On July 28th, local governments were given center stage at the United Nations Food Systems Pre-Summit dialogues to discuss how to tackle food insecurity and nutrition. The growing prominence given to local governments reflects the outcomes of longstanding processes of decentralization that have occurred globally since the early 1990s. Decentralization involves the transfer of certain administrative functions, fiscal responsibilities, and/or political autonomy to local governments. Devolution, which involves the transfer of key responsibilities to local authorities that are elected rather than appointed, is the most extensive form of decentralization. It is theoretically viewed by scholars as a way to enhance accountability to local communities, tailor goods and services to citizen preferences, and mitigate spatially-based grievances. From Kenya to Pakistan, and from Zimbabwe to Nepal, many countries have altered their constitutions in recent years to pursue devolution.
What are the impacts of such reforms on agricultural service delivery? A recent article supported by PIM examined this issue in Ghana. In 2009, Ghana’s parliament passed a Local Government Instrument that devolved responsibility for certain functions, including agricultural extension, public works, and social welfare, from the central ministries to the country’s Metropolitan, Municipal, and District Assemblies (MMDAs). By 2012, implementation of the devolution process began with a composite budgeting system whereby the budgets for devolved functions became part of the MMDAs’ budgets and under control of elected MMDA officials rather than earmarked and allocated through sectoral ministries. In addition, approximately 33,000 staff were transferred from the central government civil service to a newly established Local Government Services Secretariat. Directors of agriculture and their staff henceforth were accountable to the MMDAs rather than to the Ministry of Food and Agriculture.
The article’s analysis focuses predominantly on how these budgeting changes and shifts in reporting lines for civil servants affected agricultural extension services. The study relied on an original survey of 960 households, structured interviews with 80 of the MMDAs’ directors of agriculture, and district-level budget data from 2012-2016. The findings indicated an important paradox: devolution enhanced accountability but undermined agricultural service provision.
In particular, the study found a 36 percent decline in median agriculture expenditures across Ghana’s 216 MMDAs in the five years after the reforms, even as median expenditures on public works increased by 24 percent during the same period*. The directors of agriculture confirmed that they received fewer resources for extension services after devolution than prior to the reforms, with 65 percent of directors claiming they could not negotiate for sufficient resources during the composite budget planning process. One of the most common reasons given by the directors is that the politicians within the MMDAs who vote on the budget allocations prefer prioritizing investments in physical projects, such as roads, school blocks, and health compounds, over less visible services such as agricultural extension that are relevant to a smaller share of the community.
The household survey largely confirmed this pattern. Indeed, among both farmers and non-farmers, investments that fall under the public works mandate were favored by 36 percent of households, compared with only 2.3 percent prioritizing agricultural extension services. At the same time, 80 percent of respondents indicated that they had voted in the 2015 local elections and almost 90 percent claimed they would vote out their assembly representative if s/he did not deliver on campaign promises. A majority also indicated that they would first contact their elected assembly member over any other actor when faced with failing services in their community. This collectively suggests that local communities appreciate that devolution enables them to hold their local assembly members accountable through elections. In other words, citizen priorities reflected sectoral budgeting patterns and because devolution reinforced local government accountability, politicians were motivated in the budgeting process to favor those goods and services that citizens want most. This often disadvantaged agriculture services vis-à-vis other sectors.
One caveat is that the study focused only on a limited period—five years—after Ghana’s devolution process. Nonetheless, it affirms similar findings from studies conducted in Pakistan, Kenya, and Uganda. In addition, it offers some useful policy implications. First, national agriculture development strategies and spending targets, such as the 10 percent spending target for agriculture under the African Union’s Comprehensive African Agriculture Development Plan (CAADP), need to reflect parallel public sector reform processes, such as devolution, that can affect the viability of meeting such targets on the ground. Second, low levels of own-source revenue within most of Ghana’s MMDAs exacerbated sectoral spending trade-offs; especially given steep declines in sub-national revenue in the wake of Covid-19, this emphasizes the need to diversify sources of local government financing for devolution to work effectively.
In sum, the study suggests that while locally elected governments play an important role in meeting agriculture and food system goals—including those being discussed at the UNFSS— their success in doing so requires being not only properly financed but also motivated by a strong mandate from their constituents.
*At the time of fieldwork, Ghana had 216 MMDAs. Since then, they have increased to 254.
Danielle Resnick is a Senior Research Fellow with IFPRI's Development Strategy and Governance Division and leads IFPRI's Governance theme. At PIM, Danielle leads cluster of research on Political Economy and Policy Processes under Flagship 2: Economywide Factors Affecting Agricultural Growth and Rural Transformation.
Photo: Budget hearing at the Gomoa East District Assembly.
