Richard C. Crook
Abstract
It is generally assumed by advocates of democratic decentralisation that maximising locally-raised revenue sources will help to enhance accountability through a closer and more legitimate relationship between the local government authority and citizens. Research on Ghana and other African countries shows, however, that the dominance of clientelist forms of politics undermines the legitimacy of local taxation; where voters expect their representatives to provide specific pay-backs to themselves or their communities in return for support and payment of tax, it is extremely difficult to establish a 'collective interest' for the local government area. Citizens tend to interpret allocation decisions over expenditure of revenues as products of patronage relations rather than as a collective public good. This problem is especially acute where resources are very scarce and the revenue base limited. Central government transfers are, therefore, likely to remain the fairest and most effective way of financing local government in such contexts.
Keywords: democratic decentralisation, Ghana, clientelism, taxation, local revenue, intergovernmental transfers.
1 Introduction: accountability and central vs local revenue sources
One of the most commonly recognised problems of local or decentralised government in less developed countries is lack of resources to carry out their basic functions and services – the so-called 'unfunded mandates' or 'responsibility-resource gap' (Dafflon 2013). How to fund decentralised government in poor or developing countries is the subject of a continuing debate between those who argue that decentralisation should seek to maximise locally-raised revenue, and those who argue that central government transfers are the only realistic and equitable way to ensure adequate resources, even if those transfers become the dominant source of revenues.1
The main arguments in favour of central government funding are:
The main argument for central funding therefore rests on the assertion that the source of funding for local government is less important than its adequacy for the functions allocated, and how it is controlled and monitored, including the role of political accountability mechanisms at local level.
Those in favour of maximising local revenue mobilisation base their arguments primarily on the desirability of enhanced accountability of government, linked with democratic decentralisation. The main assumptions are as follows:
In this article, some of the assumptions underlying the advocacy of maximising local revenue mobilisation will be challenged, focusing particularly on the alleged link between democratic representation, greater willingness to pay local taxes, and improvements in the provision of public goods including investment in pro-poor development programmes. The key institution linking government and citizens in this context is that of the elected representative. To what extent does accountability through elections really ensure that local governments will be able to improve local tax mobilisation, and produce more developmental outputs? And what do citizens in today's poor or developing countries actually demand of their local representatives? The 'principal-agent' models popular with development analysts and donor agencies tend to assume that citizens as 'principals' will demand collective or 'developmental' public goods and accountability for expenditure decisions. But this is not necessarily the case (cf. Booth 2012).
These questions will be examined through a review of the experience of democratic decentralisation in Ghana, with some comparative observations on other African and Asian countries, including a rather different form of decentralisation in Côte d'Ivoire based on a Francophone model of local government. These cases show that even after many years of consolidated democratic local government, there are still enormous problems with local tax mobilisation, and with the developmental responsiveness of the local authorities. The problems derive primarily from the role of representatives in societies dominated by a patron–client form of political competition, and also from the logic of the particular forms of electoral representation which have been adopted.
2 The Ghana District Assemblies
2.1 Formal structures and functions
Ghana's District Assembly system was introduced in 1989 by the then military government of Jerry Rawlings. The 'Rawlings Revolution' of 1981 was based on radical populist ideas of direct participation and no-party people's democracy (inspired partly by Gaddafi's Libya), and the decentralisation reform of 1989 was portrayed as a fulfilment of that commitment to introduce a 'truly Ghanaian' form of grass-roots democracy. With the transition to a more conventional, representative multiparty democracy in 1992, the District Assemblies (DAs) were incorporated into the 1992 Constitution of the Fourth Republic and are now governed by the Local Government Act of 1993. Politically, the reform was clearly an attempt to create a rural power base for Rawlings, embodying as it did a privileged position for pro-Rawlings 'revolutionary organisations' (which later became the core of his party, the National Democratic Congress (NDC)), and mechanisms for co-opting rural business, professional and agrarian elites (see Crook and Manor 1998).
There are now 216 Assemblies, ranging from 170 rural DAs with populations ranging from 75,000 to 250,000–300,000, to 40 Municipal Assemblies for towns with populations of over 95,000, and six Metropolitan Assemblies for the large cities including the capital Accra, population 2.4 million. Their primary function is to be responsible for the 'overall development of the district', together with a very wide range of specific service responsibilities. In the rural DAs, these are mainly the provision of educational and health infrastructures, minor roads, public works, markets, town planning and other projects.
Administratively, they are a mixed form of decentralisation in that their other main function is to supervise the various former line ministries which have been deconcentrated to district level since 1989, to form an integrated district governmental and financial apparatus. They are therefore responsible for drawing up an overall District Plan and a composite District Budget, although this remains problematic even after 27 years of operation. The district is administered, not by an elected mayor, but by a centrally appointed official, the District Chief Executive (DCE), who is normally a political activist loyal to the government of the day, assisted by a small group of senior civil service administrators and technical officers. Other staff (e.g. revenue collectors, drivers, clerks, labourers) are employed locally.
2.2 DA funding
Since 1994, the majority of funding for the DAs (currently around 82 per cent overall) has come from central government transfers (Gilbert, Hugounenq and Vaillancourt 2013); this represents a continuing trend towards increases in both the absolute amounts and the proportion of funding coming from central government or external sources, which, as will be argued, is linked to the political imperatives of democratic decentralisation. The main funding source is the District Assemblies Common Fund (DACF), a distributable pool account which is allocated 7.5 per cent of national government revenues; other elements come from the District Development Fund (DDF) (a channel for donor funds), Highly Indebted Poor Countries (HIPC) funds, royalties from natural resources and customary land rents, and government payment of some district staff salaries. The largest source of local revenues is 'fees and charges' which includes local business taxes, followed by licences and property tax (Table 1).
2.3 Democracy and the political context
Rawlings' original populist vision of the DAs was that they would build a genuinely grass-roots form of participatory democracy which would focus on 'community-based forms of self-reliant economic and social development' (or 'self-help', as it is called in Ghana). Rawlings was especially hostile to liberal, multiparty-based competitive politics, which he saw as an inappropriate, colonial imposition on African societies, and a harmful distraction from building communal consensus around practical development measures. Hence the ban on parties in local elections, which not only echoed the Libyan system, but was similar to the Ugandan system introduced by Museveni's ex-guerrilla army National Resistance Movement (NRM) regime in 1986 (Crook 1999). (A similar approach has characterised other authoritarian regimes which have resisted political competition at the national level, such as Bangladesh under Ershad, or Pakistan under Musharraf).
These participatory aspirations were embodied in the democratic institutions of the DAs at both district and sub-district levels. At the lowest level, there are elected Unit Committees for every group of 500 electors (in practice, villages and quarters); above them are Area and Town Councils, not directly elected but composed of DA-elected members for those areas.
At the district level, which because of the size of the DAs has necessarily to be a form of representative politics, there are Assemblies which consist of two thirds-elected Assembly Members (AMs), elected by 'first-past-the-post' elections as single members for very small wards. The average ratio of AMs to electors is 1: 2,000; in the two case study districts studied by Crook and Manor in the 1990s (East Akim and East Mamprusi), the ratios were 1:1,800 and 1:2,200 (Crook and Manor 1998). This was clearly designed to foster a very close and direct relationship between elected AMs and their constituents. Together with the one third of AMs appointed by government on local recommendations, this has produced relatively large Assemblies, for example 79 members for East Akim and 69 for East Mamprusi, with populations in the 1990s of 209,710 and 183,839 respectively.
The original political context of the DAs changed substantially after 1992, following Rawlings' acceptance of the need to move to a multiparty, constitutional democracy with an elected executive president and national Parliament. Rawlings secured his continuation in power for another eight years through his NDC party, which used his incumbent power to mobilise support; control of the resources of the rural DAs through his loyal DCEs and his Committees for the Defence of the Revolution (CDRs) played a significant part in the election victories of 1992 and 1996. After the opposition, the New Patriotic Party (NPP), won power in 2000, Ghana became established as a relatively stable two-party competitive democracy, with the NDC returning to power in 2008. The two main parties represent two competing 'political traditions' associated with both class and regional power bases; the NPP is descended from the conservative nationalist movement of the 1950s, which had its main power bases in the southern Akan regions of Eastern and Ashanti, and is based on traditional elites and the established business and professional classes. The NDC grew out of Rawlings' more radical, anti-establishment movement which looks back to the Nkrumah era of state socialism and is especially strong in the northern and Volta (south-eastern) regions.
Nevertheless, in spite of these broad historic differences, both parties have adapted to multiparty democracy since 1992 in similar ways, to construct what Whitfield has called a system of 'competitive clientelism' (Whitfield 2011). This system is built on the 'mercantilist' nature of the economy. Ghana's economy still depends mainly on the export of primary commodities such as gold, timber and cocoa grown by small/middle farmers, and more recently, crude oil. The state's revenues are drawn from the rents it extracts from the export/ import economy, and the domestic business classes are relatively weak and small. Whichever party controls power uses access to state contracts and other opportunities to reward its supporters, creating a dependent or client business class, those who have 'invested' in the party's victory. Competition for power is a zero-sum game for those who depend on political patronage. And winning elections is therefore an essential prerequisite for sustaining these patronage machines. Parties, which are quite weakly organised and factionalised, also depend on regional and local elites to invest in securing election victories, and are expected to reward victorious MPs as well as the party's 'foot soldiers' at constituency level after an election win.
Most importantly, these expectations of the purpose of elections and the use of political power are generally shared by voters. One study of MPs in competitive and non-competitive constituencies found that 70 per cent of voters voted for their preferred candidate on the expectation that they would be able to provide specific local development projects or job opportunities (so-called 'club goods'); 15 per cent expected personal benefits to be given out during the election campaign, and further rewards for particular forms of support. (These expectations were higher in the more competitive constituencies; Lindberg 2003, 2010.) To be successful (and re-elected), MPs therefore have to cultivate clientelist links with higher levels of the party and government ministers if they are to deliver on their patronage promises; they also need to generate pay-back for their 'investment', through kick-backs on brokering access to government contracts and other useful opportunities. Voters themselves (especially 'swing' voters) therefore vote on clientelist principles, rather than choosing parties or representatives who will provide more programmatic public goods (cf. Keefer and Khemani 2005; Levy and Kpundeh 2004; van de Walle 2012; Vicente and Wantchekon 2009). It is the clientelist basis of relations between voters and their representatives, and between citizens and government authorities, which seriously undermines the notion that democracy will produce governments which are more directly accountable to citizens for the way they spend public money, or that will spend more on redistributive developmental policies, even if the funds are the product of general taxation rather than 'rents'.2
2.4 Representative democracy and clientelism at the local level
The clientelist assumptions of voters are reproduced at the local level in the politics of the DAs. Indeed, they have been officially acknowledged in the 6 per cent share of the DACF which was allocated to MPs to be spent on 'Constituency Development', or projects they choose for their constituencies as a way of fulfilling constituents' expectations (Lindberg 2009). MPs are also included in the DA as ex officio members, and can therefore bring their political concerns to bear on DA planning and spending.
In the election of AMs, the no-party rule and the very close, grass-roots character of the representation offered by AMs has strongly reinforced the idea that the Assembly representative is a delegate of his or her specific community – a ward of one or two villages, with only a few thousand electors. The rules for the election of an AM limit the number of candidates to five; they must be resident in the district (to exclude absentee urban elites), and have to stand on their 'personal merit, integrity and proven participation in community development' (Crook 1999). The government then organises and funds a three-week period of public hustings at which the candidates present their 'life histories' and manifestos, and take questions from the public. Once elected, one of their official duties is to encourage and develop 'self-help' projects in their communities. And they have a duty to maintain regular contact with their constituents. The record of the elected AMs in developing good participatory relationships was in fact quite good for the first few years of the new system. In the districts studied by Crook and Manor in 1992, 3.83 per cent of electors could name their AM, 32 per cent said they had attended a meeting with their AM, and 12 per cent had contacted their AM individually, which in comparative terms is quite impressive. The first elections for the DAs were also met with enthusiasm, as reflected in the 58 per cent turnout (Crook and Manor 1998).
The pressure on such representatives to fight for projects for their own areas, and assistance with their self-help projects is clearly very strong. The DAs have therefore become arenas in which there is little debate over development policy for the district as a whole; instead, they are battlegrounds over the allocation of whatever resources are available to the communities represented by each of the AMs. In a situation of extreme scarcity and inadequate revenue sources, only a very few of the hundreds of villages or towns within a district can be offered any development project in any single year.3 And in a political system dominated by patron–client linkages, whatever decision is taken will be interpreted (whether rightly or wrongly) by the unlucky ones as a consequence of the greater 'influence' or patronage connection of that community's representative; maybe he or she is close to the DCE, the MP and the ruling party in the district, or other influential government officials and 'big men' in the area. (Similar problems have been reported from Uganda: in one study of local primary health-care services, citizens did not bother to complain about poor service, or use 'participatory' user complaint mechanisms because they believed that local councillors were involved in corrupt patronage relationships with clinic staff, and therefore complaints would never be listened to fairly (Golooba-Mutebi 2005).) With the growing intensity of party competition after 2000, it also become more plausible to suspect that communities which were known not to have voted for the ruling party in Parliamentary elections were being discriminated against in development expenditure decisions.
2.5 Community representation and the legitimacy of local taxation
One of the consequences of such a system, dominated by strong community representation and clientelist politics, is that it undermines the legitimacy of local taxation, and makes it increasingly difficult to improve local revenue mobilisation. Where the fairness of district allocations of expenditures is constantly challenged, it is difficult to create a sense of the 'collective interest' of the district, and hence acceptance of the idea that taxation is for the general good. It is here that clientelist politics has its most damaging effect, insofar as it makes the link between democratic representation and acceptance of local taxation difficult to sustain, and therefore undermines any move to reduce dependence on central government transfers. As Luttmer and Singhal have shown, 'tax morale', or intrinsic motivations to comply with tax demands based on factors such as citizens' relationship to the state, and ideals of reciprocity, is an important element in the success of any taxation system (Luttmer and Singhal 2014). In the Ghanaian DAs, each community – indeed each individual voter – wants to pay the equivalent of a 'hypothecated' tax in which the community will get back what it paid as a specific service or project for their particular community. They do not want to fund benefits for other communities. This attitude was actually reinforced by a government decision to introduce 'local rebate' schemes, according to which a proportion of locally collected taxes was returned to the Area or Zonal Councils in which they were collected on a pro rata basis. The problem was also exacerbated in the early years of the Assemblies by the contradictory official roles of the AMs.
On the one hand, they were supposed to encourage payment of local direct taxes (flat levies or 'poll taxes', market and lorry park tolls, property taxes) which were justified in terms of the necessity for collective development resources; citizens were told that development progress depended upon them paying their taxes. In the case study districts, AMs took part in 'pay your tax' campaigns alongside Rawlings' CDR militants, who used increasingly heavy-handed methods in local markets. In Africa, both the association of direct taxation with colonialism, and the poor past performance of local government made it especially difficult to persuade citizens of the 'rightness' – and benefits – of paying their taxes. Yet AMs ended up being accused by the government officials and CDR militants of being 'saboteurs' of government policy when tax collection did not meet targets (Crook and Manor 1998: 221). When the promises made by AMs that paying taxes would lead to funding of local projects were routinely unfulfilled, encouraging tax payment became even more difficult.
On the other hand, AMs were supposed to lead local, self-help-based community development efforts as a way to bolster their own popularity and legitimacy. They knew they would ultimately be judged on their success in 'bringing development' to their wards. For citizens, this could be seen as a double jeopardy; they were told they had to pay their taxes in order to sustain the development programme of the local government. But when they requested specific projects they were told they had to organise their own self-help, which could also involve paying locally agreed project levies. This led AMs to increasingly withdraw from encouraging payment of tax, and instead to focus mainly on self-help projects, with whatever help they could get from other patrons.4
The longer-term consequences for the DAs of these interactions between intensely local community representation and shortage of resources have been on balance very negative.
2.6 Decline in participation and increasing distrust of District Assemblies
In the first place, disillusionment on the part of both citizens and representatives rapidly set in. In spite of the good initial improvements in participation, by 1992, 3.71 per cent of respondents in the case study districts felt that the Assembly 'could not meet their needs', a view amply proven by comparing the actual outputs of the Assemblies with citizens' statements on what they saw as the main developmental priorities for their communities (Crook and Manor 1998). Twenty-two years later, little had changed; a national survey by Ghana's Center for Democratic Development (CDD) found that 65 per cent of citizens did not trust the DAs and gave poor ratings to their performance in key areas of public service provision (Ayee, Ahwoi and Deku 2014: 152). Participation in local elections declined, from 58 per cent in 1989 to 41 per cent in 1998, and has continued to fall since (Crook 1999). Worse, the most recent assessments have confirmed that the sub-district structures – Town, Area and Zonal Councils, and Unit Committees – which were supposed to be the core of grass-roots participation – remain for the most part 'paper' institutions which have never really been activated (Ayee et al. 2014: 71). When elections for the Unit Committees finally took place in 1998, 65 per cent were uncontested. The initial cohorts of elected AMs also lost heart: by 1994, only one third on average of the original AMs stood for re-election, and it has since proven consistently difficult, except in the main urban areas, to secure good candidates for the Assemblies.
2.7 Political pressure to increase central government funding, decline in local revenue mobilisation
Secondly, there has been little or no progress in improving local revenue mobilisation. The political and administrative difficulties outlined previously, combined with the increasingly powerful clientelist basis of local and national politics, have led to continual increases in central government transfers to the local government system, both in absolute and proportionate terms. Before the DA reform, Ghana's District Councils, which were administratively run, received between 24 and 43 per cent of their revenues from central government transfers. In the first few years of the DA system, when central government transfers came primarily from ceded revenues, salary payments and earmarked grants, central government funding continued to account for around 40–50 per cent of total revenues (although this actually declined somewhat in the two case study districts) (Crook and Manor 1998). A major change occurred with the introduction of the DACF in 1994; initially set at 5 per cent of national revenues, political pressure led to the fund being increased to 7.5 per cent in 2008 after the election of the NPP government (Gilbert et al. 2013). The NDC regime (which lost office in December 2016) was considering increasing this to 10 per cent. Over the past 20 years, funding for the DAs coming from the DACF has resulted in a tenfold increase in central funds compared to the former ceded revenues system. The increases have been especially significant over the 2008–13 period (see Table 2). In 2008, total central government and external funding accounted for between 74 per cent (largest quintile by population) and 93 per cent (smallest quintile by population) of total revenues, the differences reflecting the equalisation formulae (Gilbert et al. 2013). (The three big cities, Accra, Kumasi and Tema, received 59 per cent, 68 per cent and 47 per cent respectively from central government transfers.) Latest estimates suggest that the proportions are between 80 and 99 per cent (Ayee 2014).
As noted, this significant and continuing growth in central funding has raised the average annual per capita revenue for all DAs to between US$5 and US$10 – still not very impressive, given the responsibilities of the district and Metropolitan governments. At the same time, local revenue mobilisation continues to experience the same difficulties and remains weak. One indication is the persistence of unrealistic budget estimating, whereby every year DAs set budgets based upon estimates of revenue collection which are never achieved, often by huge margins. In the two case studies, the gap between revenue estimates and actual revenues in the early 1990s was routinely between 200 and 400 per cent. In 2014, it was calculated that the actual revenues of all Assemblies were only 37 per cent of budget estimates (Ayee et al. 2014).
An example of the difficulties is to be found in the poll tax or 'flat basic rate', a legacy from the colonial era which Assemblies have retained and still attempt to collect. It is usually a very small, even token amount (Ayee suggests it is typically the equivalent of £0.10), but collection rates are very low – in the case studies, the revenue figures showed that only around 12–14 per cent of the adult population were paying the tax. Nevertheless, an official report on the performance of the decentralisation programme still recommends that the tax be retained, but at a more useful level (for example, the equivalent of one day of the official minimum daily wage), on the grounds that it provides an accountability link between the provision of services and payment of tax (Ayee et al. 2014). Clearly, the fundamental political difficulties at the root of local revenue mobilisation have not been addressed, given the context of ever-increasing pressures to provide central funding. Such a centrally-dominated system can be seen as an important element in the control of patronage which permeates the democratic political system, side by side with the laudable goal of providing realistic and adequate funding for the decentralised authorities.
3 Would a different kind of representation and electoral system work better?
If the community basis of representation on large district governments creates such deep problems, would a different form of democratic representation work better? Côte d'Ivoire introduced a 'Francophone'- style local government system between 1985 and 1996, based on 'communes'. The distinguishing feature of the commune system is that it is based on an urban conception of local government, in which each town is made a self-governing unit, incorporating only a small surrounding rural area. By 1996, Côte d'Ivoire had 196 communes, the majority of which had populations of less than 20,000. Each is run by an executive mayor who is elected by the winning council at its first meeting. A closed list system of election is used which means that the councillors do not officially have any constituency responsibility or 'community' link. But the ratio of representation is even closer and more 'grass roots' than that of Ghana: 1:1,000 on average, with ratios of 1:400 more common in the smaller communes such as the four studied in the 1990s (Crook and Manor 1998).
Unfortunately, the Ivorian list system does not work any better than single-member wards for encouraging local democratic responsibility and collective action including local revenue mobilisation for development. A survey of local residents revealed that 58 per cent could not name any elected councillors, whilst 90 per cent could name the mayor (Crook and Manor 1998). The political logic of the list means that most mayors are elite political entrepreneurs who put together their list before the election; they are his supporters and if the list is victorious will elect the patron mayor as expected. Worse, the list election is a 'winner takes all', so the winning list takes control of the commune council and there is no representation of any opposition. This means that collecting local taxes and even special levies for agreed commune projects is difficult because those who did not vote for the winning list and its mayor do not consider themselves in any way obliged to contribute. In one commune studied, a project to build a much-needed secondary school for the area had to be abandoned as it was impossible to collect sufficient funds through a commune special levy, even though the prefect was enlisted to try to enforce it. The mindset of Ivorian local electors is very much focused on central or national politics; they see 'development' as something which the mayor, as a big 'patron', should bring to the locality through his political connections with capital city politicians and officials.
4 Conclusions
The issues emerging from the experiences of decentralised government in Ghana, Côte d'Ivoire and Uganda, as well as Africa more generally, suggest that representative local democracy does not necessarily lead to improved local revenue mobilisation, particularly where it involves any form of direct or individual taxation. Indirect taxation is probably easier to collect but still does not deal with the problem of how to give greater legitimacy to resource allocation decisions by an elected council, where there is little or no sense of a collective public good at that level of action. Neither single-member electoral representation of communities on a 'no‑party' basis, nor party list elections seem to work well, within the context of a national political system dominated by competitive clientelism.
Competitive party competition at local level could perhaps help to aggregate a broader sense of collective interest, but can also lead to an even stronger sense of exclusion and resentment if resources are denied to communities which voted for the losing party, as happens in many countries. In India, which has a nationwide system of local democratic governments, what does seem to work better is the presence of real opposition, which can challenge the most blatant examples of clientelist allocation of resources; this, combined with the very high rate of turnover in Indian elections (ruling parties at state level were thrown out after one term in over 70 per cent of elections, 1980–2008) has led to the emergence of 'post-clientelist' initiatives in which politicians try to appeal to broader groups of voters with generic development programmes (Manor 2016). In India, such approaches are further encouraged by the enormous increase in national funding resources handed down to local governments by anti-poverty programmes such as the National Education Guarantee Programme and the National Rural Employment Guarantee. These initiatives can of course feed local patronage politics; but active democratic competition and popular accountability initiatives can help to make them into tools of broader developmental initiatives. There is some evidence that this may be starting to happen in parts of Africa as well (van de Walle 2012); but the continuing dominance of clientelist politics means that it is still very difficult to establish the legitimacy of local taxation. It may be argued, therefore, that attempting to increase the proportion of local government revenues deriving from local taxation faces many obstacles, and is not necessarily helped by the current form of democratic electoral politics in Ghana. Central government transfers are, therefore, likely to remain the fairest and most effective way of funding local government in Africa for some time to come.
Notes
* This article uses material from research on local government in West Africa and South Asia funded by the British Overseas Development Administration (now Department for International Development, DFID). The views expressed are those of the author and should not be attributed to DFID.
1 See Cabral (2011) for a useful summary of the main arguments for (and against) democratic decentralisation more generally.
2 Some scholars nonetheless remain puzzled that democracy has not produced more redistributive policies; Ardanaz and Scartascini's (2011) study of Argentine fiscal federalism explains it in terms of the 'overrepresentation' of richer districts in the national parliament. But this simply transfers the same argument to parliamentary representation, on the assumption that government policy will somehow directly reflect the interests represented in parliament. But there is little evidence that parliaments in either South American or African democracies have much influence on policymaking or policy implementation (see Barkan 2010).
3 In the early 1990s, the per capita development expenditure in East Akim and East Mamprusi districts was the equivalent of US$0.18 and US$0.45; even by 2008, after substantial increases in central funding, the average per capita revenue of all the Assemblies was only US$5.31, and US$10.62 for the smallest/poorest quintile (2012 exchange rates) (Crook and Manor 1998; Gilbert et al. 2013).
4 In one of the case study districts, respondents identified 66 per cent of the projects which had been started in their villages as 'self-help'; but 71 per cent were dissatisfied with them because few ever got completed (Crook and Manor 1998: 255).
References
Ardanaz, M. and Scartascini, C. (2011) Why Don't We Tax the Rich?
Inequality, Legislative Malapportionment, and Personal Income Taxation
around the World, Working Paper IDB-WP-282, Washington DC:
Inter-American Development Bank
Ayee, Joseph R.A. (2014) 'Socio-Economic Impact of Gold Mining on
Local Communities in Africa: The Ghana Case Study', consultancy
report for World Bank, unpublished, Accra
Ayee, Joseph R.A.; Ahwoi, K. and Deku, K. (2014) 'Review of
Decentralization Reforms in Ghana: Aide Memoire of the Joint
Decentralization Review Mission', report commissioned by the
Government of Ghana Inter-Ministerial Coordinating Committee
(IMCC) on Decentralization and the Development Partners,
unpublished, Accra
Aziz, A. and Arnold, D.D. (eds) (1996) Decentralised Governance in Asian
Countries, Thousand Oaks CA: Sage
Bahl, R. and Linn, J. (1994) 'Fiscal Decentralization and Intergovernmental
Transfers in Less Developed Countries', Publius – The Journal of Federalism 24.1: 1–19
Barkan, J. (2010) 'Legislatures on the Rise?', in Larry Diamond and
Marc F. Plattner (eds), Democratization in Africa: Progress and Retreat,
Baltimore MA: The Johns Hopkins Press
Bird, R. (1994) 'Decentralizing Infrastructure: For Good or For Ill',
background paper for World Development Report 1994, Washington DC:
World Bank
Booth, D. (2012) Development as a Collective Action Problem: Addressing the
Real Challenges of African Governance, London: Overseas Development
Institute (ODI), Africa Power and Politics Programme,
(accessed 17 January 2017)
Brautigam, D.; Fjeldstad, O-H. and Moore, M. (eds) (2008) Taxation
and State Building in Developing Countries, New York NY: Cambridge
University Press
Cabral, L. (2011) Decentralisation in Africa: Scope, Motivations and Impact on
Service Delivery and Poverty, Future Agricultures Working Paper 020,
London: Overseas Development Institute (ODI)
Crook, R.C. (1999) 'No-Party Politics and Local Democracy in
Africa: Rawlings' Ghana and the Ugandan Model in the 1990s',
Democratization 6.4:114–38
Crook, R.C. and Manor, J. (1998) Democracy and Decentralisation in
South Asia and West Africa: Participation, Accountability and Performance,
Cambridge: Cambridge University Press
Dafflon, B. (2013) 'Institutional and Fiscal Decentralization: Blueprint
for an Analytical Guide', in Bernard Dafflon and Thierry Madiès
(eds), The Political Economy of Decentralization in Africa, African
Development Forum Series, Washington DC: World Bank
Davey, K.J. (1994) 'Local Resource Generation: Role and Potential', in Donald Rothchild (ed.), Strengthening African Local Initiative: Local Self-Governance, Decentralisation and Accountability, Hamburg: Institut für Afrika Kunde
Gervasoni, C. (2010) 'A Rentier Theory of Subnational Regimes: Fiscal
Federalism, Democracy, and Authoritarianism in the Argentine
Provinces', World Politics 62.02: 302–40
Gilbert, G.; Hugounenq, R. and Vaillancourt, F. (2013) 'Local Public
Finances in Ghana', in Bernard Dafflon and Thierry Madiès (eds),
The Political Economy of Decentralization in Africa, African Development
Forum Series, Washington DC: World Bank
Golooba-Mutebi, F. (2005) 'When Popular Participation Won't Improve
Service Provision: Primary Health Care in Uganda', Development
Policy Review 23.2: 165–84
Keefer, Philip and Khemani, Stuti (2005) 'Democracy, Public
Expenditures and the Poor: Understanding Political Incentives for
Providing Public Services', World Bank Research Observer 20.1: 1–27
Levy, B. and Kpundeh, S. (eds) (2004) Building State Capacity in Africa:
New Approaches, Emerging Lessons, Washington DC: World Bank
Lindberg, S. (2010) Some Evidence on the Demand Side of Private–Public
Goods Provision by MPs, Working Paper 8, London: Overseas
Development Institute (ODI), Africa Power and Politics Programme,
(accessed 17 January 2017)
Lindberg, S. (2009) Member of the Parliament of Ghana: A Hybrid Institution
with Mixed Effects, Working Paper 2, London: Overseas Development
Institute (ODI), Africa Power and Politics Programme, (accessed 17 January 2017)
Lindberg, S. (2003) ' "It's Our Time to 'Chop' ": Do Elections in Africa
Feed Neo-Patrimonialism Rather than Counteract it?', Democratization 10.2: 121–40
Luttmer, E.F.P. and Singhal, M. (2014) 'Tax Morale', Journal of Economic
Perspectives 28.4: 149–68
Manor, James (2016) 'Beyond Clientelism: Digvijay Singh in Madhya
Pradesh', in James Manor, The Writings of James Manor: Politics and
State–Society Relations in India, Hyderabad: Orient Black Swan
Tanzi, V. (1995) 'Fiscal Federalism and Decentralisation: A Review of
Some Efficiency and Macro-Economic Aspects', paper presented at
the Annual World Bank Conference on Development Economics,
May, Washington DC
van de Walle, N. (2012) 'The Path from Neopatrimonialism: Democracy
and Clientelism in Africa Today', in Daniel C. Bach and M. Gazibo
(eds), Neopatrimonialism in Africa and Beyond, Abingdon and New York
NY: Routledge
Vicente, P.C. and Wantchekon, L. (2009) 'Clientelism and Vote Buying:
Lessons from Field Experiments in African Elections', Oxford Review
of Economic Policy 25.2: 292–305
Whitfield, Lindsay (2011) Competitive Clientelism, Easy Financing and Weak
Capitalists: The Contemporary Political Settlement in Ghana, DIIS/EPP
Working Paper 2011, Copenhagen: Danish Institute for International
Studies: 27
World Bank (2004) World Development Report 2004: Making Services Work for Poor People, Washington DC: World Bank
© 2017 The Author. IDS Bulletin © Institute of Development Studies | DOI: 10.19088/1968-2017.114
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non Commercial 4.0
International licence, which permits downloading and sharing provided the original authors and source are credited – but
the work is not used for commercial purposes. http://creativecommons.org/licenses/by-nc/4.0/legalcode
The IDS Bulletin is published by Institute of Development Studies, Library Road, Brighton BN1 9RE, UK
This article is part of IDS Bulletin Vol. 48 No. 2 March 2017: 'Interrogating Decentralisation in Africa'; the Introduction is also recommended reading.