livelihoods of rural areas by increasing agricultural production and household income. Inevitably, high rainfall regions which have higher water availability and thus more irrigated land as well as the commercial farms and households that specialize in the production of cash crops (who are also concentrated in the high rainfall regions) turn out to be main beneficiaries of irrigation expansion. Hence, the benefits of irrigation investments are not uniform across households and agricultural activities. The two-sub-scenarios presented in Annex 4 shows that targeting small-holder or large commercial farms have quite different impact on the economy. The main impact channel of the small-scale targeting irrigation expansion is through the increasing food crops production which in turn increases household income in rural areas. In urban areas, the declining food prices improve the welfare of poorer households by allowing them to access cheaper food products thanks to the increasing production. On the other hand, irrigation expansion that target commercial farms specialized in producing cash/export crops mainly improves the welfare of households by increasing the exports and hence allowing Kenya to import more of non-food commodities. This mainly benefits to the richer households in urban areas and rural households of the regions where cash/export crop production is higher (e.g. high rain fed region). Further, small-scale irrigation also improves the availability of water for the households, especially in rural areas, allowing better meeting the domestic needs (human and animals) and sanitation.
It should be taken into account that, particularly large irrigation schemes have suffered, also in the recent past, from: lack of sufficient water to be replicated at a large scale, complex managerial and procurement issues (including wasteful and corrupted use of public resources, possibly, problems of forced displacement and elite capturing most of the benefits. Many of these issues cannot be taken into account by the mode, so policymakers should factor them in a cost-benefit analysis of irrigation investments.
