Fiscal Decentralisation, Revenue Performance and Poverty in Ghana

Abstract

Sub-Saharan Africa (SSA) has experienced robust growth since the mid-1990s however, formidable development challenges such as poverty, unemployment, and low revenue, and inefficient public service delivery persist. Poverty particularly has become a major global concern, with its eradication in all forms and dimensions by 2030 constituting the first, and perhaps the most critical goal among the Sustainable Development Goals (SDGs). Against this background, SSA countries including Ghana have adopted decentralization policy since the 1980s as an alternative development strategy. The decentralization policy is anchored on grants from the central government and donors. Additionally, the Local Authorities are granted the autonomy to mobilize resources to fund their local development plans. The relationship between decentralization and development outcomes as espoused in the theoretical and empirical literature remains unclear. Whilst a school of thought postulates that the grant system discourages revenue effort and exacerbates poverty, others proffered otherwise. Also, some studies show that intergovernmental transfer arrangements may be susceptible to political manoeuvering that could create social conflicts. Whilst extensive empirical studies exist on fiscal federalism and development outcomes such as poverty in developed economies, very little is known about Sub-Saharan Africa (SSA). This study contributes to the literature on SSA, specifically Ghana using the recent Ghana Living Standard Survey (GLSS 7), district and community-level data for the period 2012-2016 to examine the relationship between grants, local revenue mobilization, and poverty in Ghana. It also examines the influence of political manipulation on DACF disbursements in Ghana. The contribution made to the literature in this thesis covers a number of areas. First, we examine the effect of grants on local revenue mobilization by testing the flypaper hypothesis. The system GMM results show that a 10 percent increase in total transfer leads to 2.7 percent reduction in local revenue at 1 percent. However, disaggregated components of grants into unconditional, conditional and sectorial-limited transfers have no effects on revenue. Second, The OLS regression results show that a 10 percent increase in DACF reduces district-level poverty by 6 percentage points whilst local revenue does not affect poverty. Moreover, the Logit and Probit regression results show that local services such as complementary (non formal) and primary education, construction of development projects, and improvement in job opportunities are positively associated with perceived improvement in community welfare. Third, the thesis assesses the extent to which political consideration influences DACF allocation in Ghana. The system GMM result reveals that not only politically-aligned districts are targeted in the distribution of DACF, but also politically-swing districts are well targeted. This study found evidence in support of the argument that grants crowds-out revenue whilst the former reduces poverty. It is recommended that, grants should be directed towards improving local tax administration system whilst innovative strategies are adopted to maximize local revenue. Grants should be invested in technology to ensure that local tax administration processes are digitalized. Additionally, the 5 percent minimum threshold of the DACF allocation should be reviewed to 10 percent of national revenue similar to those in SSA in order to accommodate the rising cost of living in the various localities. Furthermore, apolitical team of experts should be formed to manage the DACF in consultation with Non-Governmental Organizations (NGOs), Civil Society Organization (CSOs) and the MMDAs to minimize the occurrence of political manipulation