The pro-poor hypothesis of decentralisation with respect to the creation of districts is flawed at best and elusive at worst to the good governance and efficiency proposition. There is no evidence to the claim of district creation reducing poverty in Ghana, yet it has been used to explain the 94.4% increase in districts created since 1988. Using a cross-sectional data, this paper assesses the poverty reduction basis of district’s creation in Ghana. It concludes the creation of districts as a convenience political exercise and a trapping mechanism which maintains a poverty status quo rather than alleviation.
1 Introduction
Ghana’s decentralisation with respect to the creation of districts as a tool for strengthening democracy, fostering local development and subsequent poverty reduction (Republic of Ghana, 1992; 1993) is marred by nothing more than political convenience (Ayee, 2013).[1] There is scant evidence that service delivery and poverty indicators have improved as a result of decentralized governance (Cabral, 2011). Its pro-poor hypothesis is flawed at best and elusive at worst to the good governance and efficiency proposition. There are currently 216 decentralized local authorities in Ghana. The number of authorities created since the inception of decentralisation in 1988 to date has increased by 94.4%. As much as this brings decision making closer to its subjects (Hemdi, 2014), what is overlooked are the poverty traps laid in the quest for a more inclusive and effective governance.
Fragmentation leads to governance efficiency through competition (Ostrom, Tiebout & Warren, 1961; Ostrom, 1998). Conversely, it can also sustain and exacerbate territorial inequality if not complemented with requisite resource endowments and targeted transfers (Von Braun & Grote, 2000). Such is the bane of Ghana’s decentralisation; good in reaching out but less so in translating into poverty reduction.
There are studies regarding the correlation of decentralisation (administrative and fiscal) and economic development (Davis, 2006; Blochiger, 2013). Most empirical evidence though not conclusive (Martinez-Vazquez & Rider, 2005; Scott, 2009) are usually based on fiscal decentralisation (Blochiger, 2013). The creation of districts regarded as one of the key processes of decentralisation (Ayee, 2013) is under-explored in the decentralisation-poverty reduction discourse.[2] The contribution of this paper, therefore lies in filling this gap by addressing the question;
Does the creation of districts contribute to poverty reduction?
This paper argues that the creation of districts (disregarding existing resource endowment and targeted transfers) is a trapping mechanism, which maintains a poverty status quo rather than alleviation. Using an ordinary least square (OLS) estimation of 216 districts in Ghana, the paper identifies a negative and significant correlation between a district’s resource endowments (internally generated funds (IGF)) to its level of poverty. Two poverty traps, parallel to the district’s resource endowment are identified to confirm a league of poverty clusters arising from decentralisation and district creation.
Following the initial introduction of the paper, the next session gives a brief background of literature regarding the theoretical and empirical correlation of decentralisation and development. It looks at the creation of districts as a vital part of administrative and fiscal decentralisation and narrows down to the politics of it other than its development intent in Ghana. Establishing the convenience of district creation, the third session conceptualizes the creation of districts as a poverty trapping mechanism. This is empirically tested in the last session using Ghana as a case point.
2 Background Literature
Decentralisation has always been a well-acclaimed policy intervention in opening up governance and grassroots participation in development (Smoke, 2003; Robinson, 2003). It is laudable considering the undoubted relevance of direct and effective participation for development (Hendi, 2014). As a measure of the good governance agenda (see Andrews, 2008), its contribution to poverty reduction in developing countries has however realized mixed results in literature (Ndegwa, 2002; Martinez-Vazquez & Rider, 2005; Scott, 2009; Cabral, 2011). Crawford (2008) for instance questions its impact on development. In a case study in Africa, he reveals decentralisation as vastly a mere convenient political exercise implemented not necessarily as a prerequisite for development.
2.1 Decentralisation and district creation
Ayee (2013) elaborates the theoretical debate regarding decentralisation and the creation of districts as a vital aspect which has been overlooked. In summary, fragmented governance approaches other than consolidation (see Tiebout, 1956; Ostrom, Tiebout & Warren, 1961; Ostrom, Bish & Ostrom, 1998) have sufficed to dominate institutional arrangements in decentralisation.[3] Following neo-classical economics, it features a decentralized political arrangement that favours local competition for efficient service delivery. Based on such efficiency, it is argued that the creation of autonomous local authorities (districts) makes for a better governance through popular participation, responsiveness, effective targeting in policy interventions and a structural vertical relation with the central government that allow people (particularly the poor) to self-govern.
This has guided the rationale for decentralisation reforms in Africa over the last 30 years but behind these intentions, political motivations have been a major driving force (Cabral 2011). The ingredients for successful decentralisation imply a complimentary political, fiscal and administrative devolution (Jütting et al., 2005). However, in most instances administrative authority for development has been decentralized whilst the requisite fiscal means of it lags behind (Ndegwa, 2002).
2.2 Decentralisation and the politics of district creation in Ghana
Ghana practices a unitary presidential and constitutional republic with a multiparty democracy enshrined in the Constitution of the Republic of Ghana, 1992. Demographically, Ghana has a population of about 24 million, with a growth rate of 2.5% (Ghana Statistical Service, 2015). Decentralisation in Ghana, evident in the local government system, had gone through many trials till 1988 when it became more effective with the consolidation of local authorities from 65 to 110 (Kuffour, 2005; ILGS, 2010).
2.3 The structure of the local government system[4]
Local government is enshrined in the constitution under the Local Government Act 1993 (Act 462). Ghana has a three-tier local government system comprising: 10 Regional Coordinating Councils (RCC), and currently, 216 sub-regional entities (loosely referred to as districts) made up of districts, municipals and metropolitan areas (see Figure 1).
Figure 1: Hierarchies of the decentralisation in Ghana
Under the Ghana decentralisation policy and local government reform, development is a shared responsibility involving the Central Government, Local Authorities and the Communities (MLGRD, 2010). The local authorities are the highest political and administrative authority at the sub-national level. They all have the same internal political structure with District/Municipal/Metropolitan Chief Executive (DCE/MCE), a position similar to an executive mayor, as the political head. Their mandate ranges from legislative, administrative, service delivery, spatial planning to most importantly socio-economic development (Kuffour, 2005). As development authorities, the Metropolitan, Municipal and Districts (MMDs) have two main revenue sources. These are summed as internally generated funds (IGF) and externally generated funds. The IGF is made of jurisdictional property taxes, returns from local government capital investment, business operation taxes, fee, fines and issuance permits. The externally generated funds are made of central government transfers, and other grants from both national and international sources. Significantly, the government transfers are from the District Assemblies Common Fund (DACF) which is a 7.5% fiscal transfer to MMDA’s from the national consolidated fund for developmental projects (GOG 2016).
2.4 The convenience of district creation in Ghana
The creation of districts has been argued to facilitate development and poverty reduction by devolving power, competence, and resources to the district level (Ayee (2013) citing Rawlings, 1988a; Kuffour, 2005; Ofosu- Ampofo, 2012). Though consistent with the proponents of fragmented governance, the capacity of the created districts in terms of a requisite resource base remains a key question.
Agreeing with Ayee (2013), emphasis on democracy and inclusive governance has overshadowed attention to the functional effectiveness of districts created. Districts are proliferated with disregard to their economic viability necessary to stimulate development and poverty reduction (Ahwoi, 2011). With unlimited power of the president to create new districts (ibid), the politics surrounding the creation of districts cannot be underestimated. Authorities created since the inception of decentralisation in 1988 to date has increased by 94.4%. There is no proof of gerrymandering intents but the coincidental creation of districts before every general election is suspect of manipulations and convenience (see Table 1).
Table 1: Trend of district creation in Ghana’s decentralisation
Source: The authors.
Also, though the districts are established structures of sub-national governance which are independent, there is a continued dominance by the central government. Central government maintains its presence in local decision making with appointees and nominations of mayors to the districts (Jütting et al., 2005).
More importantly fiscal autonomy remains to be addressed. Fiscal constraints also persist as a common and recurring challenge cited in the annual progress reports of districts (NDPC, 2016). This undermines their viability in making a meaningful contribution to poverty reduction. That notwithstanding, the resource base needed to drive development in districts is not considered as prime, prior to their creation.
A remedy lies in central government transfers (DACF) which majority of districts overly rely on (Crawford, 2008; Ahwoi, 2010; Awortwi, 2010; Ayee and Dickovick, 2010). This is however positively related to population as its disbursement is predominantly based on needs (see GOG, 2016 for DACF calculation and disbursement). Considering that districts are created with respect to population size, the rich will become richer or at worst stay rich while the poor at best maintains their status quo. IGF hence becomes the most critical and differential resource base to achieve any meaningful development in the districts. The overarching hypothesis to be tested hence is that the creation of districts with disregard to their respective capacity in generating enough IGF propagates a poverty status quo (traps) other than reduction.
3 Conceptualizing the creation of districts in Ghana as a poverty trapping mechanism
There are three levels of local government structures (Districts) in Ghana. Based on the size of the population there are Metropolitan areas (Urban), Municipal areas (predominantly peri-urban) and districts (predominantly rural in terms of services). The categorisation does not reflect resource endowment prior the creation of the sub-regional governments (districts). Yet, resource endowment and financial transfers are basic conditions for effective decentralisation (Cabral, 2011) and poverty reduction at large. Considering this, the first consequential hypothesis to be tested is that;
Hypothesis 1
· Districts with low IGF will have a high rate of poverty and districts with high IGF will have low rates of poverty.
Secondly, population size is positively correlated with the generation of IGF which forms a significant resource base for developmental programmes and projects. A higher resource base (IGF) leads to a better capacity in providing developmental services that translate into poverty reduction. The proceeding hypothesis then stands that;
Hypothesis 2
· If the creation of districts is based on population which signals a capacity to generate resources, poverty will be predominantly severe in districts (the lowest population level).
Creating a district without a regard to its capacity for raising financial resources hinders the operationalisation of local governance for development services and subsequent poverty reduction aims. With a variation in the IGF base, two poverty traps can emerge. A district poverty trap and another trap that engulfs municipal and metropolitan areas (see Figure 2).
Districts (161 in total) represent the lowest categorisation of sub-regional governments in Ghana. Usually rural, they have the lowest population (minimum of 75,000 people) and level of available services. With low resources to provide basic and higher order services, poverty headcount is high and will remain high in districts since they lack the attractiveness to business and population. Rural urban migration compounds the problem and revenue generation capacity drops to worsen the problems of poverty.
Municipalities (49 in total) have a higher population (minimum of 95,000 people) and as such a higher resource base. They provide high order services and opportunities that attract the population from the rural districts who migrate for better livelihood. Compared to metropolitan areas (6 in total) resources are however not enough to provide higher order services that facilitates business. Usually used as dormitory towns, they tend to have floating populations who work and contribute to substantial business operating taxes in the metropolitan areas than the origin (municipalities). Other taxes like property tax accrues to the municipal areas. This makes available a substantial resource base for development, so poverty is comparatively lower in terms of headcount.
There is a strong interaction between municipal areas and metropolitan areas. Metropolitan areas (minimum of 250,000 people) have the resource base large enough to support the provision of higher order services that sustain businesses. They generate the highest amount of internalrevenue, however, there is an influx of population (from both the districts and municipalities) to these areas. This puts pressure on developmental services and infrastructure which deteriorate in time and are under constant rehabilitation. Resources to alleviate or address poverty reduction issues are hence channelled into rehabilitation and repairs which contribute less to poverty reduction. Poverty hence persists in the metropolitan areas but this is comparatively low in relative terms.
In sum, fiscal constraints (capacity to generate internal funds) limits what local government entities – arising from district creation – can effectively do to realize the poverty reduction intent of decentralisation. Poverty headcounts are hence likely to be higher in areas with low IGF base in an ascending order of districts, municipalities and metropolitan areas.
4 Data and Methodology
4.1 Data source
The paper uses cross-sectional data from three data sources. These include;
1. Sub-national population, poverty headcount and inequality from the Ghana Statistical Service report (Ghana Statistical Service, 2015). Socio-economic indicators are drawn from the same institution based on the 2010 Ghana Population and Housing Census.
2. Resource base (total revenue and internally generated fund for 2013) at the sub-national level is drawn from the Ghana Audit Service report (2013) on 216 districts.
Table 2: Variable Description The poverty headcount (P0) is the proportion of population living below the national poverty line (GHC 1,114 per person per year) The proportion (actual values) of the total amount of revenue which is internally generated from local taxes, rents, fees and fines, investments, among others. This is also for the year 2013 This groups the 216 districts into their respective region of which there are 10. The total number of people resident in the districts under study Total amount of revenue for all source accrued by the district in the year 2013 A categorization under the three levels of sub-national divisions. District, municipality and metropolitan areas The distribution of the Gini coefficient, a measure of inequality, across districts
Source: The authors.
4.2 Estimation method
The estimation is a simple Ordinary Least Squares regression (OLS). This is used to clarify the extent to which the resource generating capacity of a sub-region (using IGF) explains the prevalence of sub-national poverty while controlling for population, government transfers (DACF), other funds (grants), status of the sub-national authority and inequality indicated in equation (1)
4.3 Data description
The data covers the 216 sub-national local government divisions in Ghana. Information on IGF and total grants were however short on 29 observations which were basically newly created district which had no compiled data as at the time of reporting (see Table 3).
Source: The authors.
There is a negative and significant correlation between poverty headcount (see Table 4) and all but two independent variable (Gini index and Grants) which are positively correlated.
Source: The authors.
5 Results and analysis
Using robust (2) and without robust (1) standard errors, the OLS estimation in Table 5 confirms that a disregard to revenue generating capacity in the creation of a district, reinforces poverty. A unit increase in IGF reduces poverty headcount by 3.3% which is significant in the two estimations.
Table 5: District resource base and the pertinence of poverty
Standard errors in parentheses (1) *** pRobust Standard errors in parentheses (2)
Source: The authors.
Population which is used as a basis for the creation of districts has a stronger negative correlation (12.2%) which confirms that a supportive revenue generation base should be a pre-requisite if the creation of districts can have a significant translation of inclusive governance to actual poverty reduction. Plotting IGF against poverty headcount weighted by population size, Figure 3 depicts that the higher the IGF of a district in 2013, the lower the pertaining poverty headcount. Source: The authors.
5.1 Trap 1: District poverty trap
Beyond the mean poverty headcount of all districts in Figure 3, there is a cluster made of mainly districts (4th quadrant) which are the smallest in terms of population and a corresponding low IGF base. This conforms to the district poverty trapping hypothesis (trap 1) where poverty remains highest and skewed towards districts. They are usually rural, and lack the resource base to undertake meaningful poverty reduction programmes. Population remains low with out-migration to municipal and metropolitan areas for better services and livelihood. Using a municipal dummy[5] with districts as a control, municipalities are significantly more likely (8.6 times as indicated in Table 5) to reduce poverty headcount than district because of their IGF capacity. Government targeted transfers could have been a cushioning but the DACF (a major source of central government transfer) though positively correlated with poverty headcount, it is insignificant in explaining it.Districts with IGF above the mean include all metropolitan areas and a predominant majority of municipal areas (1st quadrant). They have a high IGF as a resource base and lower poverty headcounts (below the mean) which conforms to the “Muni-Metro” poverty trap (trap 2). The high interaction between municipal areas and metropolitan areas terms of population flows blurs form distinction between the two. Both have a substantial resource base (IGF) that corresponds to low poverty headcounts (proportion of the total population poor) but in absolute terms, poverty is more pronounce. They host a bulk of the population which put pressure on existing facilities. Urban problems are more pronounce and rehabilitation of ever deterioration infrastructure services becomes an opportunity cost to poverty reduction programs. Poverty hence pertains.5.3 Other instances
From figure 3, there are limited observations (2nd and 3rd quadrant) where IGF is low (below the mean) yet poverty headcount remain low and in others where IGF is high yet poverty remains high. These are contrary to the first hypothesis of negative correlation. Considering that these are mainly district which are predominantly rural, an explanatory factor can be the level of inequality, other sources of funds and direct interventions that do not go through the district assembly establishment.The alternative case of high IGF yet high poverty (2nd quadrant) borders the extent to which expenditures from grants are pro-poor. An instance can be made of the Garu Tampene, a district in one of the poorest regions of Ghana (Upper East). It generated only GH₵ 190,859.52 (approximately 45,442.74 USD) funds internally but had external funds amounting to GH₵ 6,312,454.28 (approximately 1,502,965.30 USD). Yet, its poverty headcount stands at 54.5%. Its external revenue is higher than two metropolitan areas in Ghana (Cape Coast (GH₵ 1,876,156.53) and Tamale (GH₵ 6,620,193.00) whose corresponding poverty headcounts are 2.6 and 24.6 respectively.6 Conclusion
Creation of districts is vital in decentralisation. It opens up governance to the grassroots and facilitates development through inclusiveness. The idea of creating districts as part of the decentralisation reform in Ghana is justified on a frail veil of poverty reduction. There is no evidence to the claim of decentralisation reducing poverty in Ghana, yet it has been used to explain the 94.4% increase in districts created since 1988. Ayee (2013) put into perspective the political economy behind the creation of the districts. This paper looks into the poverty reduction basis of district’s creation and suggests that it is rather a trapping mechanism.Creation of districts without requisite consideration of capacity in generating funds for development reinforces poverty. The paper shows that districts with low IGF have a high rate of poverty and districts with a high IGF have lower rates of poverty. In this regard, districts (lowest demarcation level) in Ghana fall in a poverty cohort where there is a low IGF and a high rate of poverty headcount persist. On the other hand, metropolitan areas and municipal areas which have substantial IGF have a lower poverty headcount. They however contend with an influx of population which makes poverty in absolute terms higher. They form a second poverty trap which pertains to an opportunity cost of rehabilitating infrastructure other than poverty reduction interventions. External fund is an alternative to complement IGF for poverty reduction interventions. However the case of Ghana indicates that a unit increase in grants rather raises poverty headcount by 3.3%. A question thus lies in the extent to which the utilization of external funds (both international and national) are pro-poor.Though a panel data is required to make a conclusive statement, the preliminary evidence from the cross sectional data suggests that, the remedy to poverty reduction in Ghana does not lie in political and administrative fragmentation. Rather, emphasis should be placed on strengthening the institutional and bureaucratic quality of the already existing districts.Ahwoi, K., (2007) ‘Overview of local government system in Ghana: Prospects and challenges’. In Decentralization in Ghana. Papers presented at the Workshops held at GIMPA, Ghana from (Vol. 2010, pp. 44-62).Andrews, M. (2008) ‘The good governance agenda: Beyond indicators without theory’. Oxford Development Studies 36(4): 379-407.Awortwi, N. (2010) ‘The Past, Present, and Future of Decentralisation in Africa: A comparative case study of local government development trajectories of Ghana and Uganda’. International Journal of Public Administration 33(12-13): 620-634.Ayee, J. R. (2012) ‘The Political Economy of the Creation of Districts in Ghana’. Journal of Asian and African Studies 0(0):1–23Ayee, J., and Dickovick, J. T. (2010) ‘Comparative assessment of decentralization in Africa: Ghana Desk Study’. Washington, DC: United States Agency for International Development.Blöchliger, H. (2013) Decentralisation and Economic Growth – Part 1: How Fiscal Federalism Affects Long-Term Development, OECD Working Papers on Fiscal Federalism, No. 14, OECD Publishing.Cabral, L. (2011) Decentralisation in Africa: Scope, Motivations and Impact on Service Delivery and Poverty. Working Paper 020. Brighton, Sussex: Overseas Development Institute.Crawford, G. (2008) ‘Decentralization and the Limits to Poverty Reduction: Findings from Ghana’. Oxford Development Studies, 36(2), 235-258.Davis, J. R. (2006) Evaluating and Disseminating Experiences in Local Economic Development: Observations on integrated development programmes of the Free State, Republic of South Africa. Working paper Natural Resources Institute.[http://papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID944029_code459916.pdf?abstract id=944029&mirid=1 [accessed 12th March 2016]Government of Government (2016) Formula for Sharing the District Assemblies Common Fund.http://www.commonfund.gov.gh/index.php?option=com_content&view=article&id=337&Itemid=437. [Accessed 14th March 2016].Institute of Local Government Studies (2010) A Guide To District Assemblies In Ghana. ISBN 9988-572-13-1Friedrich-Ebert-Stiftung. GhanaJütting, J. P., Corsi, E. and Stockmayer, A. (2005) Decentralisation and Poverty Reduction. OECD Development Centre Policy Insight No. 5. Paris: OECD Development Centre.Kuffour, J.A. (2005) ‘The Role of Local Government in Decentralisation’, Presentation at Commonwealth Local Government Conference. Aberdeen, Scotland, UK, 15th March 2005.Martinez-Vazquez, J., and Rider, M. (2006) ‘Fiscal Decentralization and Economic Growth: A Comparative Study of China and India’. Indian Journal of Economics and Business, Special Issue: India & China: 29-46.Minstry of Local Government and Rural Development (2010) ‘Draft Decentralisation Policy Framework: Accelerating decentralisation and local governance for national development’. https://www.giz.de/en/downloads/en-national-decentralization-policy.pdf. [Accessed: 4th May, 2015]Ostrom, V., Bish, R. L., and Ostrom, E. (1988) Local Government in the United States. San Francisco: ICS Press.Ostrom, V., Tiebout, C. M., and Warren, R. (1961) ‘The Organization Of Government In Metropolitan Areas: A Theoretical Inquiry’ American Political Science Review, 55(04), 831-842.Republic of Ghana (1992) Constitution of the Republic of Ghana, 1992. Accra: Assembly Press, Ghana Publishing Corporation.Robinson, M. (2003) ‘Participation, Local Governance and Decentralised Service Delivery’. In Workshop on New Approaches to Decentralized Service Delivery, Santiago, Chile, March 2003.Scott, Z. (2009) Decentralisation, Local Development and Social Cohesion: An Analytical Review. University of Birmingham Governance and Social Development Resource Centre (GSDRC) Research Paper.Smoke, P. (2003) ‘Decentralisation in Africa: Goals, Dimensions, Myths and Challenges’. Public Administration and Development, 23(1), 7-16.Tiebout, C. M. (1956) ‘A Pure Theory of Local Expenditures’. The Journal of Political Economy, 64(5) 416-424.Visser, J. A. (2002) ‘Understanding Local Government Cooperation in Urban Regions: Toward a Cultural Model of Interlocal Relations’. The American Review of Public Administration, 32(1), 40-65.Von Braun, J. and Grote, U., 2000 ‘Does decentralization serve the poor?’ In International Monetary Fund Conference on Fiscal Decentralization, Washington, DC, November 2000 (pp. 20-21). [1] Decentralisation is defined as a mode of governance which involves the devolution of decision-making powers and development responsibilities to the lowest unit of government (Von Braun and Grote 2000 citing Litvac 1999).[2] Exception is made on Ayee’s paper on the political economy of the creation of districts in Ghana.[3] Visser (2002) summarizes the reform- consolidation model as the prior alternative to the fragmented model which stipulates a unified action under a single authority.[4] See Ayee (2013 p.628) for the architecture of decentralisation in Ghana.[5] The alternative dummy for metropolitan areas is insignificant. Metropolitan areas however only forms 2.7% of the total number of districts in Ghana.Respectively, PhD fellow in Economics and Governance, United Nations University – Maastricht Economic and Social Research Institute on Innovation and Technology (UNU-MERIT ([email protected])) and PhD fellow in Economics and Governance, United Nations University and Maastricht Economic and Social Research Institute on Innovation and Technology, and IMANI Africa (UNU-MERIT ([email protected])). The views expressed, and any errors, are entirely theirs.