The Enigma of the Central–Local Government Relationship and Its Impact on Property Tax Administration in Developing Countries: The Ghanaian Perspective

Summary Property tax administration is the bedrock for effective revenue mobilisation, development, and good local governance for local governments. Yet administering property taxation continues to be a major problem, especially for many developing countries. Scholarly explanations for this poor state of affairs have focused on limited capacity, poor quality local cadastres, corruption, and local political resistance to effective property tax administration, among others. This paper moves away from these explanations to focus on a less trodden area: the relationship between central and local government and how this relationship affects property tax administration. Property tax administration involves some collaboration and overlap between different levels of government, and thus depends very much on a good and functional relationship between both levels of government, especially when local governments derive their authorities from the largesse of central governments. This relationship may have powerful implications for the ability of local governments to effectively undertake property tax administration due to the central government’s policies and politics. Using Ghana as a case study, the paper illustrates how a dysfunctional relationship between central and local governments has undermined, and continues to undermine, effective property tax administration in the country, which should serve as a lesson for other developing countries.


Introduction
Property tax is seen as one of the best ways to enhance local government revenue for human development (World Bank 2019) and has been the subject of both increased policy attention and an expanding body of research (Franzsen and McCluskey 2017;McCluskey and Franzsen 2016). With the dwindling of both central governments' transfers and external resources in developing countries, focus has gradually shifted to effective administration of property tax as the next most viable source of revenue for local authorities in these countries (Ali, Fjeldstad and Katera 2017;Martinez-Vazquez and Bird 2014). Property taxation is an ideal tax for local governments because property is immovable and 'to the extent that there is a visible connection between the types of services funded at the local level and the benefit to property values, the accountability of local governments to local residents may be substantially improved' (Slack 2013: 142). Yet despite growing attention, property tax administration (PTA) is almost always poorly structured and administered (Bahl and Bird 2018). PTA is defined in this paper as 'a process of identifying and registering properties, assessing their value for taxation purposes, defining the tax rate, issuing billing, collecting payments and enforcement and implementing sensitisation and other citizen engagement activities to induce tax compliance by tax authorities (central and local governments).' In several countries, PTA is often a shared responsibility between central and local governments with local governments having the power to collect this tax in decentralised systems. Nevertheless, as property tax administration is poorly structured, 'property tax revenues continue to be less than 1 per cent of GDP and less than 4 per cent of tax revenues in most developing countries. Despite the mountain of literature on this subject… property tax is still, for the most part, a non-starter' (Bahl and Bird 2018: 227). This appears to be particularly the case in Africa, where property taxes generate only 0.38 per cent of GDP on average across a sample of 32 countriesand an even smaller share of recurrent property taxes (Franzsen and McCluskey 2017). What accounts for the weak property tax administration in developing countries, and Ghana specifically?
In explaining these problems, there are a host of possible explanations, including limited capacity, poor quality local cadastres, corruption, and local political resistance to effective PTA, among others (Jibao and Prichard 2015;Kelly, White and Anand 2020). Relatively overlooked, however, has been the relationship between local and central governments, which may have powerful implications for the ability of local governments to effectively undertake effective PTA due to policies and politics (Fjeldstad, Ali and Katera 2019;Bahl and Bird 2018). At the extreme, this has been reflected in tugs of war between central and local authorities over who controls property tax collection, the resulting revenue (Fjeldstad et al. 2019), and how best to support capacity building for local governments (LGs).
PTA inevitably involves some collaboration and overlap between different levels of government, and thus depends very much on the relationship between both levels of government. However, many central governments in developing countries are highly reluctant to collaborate and more so to devolve PTA to local authorities (Ankamah and Yao 2013). In cases where the central government has devolved such powers to local authorities, these authorities have been ineffective in undertaking this administration due to central government interference, 'perverse incentives, inappropriate property tax policy, the lack of property administration systems, trained personnel, and synchronization of improved local service delivery with enhanced revenue mobilization' (Kelly 2000: 37). Despite this, there is a dearth of studies examining the relationship between the two levels of governments and how it affects PTA.
The importance of examining the relationship between central and local governments in shaping PTA is reflected in the broader literature highlighting the importance of this relationship for successful decentralisation from the central government. Such a successful decentralisation will enable both levels of government to effectively engage citizens and promote efficient and effective service delivery (Martinez-Vazquez and Vaillancourt 2011;Smoke 2015). As noted by Crook (2003: 78), 'the politics of central-local government relations explain what interests might gain or lose from any set of institutional opportunities, policy initiatives, and resource allocations and relate these factors to the political purposes of decentralisation', which may affect how property tax is administered. Successful local development, however, depends on deep collaboration across all levels of government (Bahl and Bird 2018), which may impact on LGs' ability to deliver needed public services to citizens.
This paper is interested in investigating the impact of the central-local government relationship in shapingand underminingeffective PTA in developing countries, with a focus on Ghana, which has been among the most successful development performers in Africa in recent decades. Ghana has undergone substantial formal decentralisation since 1992, but PTA has continued to perform poorlyaccounting for only about 14 per cent of LG internally generated funds (IGFs) in any given financial/fiscal year (Biitir and Assiamah 2015). 1 Scholars have attributed this problem to weak capacity, enforcement, and compliance (Armah-Attoh and Awal 2013; Ayitey, Kuusaana and Kidido 2013;Boamah and Okrah 2016). This focus is important, as existing studies continue to attribute the problem of PTA to poor record-keeping, the level of compliance, and corruption among others, thus failing to note possible deeper causes of this dysfunction.
Ultimately, the paper argues that a dysfunctional relationship between central and local governments has undermined, and continues to undermine, effective PTA in Ghana. This dysfunctional relationship is rooted in problems with the institutional framework on decentralisation under the 1992 Constitution (Republic of Ghana 1992), as well as the central government's half-hearted approach to decentralisation, which continues to hamper the successful and coherent administration of property taxation in local areas in the country. The dysfunctional relationship has left LGs with the lack of capacity and clarity to undertake the administration of property tax, which is the most critical component of their IGFs (Oppong 2021). We substantiate our argument by analysing this central-local government relationship across five areas: (1) the institutional quagmire in decentralisation; (2) problems of property valuation, in particular the monopoly by the Land Valuation Division; (3) general capacity problems, including those affecting PTA at the LG level; (4) the politicisation of local governments by the central government; and (5) institutional fragmentation at the central government level.
To achieve our objective, we conducted an in-depth qualitative investigation of the realities of PTA in four LG areas with significant property tax potential: the two top-ranked assemblies in the country in delivering development successfully 2 -La-Nkwantanang Municipal and the Tema Metropolitan Assemblies, both in the Greater Accra Regionalong with Sekondi-Takoradi Metropolitan Assembly (STMA), which is the hub for significant new oil and gas production and has recently undergone a large scale property tax reform process supported by USAID, and La-Dade Kotopon Municipal Assembly (LAKMAA), which has perhaps the greatest property tax potential of any district in the country. In looking at these areas, we are able to capture the challenges of effective administration even in districts with significant financial, administrative, extractive, and technical capacitiesand in doing so further underline the likely depth of these challenges elsewhere. We also conducted interviews as well as focus group discussions with a number of ministries, departments, and agencies -1 The IGF includes market tolls and other income generated ventures prescribed by the constitution and Local Governance Act, 2016, Act 936. For more on what is happening on the continent, see Fjeldstad, Chambas and Brun 2014. 2 This ranking is from the 2017/2018 District League Table, which is jointly produced by UNICEF and CDD Ghana annually. https://www.cddgh.org/ghanas-district-league- table-2017 the Local Government Services Secretariat (LGSS), the Ministry of Local Government and Rural Development (MLGRD), the Land Valuation Division (LVD) of the Lands Commission, the Land Use and Spatial Planning Authority (LUSPA), the Ministry of Finance, the Ministry of Environment, Science, Technology and Innovation (MESTI), the Inter-Ministerial Coordinating Committee (IMCC), 3 and the Institute of Local Government Studies (ILGS), as well as two think tanks, 4 in order to ensure that the findings reported here reflect broader national realities, and a broad cross-section of perspectives. We also interviewed two experts on local government and PTA. Within each assembly, we interviewed bureaucrats, using a combination of one-on-one interviews and focus group discussions.
The paper is organised as follows: the section following this introduction provides a historical context to PTA in Ghana. It traces the development of local government authority and its need for fiscal sustainability in Ghana and the various attempts in PTA. In the section that follows, we utilise the data collected to analyse the relationship between the two levels of government and show the impact of this relationship on PTA in the country. The final section concludes the paper with some suggestions for future studies.

The challenges of strengthening property taxation in Ghana
To understand the contemporary challenges of PTA in Ghana, it is useful to go back to the institutionalisation of property taxation. The tensions and challenges that have continued to exist between central and local governments in managing local revenue collection can be found in the history of Ghana's development. Under colonialism, Ghana was governed under a system of indirect rule, in which local Native Authorities were responsible for local government, but subject to broader control by the central colonial government (Lugard 1965).
It is within this context that the first attempt to impose a form of property taxation, the poll tax in 1852, was made. However, the tax could not be fully implemented as the indigenes embarked upon a series of campaigns against the tax, which led to riots and disturbances in the colony and the subsequent withdrawal of the tax in 1861 (Nti 2002). Subsequent years saw repeated efforts to introduce a local tax on immovable property, in the form of 'ntokura tow' (literally, 'window tax'), which was based on the number of windows in a building (Frimpong-Ansah 1991;Jibao 2017). The tax empowered LGs to impose a rate on buildings, provided the said building had an annual rental value greater than six pounds (£6) (Ayitey et al. 2013, Adem andKwateng 2007). These efforts encountered consistent challenges of administration in the form of jurisdictional disputes. Building complete property rolls was viewed as inequitable due to the focus on only the number of windows, and the inability of local authorities to enforce it (Hailey 1951;Kuusaana 2015;Maasole 2017).
Despite these problems, the fundamental commitment to decentralising revenue raising for local governance and development persisted. For example, the Coussey Committee (1949) 5 recommended that local authorities 'impose a tax on property and/or other indices of wealth in the locality, which will have the effect of making the wealthier members of the community contribute more to revenue than the less wealthy'. Furthermore, it noted that 'in order to 3 We had intended to interview some Parliamentarians, but those contacted did not respond to our request for interview. Also, time constraints did not allow us to interview officials from the National Association of Local Authorities of Ghana and the Office of the Common Fund Administrator. The focus group discussions were at the LGS, ILGS and the four assemblies. 4 The think tanks were the Centre of Democratic Development and the Institute of Democratic Governance. 5 The committee was established to develop a constitution for an independent Ghana. ensure uniform standards of valuation of property, … the central government through the Regional Administration should provide a trained corps of valuers whose services should be made available to localities at agreed fees ' (29-30). In 1951, the first Gold Coast (Indigenous) government was elected, and it introduced the Local Government Ordinance of 1951 (Ayee 1990;Harvey 1966;Nsarkoh 1964), which created the path to the practices of charging local governments a fee for property valuation.
This legislation outlined the level of property taxation for LGs and defined the relationship between the two levels of governments, again presaging future conflicts. For example, while the law provided the council with substantive powers including the levying of taxes, it at the same time ensured that byelaws of the councils and those related to the collection of property tax were not to have any effect until approved by the minister responsible for LG, who had the power to amend or revoke byelaws (Harvey 1966). It also marked the beginning of the present rating system, with the LG (Immovable Property Rate) Regulations 1954 prescribing that valuation be carried out by a Land Valuation Department and assessed as 10 per cent of the replacement cost of the premises (Greenwood 1962). Two issues were critical. First, local authorities were not really empowered to set their own rateable value, or to carry out property rating. Second, little was achieved due to the 'shortage of qualified personnel to train officers in the new valuation methods' (Greenwood 1962: 24) by the central government.
The post-independence period brought further reforms, but in many ways the basic pattern remained unchanged: key responsibilities and activities related to property taxation remained at least partially in the control of the central government, thus complicating PTA. Following independence, the government passed a series of laws reaffirming the broad role of local governments and their right to levy taxes on immovable property, especially buildings. Nevertheless, in all matters of rating local authorities had to seek the prior approval of the minister. In addition, a list of tenement structures was exempted from assessment and rating by the government (Nsarkoh 1964). Under the Act 'the responsibility for the valuation under the new system [was] with the central government and valuation officers [were] employed by the Ministry of Interior' (Nsarkoh 1964: 60). In 1964 the government amended the 1959 Municipal Rating Act (No.4) (GoG 1964) to enable individuals to challenge the rating of their property(ies) not with the local authorities but with a central government valuation committee for redress. At the same time, the LGs could not enforce the payment of property rates by central government entities, as well as politicians in their areas, due to what has become known as the big men syndrome, centrally appointed weak administrators, corruption, and political interference (Werlin 1972: Nugent 1995. In 1967, the ruling government set up the Mills-Odoi Commission, which recommended the further decentralisation of power to local authorities. However, these recommendations again failed to be realised, as a new Local Government Act passed in 1971 instead reaffirmed central government authority over various aspects of PTA by creating institutions that affected the ability of LGs to undertake effective PTA. 6 Furthermore, the government indicated that 'the Minister may by legislative instrument prescribe either generally or in respect of any particular district a basis for the assessment of rateable value of premises' (GoG 1970). This power was in addition to several other exemptions, which were subject to the approval of the minister. 6 In Section 24, the Act noted 'subject to the approval of the Minister, a Council may charge fees for any services or facility provided by the Council or for any license or permit issued by the Council'. Article 67 Subsection 2 entreated local authorities to seek approval from the Regional Council, controlled by central government, for levies in the local area. Section 68 limited the amount of levy to be collected, while Subsection 8 of Section 69 noted that 'it shall be the duty of the Minister to cause to be determined the ratable value of premises for the purposes of this section and for the said purposes, he shall cause a Valuation List to be prepared for each district'. Thus, based on these articles, the Councils were made subservient to the central government in terms of their ability to internally generate their own revenue, including property taxation, for services. The next important change came in 1975, with the National Redemption Council government issuing new regulations for property rating. A key aspect of these regulations was the appointment of a Commissioner for Valuation, with greater power over property rate administration relative to the local authorities (GoG 1975). For example, the Commissioner was charged with the responsibility of preparing the valuation list for each district and oversight over administering the immovable property rate. The Commissioner was also empowered to appoint an assessment committee at each local level with the power to appoint a Clerk for them. This committee was not responsible to the local council but rather to the Commissioner. This new regulation weakened the local authorities' ability to collect property rates in their jurisdictions and strengthened the hands of the central government in this respect, creating what has been described as 'lock-in effect' (Liebowitz and Margolis 2014).
Subsequent years saw important further decentralisation reforms led, most notably, by the Provisional National Defence Council Law (PNDCL) 207 (GoG 1988) and Local Government Acts of 1993(GoG 1993GoG 2003). However, the broad landscape for PTA remained relatively unchanged. The Acts noted the importance of property taxation and the right of local governments to set rates and to collect revenues. However, the Acts continued to impose restrictions on these powers, most notably in relation to the identification and valuation of propertiesthe most basic functions for effective PTA. Most notably, Subsection (8) of Act 1993 identified the Land Valuation Board (LVB) as the only authority to determine the value and rates on properties, as well as prepare the valuation rolls. The Acts maintained many centrally-determined exemptions, while giving the Minister of LG the responsibility of setting up a rate assessment committee for LGs, to which an individual aggrieved with the rate imposed may apply for a review (GoG 1993). 7 The combined effect has been that LGs have had little recourse in cases where the LVB did not effectively develop accurate property rolls, which has then affected property identification, assessment, and collection.

Central-local government relationship and property tax administration
In this section, we will use the themes identified in the introductory section to investigate the relationship between the central and local government and examine how this affects property tax administration. As already noted, five main themes were derived from the interview data. These are: 1) the institutional challenges underpinning decentralisation and property tax administration; 2) problems with valuation and data management; 3) human resources and capacity problems at the local government level; 4) the over-politicisation of local governance by the central government; and 5) institutional fragmentation (silos) at the national level. It must be understood that these themes and separate variables are interrelated and significantly affect each other.

Institutional challenges
Various institutions, i.e., laws and regulations, govern the relationship between central and local governments, as well as PTA in both developed and developing nations. While these institutions play an important role in PTA, they also pose significant challenges for local governments, especially in countries where such governments are subject to the whims and caprices of the central government (Ahmad, Brosio andPöschl 2015: Kelly 2014). For example, they may affect what sorts of fiscal transfers are sent or received, the level of taxation and other levies that LGs can undertake, the property registration system, tax obligations, land and property taxation, statutory land use, and the dispute resolution system (Tang, Wong and Liu 2011). In addition, they allow central governments to dictate through mandates and restrictions what LGs can or cannot do including tax exemptions and, in some cases, rates to be charged, as well as a host of other things, and yet expect the local government to take the financial burdens associated with such policies in the provision of certain public services (Fan 2014;Mikesell 2018).
In Ghana, the institutional arrangements setting the relationship between the central and local governments are enshrined in Chapter 20, articles 245-256 of the 1992 Constitution. The chapter lays out the roles and responsibilities of local government, and in turn gives the central government responsibility for passing laws to put these roles into practice. This arrangement has, in practice, left the space for the central government to grant de jure autonomy to LG, but de facto to strictly constrain it, with the attendant broad array of problems for LGs. 8 The constitutional provisions regarding this relationship have led to what may be described as 'centralisation in decentralisation'. While central government has decentralised administratively, it has recentralised in a number of areas, especially finances and personnel due to its ability to enact laws, which enables it to exert itself in LG administration. Thus, while the constitution calls for a broad devolution from the central government, the powers it grants to the central government in terms of controlling the assemblies, have created ambiguities in the process and the failure to decentralise PTA effectively. Several interviewees expressed displeasure about these constitutional provisions. An interviewee said: The constitution has created more problems for LG development. It gives us five different features of our LG system. It talks about adequate financing, etc. It doesn't talk about property tax and who should administer it. The constitution should have been clear, as to which level of government should have this responsibility, since it gives both levels of government some level of power to tax. This omission has led to the central government's role in PTA through the various parliamentary Acts, policies, etc. 9 Consequently, another interviewee was of the view that if Ghana is to develop as expected, then parliament must amend the constitutional provisions that constrain the functioning of assemblies, so as to empower them, especially in areas of revenue administration (including property taxation) within their jurisdictions and use the revenue yields for development. 10 This intergovernmental framework continues to create confusion when it comes to property tax as many still believe that PTA is the duty of central government. There is a lack of clarity around the various regulations and how to undertake PTA. Hence, processes have been laid down in the law to guide local governments in the collection and utilisation of the resources. Property tax forms an important source of revenue mobilisation for local government, especially for urban local governments compared to rural ones. Rural LGs either do not have enough properties to tax or lack the capacity to identify appropriate properties to be taxed. 8 This is common in many countries.
LGs derive their powers, roles, and responsibilities from the CG. Thus, the quality of the local government system is only as good as the enabling and support from the CG. 9 While Article 146 of the Local Government Act, 2016, stipulates the method of rating, there is serious ambiguity about the roles of each level of government in PTA. For example, while the central government, through its agencies, is to do the valuation, local governments are to collect rates, and if there is any problem, citizens are to contact the central government. The ambiguity of these rules has opened the space for different interpretations of what PTA is all about. 10 See Article 35, for example, that requires the central government 'to take appropriate measures to ensure administrative and financial decentralisation and to give opportunities to people to participate in decision-making at every level in national life and government'. However, some level of autonomy as so to enable LGs to function very well is needed. As explained in this paper, the assemblies cannot even employ their own staff, as pertains in most developed countries. Every LG, however, is to collect property tax as part of its IGF to complement or support what central government provides in the form of transfers.
The constitution makes the central government the embodiment of accountability in the country, as expressed in the Financial Administration Act, 2003 (Act 654). There are two elements in dealing with accountability issues when it comes to PTA and the relationship between the two levels of government. First, LGs have the leverage and freedom to use and manage the property taxes they collect and thus do not remit property tax revenue collected to the national treasury. Such taxes are, therefore, for local development.
LGs are expected to budget for the collection and use of such taxes in advance, and declare what is collected and how it was used in the previous year, as well as disclose how much they intend to collect and use for the next fiscal year to the central government. Second, as part of this process, at the end of the fiscal year, LGs have to account for this taxation in their financial statement to the central government to indicate how much of what was budgeted for was collected, why they were not able to collect (if there is a shortfall), how collected taxes were used, and the areas in which they have funded development with these funds under a Public Financial Management Framework (PFMF) (GoG 2016). This is how central government regulates and provides oversight of the fiscal decentralisation relationship between the two levels of governments. The constitutional provisions and the PFMF continue to enable the central government to control the assemblies.
While there is nothing wrong with the central government setting regulations, interviewees believed such regulations affect how the assemblies can perform. For some, under the PFMF, accountability is an upward issue rather than a downward one. Assemblies are more eager to show to the central government rather than the locals what they have done, leading many to question what property taxes are used for and making them reluctant to pay these taxes (Bratton 2012). 11

Problems of property valuation: the monopoly of the Land Valuation Division
A fundamental issue in PTA is property valuation. Without an effective and efficient system to accurately value and revalue properties at periodic intervals, undertaking property tax administration will be a mirage (Effiong 2015). The issues relating to property valuation and the problems associated with how it is carried out have been discussed in the literature (Grover, Törhönen, Munro-Faure and Anand 2017;Kealy, Dovidio and Rockel 1988;Turner 2010). Interviewees unanimously agreed that one of the main issues undermining PTA in Ghana is property valuation and data management, which they blamed on the central government's heavy hand on LGs (see Mends 2006).
The LVD is a central government institution mandated to collate research, manage, and record all data on properties in the country under the Lands Commission Act, 2008 (Act 767). Prior to Act 767, the LVD operated under Section 43 of PNDC Law 42 (1986). 12 The LVD determines the values of properties rented, purchased, sold, or leased by or to government. The LVD has historically been given a monopoly over the valuation process, at a cost to LGs.
Nevertheless, PNDCL 42 only outlined that there shall be a LVD, which shall be responsible for valuing all government properties, but did not provide details about the agency's 11 According to Michael Bratton (2012: 518), these LG entities 'tend to respond to the preferences of the state authorities, international donors, or foreign-funded NGOs that provide the necessary resources' due to weakness of the local tax base and their continuous reliance on transfers from national treasuries. 12 Before 1986, the Valuation Division of the Ministry of Local Government was responsible for carrying out rating valuations in the country. Then in 1986 under the PNDCL (Supplementary and Consequential Provisions), 1982 (Section 43), the Valuation Division was transformed into the Land Valuation Board (LVBlater Division) and charged with preparing valuation lists for property rating, and determining compensation for land acquired by government or any public corporation (Kasanga and Kotey 2001). composition, functions, its roles, etc. It was this lack of clarity that led the government to bring all land institutions together with a detailed framework on the functions of the LVD under the new Lands Commission Act (Act 767). Nonetheless, Act 767 also made the LVD the sole valuer of properties in the country thus enhancing the monopoly on property valuation in the country, although there are existing private entities that assemblies should be able to engage to conduct valuation of properties in their jurisdictions.
There are two significant problems with this monopoly. First is the inability of the LVD to value properties in the country in a timely manner. The LVD has its own problems, especially capacity issues. For instance, it has only a small number of offices nationwide. In addition, there is a severe shortage of qualified staff including properly trained valuers, lack of logistical support, poor staff remuneration, and delays in compensation payments. Consequently, valuing properties has become a major headache for LGs. An interviewee explained, 'The LVD does not have the capacity to tackle over 265 metropolitan, municipal and district assemblies (MMDAs). Now the legislation allows [the] LVD to engage services of private valuers to support them, but that requires financial resources which [are] also not available to them.' 13 While properties are supposed to be valued every five years as per the law, this has not been true in practice, and in some local assemblies, valuation of old properties has not been carried out in more than 20 years, 14 due to administrative, technical, and financial capacity issues.
The LVD has limited resources at the national level, and this is worse at the LG level. An interviewee from the LVD explained: 'At the district level, we do not have what it takes to maintain high level personnel. So, at best the assemblies cannot engage valuers on a permanent basis. Our capacity constraints are in human, as well as financial resources.' The lack of technical capacity continues to constrain the ability of the LVD to compile the valuation rolls for local governments and to provide adequate guidance for appeals and negotiations of valuations. In addition, the operations of the LVD are also fraught with several other challenges such as unavailability of accurate land ownership records, operational logistics, and low staff morale. This hinders their ability to perform valuation effectively. 15 Another key issue arising from the LVD monopoly on valuation is the cost associated with such valuation, which is borne by the assemblies from their already meagre resources, as mentioned by interviewees.
LGs are expected to either pay the LVD upfront or after the work is done. Without such payment, the LVD can withhold the completed valuation rolls from the assembly. The majority of LGs, however, cannot pay upfront for the valuation as 'valuation is very expensive', say some interviewees. This poses a serious legal problem for LGs as the law on property taxation says assemblies cannot charge property tax if the property has not been properly valued and allows individuals to even seek redress at the courts if the law is not adhered to. In short, the allocation of valuation responsibilities to the low capacitated LVD, and the cost of valuation, prevent local governments from conducting valuation in a timely fashion.
To overcome this problem, the central government is considering amending parts of the existing law to ensure that LGs are able to do a 'cost benefit analysis' of their overall financial situation, including income and expenditure, to establish the cost of collection of property taxes, and then how much they are likely to mobilise with respect to other taxes, says an interviewee. Another respondent explained that: 'If the cost benefit analysis indicates that if an assembly spends, for example, $10,000, it would be able to recoup such an amount within the next five-year period, as well as make excess money from those taxes, then the 13 The LVD should not have offices at every assembly. Some form of organisational clustering by the LVD, where an office can look after five to ten assemblies, may suffice in addressing this problem. 14 According to Dafflon and Madiès (2013), the last valuation rolls for Accra, for example, were produced in 2006.
LGs should be empowered to hire private valuers to undertake the valuation for them. 15 This is a problem that public bureaucracy in Ghana is facing. Thus, the problem is not unique to the LVD. local government can go ahead and undertake property valuation.' 16 Based on this cost benefit analysis report, the assemblies will be able to go to banks to borrow to finance the valuation of the properties. In addition, the central government is developing a LG finance project that will facilitate the ability of the assemblies to approach the capital market to borrow to finance property valuation for tax purposes.
Furthermore, as part of this amendment, the central government is considering liberalising the valuation space altogether. 17 This liberalisation would enable the assemblies to engage qualified private valuers to undertake the valuation exercise to lower the cost of valuation for assemblies. A caveat though is that any valuer (be they individuals or organisations), must be certified and in good standing with the Ghana Institute of Surveyors (a monopolistic organisation), thus adding another layer of bureaucratic red tape to the valuation process.

Human resources and capacity building issues in local government
The literature on developing countries has identified four main types of capacity issues that need to be addressed if these countries are to overcome their developmental problems. These capacities are administrative, technical, regulatory, and extractive 18 (Brautigam 1996;Haldenwang 2017;Lodge 2014). Local governments need all four types of capacities if they are to ensure effective delivery of public services. This is because the lack of administrative capacity, for example, will seriously hamper their ability to undertake their extractive and regulatory duties, i.e., tax collection, building permits, land use, etc. At the same time, without the technical capacity, property valuation will be a serious problem.
Our interviewees consistently pointed out these four capacity issues and how they continue to affect LGs; they noted that without addressing these capacity gaps, not much can be achieved with regards to property tax administration. Two issues are of essence here: (1) the recruitment of personnel and the power to fire non-performing personnel; and (2) the design and implementation of capacity building initiatives for local authorities.
Local governments continue to have serious human resource issues; they are not able to recruit and build the capacity of employees on their own. Part of the inability of LGs to build their human resource capacity is the central government's hold on personnel. Under the decentralisation system in Ghana, LG officials are central government employees, who until recently were under the civil service. When the LGSS was set up as a semi-autonomous institution in 2003, these employees were automatically transferred there. Thus, the central government continues to have power over human resources management at the local level and the power to transfer, fire, reprimand, etc. LG employees without any recourse to the local authorities or even the bureaucracy at the local level. An interviewee bemoaned: 'The challenge we have with this [property tax administration] is the fact that we do not have the personnel to undertake what is expected of us… This problem delays the collection of determined rates by the assemblies.' Another respondent claimed: 'The underlining issue is that the right people are not being funded (training), the right people are not leading the processes, the right technical capacities are not being fostered and local authorities have not 16 It is important to ensure that the property tax is not seen as a 'nuisance tax', where the costs of administration are more than the revenue mobilised. A critical part of any reform should be how to streamline the administration to reduce both the administration and the compliance costs. 17 As at the time of our interviews, a draft bill on the liberalisation of the valuation space to enable individuals and organisations to participate in the process was being developed by the MLGRD. 18 Administrative capacity refers to 'the ability of states to plan and execute policies and enforce laws cleanly and transparently' (Fukuyama 2004: 9), while regulatory capacity is about the 'ways in which resources are allocated to ensure that systems of control maintain their well-functioning in often uncertain environments' (Lodge 2014: 65). On the other hand, extractive capacity refers to a capacity for interaction between government and social resources (Fewsmit 1995), while technical capacity is the ability to plan and translate broad objectives into programmes and projects and to ensure and monitor their implementation. been given the right kind of authority to undertake their assignments, and that is the current state of things.' One expected this to change with the passing of the Local Government Act 936 in 2016 after numerous complaints about the human resources issue. Sadly, Act 936 still empowers the Office of the Head of Local Government Services (OHLGS) over recruiting, developing, and transferring of LG personnel. Consequently, the OHLGS can simply transfer personnel in district A to another on the basis that the person has been there for four years, without taking into account the performance of the personnel and the needs of the LG involved. As one interviewee put it, 'No one is an employee of any assembly; they are all employees of the LGS, and this allows the centre to control the activities of the LG at the detriment of local development.' 19 A second problem of capacity building is that LGs are unique entities based on their jurisdictions and the local economy. On this basis, they are classified by the MLGRD into metropolitan, municipal, and district assemblies. Consequently, their capacity needs are different. Nonetheless, government has not empowered them to develop and build their own capacities but rather continues to promote a one-size-fits-all approach through the office of the OHLGS. The OHLGS designs and then implements capacity building initiatives, which are mostly determined by the central government agencies such as the MLGRD, which does not take into consideration the local environment prevailing in a particular locality. It is thus difficult for these assemblies to design their own capacities when it comes to addressing PTA. In situations where they can do so, interviewees explained that such strategies are overridden by central government agencies, such as the case of the Tema Metropolitan Assembly that was made to suspend its IT-led strategy on PTA by the central government, as it had not been authorised by the Minister for Local Government and Rural Development to implement it.

Local governance and property tax administration
Local governance encompasses 'the direct and indirect roles of formal institutions of local government and government hierarchies, as well as the roles of informal norms, networks, community organizations, and neighborhood associations in pursuing collective action by defining the framework for citizen-citizen and citizen-state interactions, collective decision making, and delivery of local public services' (Shah and Shah 2006).
A key aspect in achieving good local governance is decentralisation with significant autonomy from central government (Prichard 2015). However, most developing countries lack this autonomy, which has led many central governments to politicise local governments through political capture. This, in turn, has affected property taxation, as well as local revenue sources (Masaki 2018). As noted by Haque (1997), the politicisation of local institutions and the imposition of various programmes on them by the central government makes it harder for local institutions to be responsive to local needs and affects their ability to even embark upon effective revenue mobilisation, including PTA. Similarly, Haldenwang (2017) notes that effective property taxation can be hampered by political settlements led by the central government, where local politics or tax administrations are captured by powerful groups and any decision to tax is taken in the light of the political costs it entails.
In Ghana, this problem manifests itself in the power of the President to appoint and dismiss metropolitan, municipal, and district chief executives, as well as one-third of the General Assembly members (Conyers 2007), under Article 243(1) of the Constitution. This provision 19 It must be noted that assemblies are allowed to hire their own tax collectors, but most of them are part-timers and poorly paid. These tax collectors are, however, not involved in property tax collection. Individual property owners are expected to go to the offices of the assemblies to pay their taxes when they receive notifications. has fuelled the over-politicisation of the assemblies and hampered the ability of LGs to effectively administer property taxes. The reluctance to pay taxes, including property taxes, is not taken seriously by central government appointees. To such appointees, enforcement of tax obligations on citizens may lead them to revolt against the central government in future elections. This over-politicisation was a serious concern to our interviewees. One noted: I think that the politicisation of the local government system is a bane. If the constitution is amended, probably it will lessen the over-politicisation of matters of the assemblies, because we are not getting it right. The assembly is a government for the local people, and they must act as such. If you look at what the constitutional provisions state, it means that they (assemblies) must have their own policies even if they want to tie them into a central government policy.
Chief executives are the principal agents charged with the day-to-day administration at the LG level. There are a number of provisions under the LG system, which ordinarily should not exist under a democratic system, as under this system chief executives are more accountable to the President than to the local populace. Second, the selection and appointment processes are very opaque: unelected local party executives present a name to the President, who then nominates the person, and he or she is then confirmed by two-thirds of the assembly members. In the majority of cases, these nominated individuals easily obtain the two-thirds vote, as they only need half of the elected assembly members' voteone-third of the appointees are always 'locked-in' due to party affiliation. Where such a nominee is unable to obtain the required votes due to 'politics', the President has the power to withdraw the one-third of appointed assembly members and replace them with party officials that will vote favourably for the nominee.
In addition, the President has the power to stop any activities that are being carried out by the assemblies, which he or she may deem as detrimental to his or her chances of reelection or to the re-election of their party (Gyampo and Graham 2017). This power significantly affects how the assemblies carry out revenue generation activities. An interviewee explained that 'there have been instances where chief executives may try to sack people selling on pavements [and] for them to go into the markets so that the assemblies can tax them. As soon as they start the process, the powers that be will say that "Look: this will affect our chances in the next election so stop it".' Another also explained that 'If somebody is not paying property tax, maybe a party office, they have not paid their property taxes or they are building without a building permit, you go in there and the party hierarchy from Accra will tell you that this is a party office so you can't do it.' Another major problem is that chief executives do not have 'tenure'. Although the constitution says that chief executives are to serve for four years, this is normally not the case, as they serve at the pleasure of the President. As rhetorically explained by an interviewee, 'They can be appointed this morning and sacked the following evening.' To another interviewee, 'A district chief executive (DCE) becomes a DCE at the whims and caprices of the President and doesn't even know when they are going to be sacked.' Thus, this politicisation through the appointing power of the President continues to hamper the ability of chief executives to perform as expected, including the enforcement of property tax payments in particular, with defaulters who are politically connected. Consequently, this has led to what Haldenwang (2017) has described as the shying away of local governments from the political costs associated with more active revenue mobilisation in developing countries.
To overcome this problem, most interviewees advocated for the election of metropolitan, municipal and district chief executives (MMDCEs), as well as the adoption of an institutional framework governing the relationship between the centre and the local. Interviewees believe that such an election would reduce the political interference from the central government, ensure that MMDCEs are accountable to the electorate, and help assemblies to undertake reforms such as property tax reform. In short, such an election would force MMDCEs to focus on the developmental needs of the people and push them to develop strategies for the administration of property tax for development (Agomor, Adams and Taabazuing 2019). Many believe that once a MMDCE was elected, the lack of party constraints would allow them to recruit the right people and prioritise the 'correct' projects, as they would be judged by the electorate in the next election cycle on their achievements. This creates an incentive for all parts of the assembly to achieve its developmental goals right from the top.
In all, the enormous political control the central government continues to exercise over LGs not only over hiring, but through its appointment of chief executives and one-third of the assembly members, has a significant impact on assemblies' ability to pursue any meaningful reforms, especially those related to the assemblies' finances. Consequently, any serious reforms which the central government may deem threatening to its political interests and survival are likely to be blocked by those short-term, and narrow, central government concerns. Hence, the seeming neglect of reforms that could create more financially sustainable, efficient, and effective assemblies including local tax and property tax reforms.

Institutional fragmentation ('silolisation') at the national level
One of the major debates in PTA concerns whether institutional consolidation or fragmentation is beneficial or not for local authorities (Krupa 2017;Tavares 2018). Institutional fragmentation refers to the way in which distinct institutions exist and interact (Biermann, Pattberg, van Asselt and Zelli 2009), or simply, the number of governmental units in a given area (Goodman 2015), whereas consolidation signifies the amalgamation of such institutions.
When it comes to PTA, overly fragmented assessment systems, for example, do not produce equitable and timely property tax assessments, ratio studies, and reports (Krupa 2015). An important issue with fragmentation is the lack of trust among institutions. Without trust, individual organisations may refuse to share information, which may affect local governments in their attempts to effectively administer property taxation. Trust becomes an important issue in institutional fragmentation (Fjeldstad, Ali and Katera 2019).
This institutional fragmentation was highlighted as one of the major problems affecting PTA in Ghana. A major problem with this fragmentation is conflicting and overlapping mandates and the lack of information sharing among agencies such as the Office of Stool Lands, the Administrator of the District Assembly Common Fund, 20 the Lands Commission, MLGRD, and the Ministry of Environment. These agencies hold different information, including on land acquisition by both individuals and companies in a particular local jurisdiction, which could help in the gathering of information related to property tax assessment and possibly the persecution of offenders. Such fragmentation and lack of coordination affect the ability of local authorities to consult other institutions in the development of relevant policies in relationship to property taxation and other financial issues for the assemblies.
In PTA, the collaboration between these ministries and agencies, and between agencies themselves, especially those involved in the valuation and planning processes, is critical. A respondent at LUSPA explained: 'Sometimes some ministries think that [other ministries] are usurping their roles. Let me take the MLGRD for example. They are doing a lot of spatial planning initiatives, which we think we should have spearheaded. We are a different ministry. There are frictions between some of these ministries and most of the other sectors. What we 20 The District Assemblies' Common Fund is a pool of resources created under Article 252 of the 1992 Constitution of Ghana. It is a minimum of 5 per cent of the national revenue set aside to be shared among all District Assemblies with a formula approved by Parliament. It is a development fund. See http://www.commonfund.gov.gh need to do is a realigning of the roles and responsibilities, which we are now lobbying to go to local government.' On the question of the collaboration between agencies and ministries and how silolisation affects them, a respondent said: They [LVD] see themselves as an authority on their own. They have their board and so one cannot dictate to them what they should do. I think they have a problem with [the] Lands Commission, even before they were bundled together to form the Lands Commission. We had [the] Survey Department and Land Valuation, but now they are all together under one executive secretary. So, they can work together, but not with the other agencies such as Town and Country Planning. They can't work together because of the different roles they play, even though it would have been ideal if they are all together so that the processes would be easy.
Another interviewee explained: From my personal experience with LVD, the relationship is not all that friendly. They are supposed to collaborate with us, especially the MMDAs. Where they have done street naming, they are to evaluate the properties so that we can know the right property tax to be implemented. You can go to them several times that 'let's collaborate and do it', but they want to do things their own way, although there are timelines for MMDAs.
Clearly, there is a lot of hostility because of 'turf wars' between LUSPA and other central government institutions and even between the MMDAs, the LVD, and other institutions that are supposed to be part of the assemblies but report to their head offices rather than the assemblies. Consequently, there is no fluidity or a harmonious relationship between all the institutions in the field. 'Everybody wants to see himself as the champion', says an interviewee. The problem with this fragmentation is that assemblies are caught in between with no power to deal with any of these institutions, or with the conflicts. Thus, the proverbial idea that when two elephants fight, it is the grass that suffers is applicable in this situation.
To overcome this fragmentation and to some extent the turf wars, an interviewee suggested a strong 'intersectoral approach', where 'We can equally talk about the quality of human resource, the assemblies employing their own personnel, as well as better decentralised agencies, which will report to the assemblies and uphold the assemblies' authority.' Another interviewee was of the view that the IMCC should take the lead on issues of coordination and responsibilities because of the power it wields, with the hope of reducing 'silolisation' and promoting the sharing of information, especially between LVD and LUSPA. Another interviewee indicated that: 'There are unilateral decisions [by MoF, for example] where everyone has to go along which can lead to rocky implementation at the ground level, but in cases where everyone is on board at the same time, if there is harmonisation, then there will be a positive impact on the ability to do property tax administration.' Thus, consolidating institutions is an essential element in addressing problems associated with central-local government relationships in PTA in Ghana, and perhaps in other developing countries. This could help resolve silolisation problems, issues of turf wars and mistrust among these institutions, and promote good local governance.

Conclusion
Property tax revenues continue to be seen as a major source of income for local governments in both developed and developing economies. The literature on this subject has consistently shown how an effective PTA can bring many benefits to citizens. Despite this, effective administration of property taxation continues to be problematic for local authorities. In developing countries, the many attempts to develop and implement reform programmes to address the issue have not yielded the needed results. While existing studies continue to focus on the institutional framework underpinning PTA, capacity building, property valuation, and the issue of corruption, one important issue that affects PTA and which has been neglected is the relationship between central and local governments, which may impede the effective and efficient administration of property taxation (Dillinger 1991;Kelly et al. 2020;Mikesell 2007).
The objective of the paper was, therefore, to examine and understand this relationship and how it may be impeding the ability of local authorities in Ghana to administer property tax effectively and efficiently. Drawing on qualitative data collected from the field, we identified five major areas in central-local government relationships, which negatively affect local authorities in PTA. These are: (1) the institutional quagmire in decentralisation; (2) problems of property valuation, in particular the monopoly by the Land Valuation Division; (3) capacity problems at the local government level; (4) the politicisation of local government by the central government; and (5) institutional fragmentation at the central government level.
The study demonstrates how these five areas continue to enable the central government to strongly hold on to Ghana's decentralisation project and the local government system, and how it thus serves as a stumbling block to effective PTA at the local level. Similar to the findings of Fjeldstad et al. (2019), the paper has shown that political expediency from the central government has become a bane for PTA. Such political expediency has manifested itself in several ways including the way the constitutional framework defining the relationship between the two levels of government has been implemented, the serious politicisation of local government, the role of central government agencies in constraining the ability of local authorities in PTA, institutional fragmentation at the central government level and capacity problems at the LG. This relationship has created some level of despondency among bureaucrats at the local level who are expected to be at the forefront of the development and implementation of models for PTA. In a nutshell, this central-local government relationship has created a conundrum in PTA, which can be well illustrated as follows. First, the constitution and other institutional frameworks give the assemblies the autonomy to collect taxes and fees, as well as guidelines on how to utilise the resources for their operation and development. The central government supports this through the development of implementation guidelines, as well as valuation services for local governments. Given that one cannot impose a tax on a property if it has not been properly valued, LGs are limited in their ability to effectively undertake such an important exercise. However, the monopoly on property valuation by the LVD has affected the ability of LGs to engage private and/or non-LVD valuers to undertake property valuation. As a result, properties are not valued periodically, which makes data on property incomplete or out of date. The LGs lack the capacity to continuously check on property development. The same applies to the appropriation of land for development. Thus, property development, valuation, and taxation continue to be done haphazardly.
To address these intractable problems and ensure effective PTA, improving the relationship between the two levels of government is essential. This can be done through an effective institutional framework that should constrain the ability of the central government to frequently interfere in the activities of LG. Doing so will allow for the effective decentralisation of central government agencies and the empowerment of local authorities to have some level of control over these agencies. This may help reduce coordination problems among the various institutions as they exist now at the local level.
Furthermore, central government and development partners should assist local authorities in building local capacities through effective decentralisation, where LGs should be able to recruit their own staff, as well as design capacity building initiatives that fit their needs. This can be done through an effective collaborative approach that respects the boundaries of the relationship between the two levels of government (Ricciuti, Savoia and Sen 2019). There is also the need for such partners to incentivise local authorities to understand the importance of PTA. But this can only be truly effective if the central government reduces its strong hold on LGs.
It has been argued that 'local-central [governments] relation is a "murky" business and there is no short-cut to improving it. "Murkiness" calls for unbundling' (Oduro 2018: 900) of this relationship for effective development, especially if we accept the notion that development from below is the way forward to addressing wicked problems in developing countries. If this is the case, then future studies should examine how unbundling both administratively and politically can be done effectively and efficiently. What sort of institutional framework should scholars develop to examine this relationship and provide useful recommendations to government to enhance effective and efficient central-local government relationships in developing countries? How can local government develop the necessary capacity to enable them to enforce property tax payment in their jurisdictions? These are some questions that future studies may attempt to answer.
In addition, future studies should employ the mixed methods approach with the view of surveying revenue collectors and interviewing other officials to see whether the same result will be obtained. With such a method, the cases can be expanded to include more rural areas, as their needs may differ from well-endowed ones. Furthermore, undertaking citizen surveys to gauge their perception and understanding of PTA will also enrich the perspective of the issue better.
It is our hope that this study will ignite a more forceful debate over the issue and perhaps turn the attention of scholars, administrative reformers, and external organisations interested in local development to examine this relationship and develop theoretical frameworks that will enhance the study of central-local government relationships in the current decentralisation and PTA projects. Such a study may unearth further issues that can forcefully help to empower LGs to effectively mobilise local revenue including property taxes, which may help them to deliver better public services and address 'wicked problems,' which are intractable (Rittel and Webber 1973;Head, 2022), especially in an era of 'glocalisation' (Swyngedouw 2004).