7 - The Political Stakes of Decentralisation

Summary

With economic reforms in the 1980s, the opening up of political space and the end of war in the early 1990s, Mozambique embarked on a decentralisation process. As in other sub-Saharan Africa countries, the impact of the decentralisation reforms on local development and the strengthening of democracy has been modest. How can this be explained? This chapter addresses this question analysing how institutional dynamics shaped their results. The main argument is that the nature of the political system shaped reform results, in the context of institutional dynamics. Of these, those linked to state capacity and independence from private interests, stand out. Reforms are implemented according to group interests, particularly political parties’ interests, which capture the state and use them for maintaining and bolstering political power. Rather than being a means of improving the provision of public goods and strengthening democracy, decentralisation reinforces state control and panders the elite. Probably the biggest challenge facing decentralisation, this makes it a fundamental issue in any reform, within the context of strengthening democracy and promoting local development.

I Introduction

In the context of the economic and political changes in the late 1980s, made visible through economic restructuring and pro-democracy movements, several countries in sub-Saharan Africa embarked on decentralisation reforms. The majority of these reforms can be seen as a response to state weakness (Olowu and Wunsch 1990; Osaghae 2007) in terms of political regulation and service delivery, particularly in the education, health, water, and sanitation sectors. Thirty years later, many scholars find that the successive decentralisation reforms in sub-Saharan Africa had little impact on the resolution of governance problems (Manor 1999; Crook and Manor 2000; Tilburg 2008; Bierschenk 2010; Booth 2010; Crook 2010; Olivier de Sardan 2011; Batley et al. 2012; Mohmand and Loureiro 2017). In fact, the extensive literature produced in recent years and the different instruments of governance measurement in Africa (Ibrahim Index of African Governance, Afrobarometer, Democracy Index) show not only the little progress made in terms of governance, but also the complexity of the dynamics and the institutional factors underlying the implementation of decentralisation reforms (Mattes and Bratton 2016; Mo Ibrahim Foundation 2016).

With the introduction of the economic reforms in the late 1980s, the opening up of the political space and the end of the war in the early 1990s, Mozambique embarked on decentralisation reforms, based essentially on two dimensions: administrative decentralisation, under the scope of the Local Government Act (Law No. 8/2003, of 19 May) and political decentralisation, in the context not only of establishment of local authorities (Law No. 2/97, of 18 February) but also of the approval of the so-called decentralisation package, which calls for the election of provincial governors (Laws Nos. 3/2019, 4/2019, 5/2019, 6/2019, and 7/2019, all of 31 May, and Decree No. 2/2020, of 8 January).

With strong support from international cooperation partners, through different programmes focusing not only on the districts but also on the municipalities (Weimer 2012), the decentralisation reforms in Mozambique, according to research carried out in recent years, are still far from promoting citizen participation in the solution of local problems and the broadening and consolidation of democracy. In fact, as is the case in other countries in sub-Saharan Africa, the impact of the decentralisation reforms in Mozambique has been modest in the resolution of governance problems (Forquilha 2008; Forquilha and Orre 2011; Weimer 2012; Weimer and Carrilho 2017). Indeed, data from the most recent rounds of Afrobarometer surveys, for example, show a significant reduction with regard to demand for democracy, having gone from 25 per cent in the 2011–13 round to 9 per cent in the 2014–15 round (Mattes and Bratton 2016). In addition, the democracy index for 2018 shows that Mozambique’s classification has fallen significantly, going from a hybrid regime to an authoritarian regime (The Economist 2019).

How can the modest impact of the decentralisation reforms on the promotion of local development and the strengthening and consolidation of democracy in Mozambique be explained? Based on evidence collected from interviews, Afrobarometer data, assessment reports on support programmes/projects for decentralisation, and literature produced on decentralisation in Mozambique in recent years, this chapter aims to answer the question asked above by looking at institutional factors that constitute constraints to decentralisation reforms. It argues that the results of the decentralisation reforms are shaped by the nature of the political system and the way it operates, in the context of institutional dynamics. The chapter develops the argument exploring institutional dynamics by examining five institutional dimensions, namely (a) rule of law; (b) participation and political accountability; (c) political stability, violence, and the legitimacy of the state; (d) state capacity and independence from private interests; and (e) sovereignty and independence. The argument is looked at essentially in two parts. In the first part, the chapter analyses the process of formation of the Mozambican political system and the dynamics behind the decentralisation reforms. In the second part, the chapter focuses on the decentralisation reforms, seeking to analyse their results based on the way the political system operates, looking at the main institutional dynamics.

II The Process of Formation of the Mozambican Political System

Mozambique won its independence from the Portuguese colonial power as a result of a lengthy armed struggle led by the Mozambique Liberation Front (FRELIMO). Transformed into a Marxist-Leninist political party in 1977, FRELIMO has been ruling the country since independence in 1975, first in a one-party context and later, following its successive re-elections, under the scope of the multiparty system in force since 1990. Despite its revolutionary discourse on the ‘destruction of the colonial state’, in the period following independence, FRELIMO maintained the centralising rationale and practices of the colonial state. Indeed, not only did the new organisation of the Mozambican state fail to bring autonomy to the subnational levels, but the national unity discourse in the one-party context also reinforced the rationale of a centralised state (Brito 2019). How can this centralisation be explained and why, at a certain time, were decentralisation reforms begun? The answer to this question will help us understand the way institutional dynamics shape the results of the reforms. We will be discussing this below.

A The Historical Trajectory of the Postcolonial State Building in Mozambique

The historical trajectory of postcolonial state building in Mozambique is largely confused with the internal dynamics of the constitution and the development of the anticolonial movement led by FRELIMO. These dynamics were marked by rifts, conflicts and violence, which became more acute particularly in the late 1960s and early 1970s, and structured the institution building process in the postcolonial period (Forquilha 2017). In fact, the internal crisis within FRELIMO¹ during the anticolonial struggle contributed significantly to the reinforcement of the rhetoric exalting national unity to the detriment of ethnic differences. These differences were very often regarded as tribalism or regionalism and therefore incompatible with the political project based on the idea of transformation of mentalities and social relations in the context of building the Homem Novo² and a new society (Cahen 1987; Brito 1991; Meneses 2015;).

With the advent of independence in 1975, the role of FRELIMO in the construction and transformation of Mozambican society was reinforced and enshrined in the Constitution of the Republic.³ This fact contributed to the centralisation of the state, a process that soon gained consistency, particularly after 1977, when FRELIMO officially became a Marxist-Leninist party. In the programme presented for its 3rd party congress, FRELIMO’s role in the materialisation of the revolution and the centralisation of state power were important aspects, as can be seen from the extract below:

With regard to the state, the party’s task is to destroy the colonial state and create a new type of state apparatus that reflects the interests of the working classes in its structures and its activities. The state is the main instrument for the materialisation of the party’s revolutionary politics. Therefore, the party directs and guides all state activities. The leadership role of the party in relation to the state is as follows: […] b) the party outlines the fundamental guidelines for development in all walks of social life. These guidelines are reflected in the state laws. The state laws express the practical application of the party’s directives; c) the party creates conditions for all state bodies to become party organisations so they can ensure the party’s political line; d) the party mobilises and politically and technically trains staff loyal to the party to hold positions in the state apparatus […]. The People’s Republic of Mozambique is a revolutionary democratic dictatorship state of workers and peasants. It exercises its power against the enemies of the people, capitalists, against imperialism and its agents and in the interest of protecting and defending revolutionary conquests and territorial sovereignty and integrity.

(FRELIMO 1977: 28–29)

In this context, the centralisation of state power arose due to a combination of two important factors, namely the inheritance of the centralising nature of the colonial state and FRELIMO’s political project, culminating in the centralised Marxist-Leninist state at the service of the interests of the revolution. In the years that followed independence, the centralisation of state power significantly marked not only the functioning of the institutions, but also the operational practices and rationale of state employees and agents. This is an important aspect for understanding the result of the decentralisation process in the context of the market-oriented policy reforms underway in the county since the early 1990s.

B The Origins of the Decentralisation Reforms in Mozambique: Actors and Dynamics

The economic and political reforms taking place in Mozambique in the late 1980s and early 1990s, including the switch to a market economy, the structural adjustment programme, and political liberalisation, brought important consequences from the point of view of the institutional set-up of the state, resulting in the decentralisation process.

In the literature, the concept of decentralisation refers to the transfer of power, authority, functions, and competencies from the central state to lower levels (Rondinelli et al. 1983; Adamolekun 1999; Manor 2011). However, it is important to point out that the concept has different meanings. There are at least three. In fact, when authority and resources are transferred to agents of the central state located at different levels (region, province, and district), this is administrative decentralisation, which is also known as ‘deconcentration’ (Rondinelli et al. 1983; Adamolekun 1999; Manor 1999). When power and resources are transferred to elected, independent subnational units with a legal personality separated from the central state, this is political decentralisation (Rondinelli et al. 1983; Adamolekun 1999; Manor 1999). This is the case of local authorities (and, more recently, provincial authorities) in Mozambique. Finally, decentralisation can also mean the transfer of responsibilities regarding budgets and financial decisions (Rondinelli et al. 1983; Adamolekun 1999; Manor 1999; Forquilha 2016). This is fiscal decentralisation. With regard to Mozambique, the Decentralisation Policy and Strategy (PED) states that ‘in Mozambique, decentralization is carried out through local authorities, local state bodies and institutions with community participation and consultation’ (Resolution No. 40/2012, of 20 December). According to the PED, the district development fund (known as ‘the 7 million’) is limited to the scope of the governmenťs willingness to decentralise to empower the communities in the fight against poverty’ (Resolution No. 40/2012, of 20 December). Therefore, when we speak of decentralisation in this chapter, we are referring primarily to the two main types of decentralisation, which are administrative decentralisation and political decentralisation.

If it is a fact that the centralised state model had a significant impact from an ideological point of view, namely in terms of associating the population with FRELIMO’s political project, it is no less true that, from an administrative point of view, the results were modest. Indeed, the heavy centralisation of public administration, associated with the scarcity of financial, material, and human resources and the serious effects of the war of the 1980s, contributed to making the state distant from the citizens, particularly in rural areas, insofar that more and more difficulties arose in the provision of basic services, especially in the education, health, water, and sanitation sectors. In this context, in parallel to the process of implementation of the structural adjustment programme, which began in 1987, the Mozambican government took steps aimed at improving the functioning of public administration, underlining the need for less centralised management. That was how the law no. 2/87, of 30 January, on the general state budget for 1987 was approved. In fact, its preamble underlines the following:

The general state budget for 1987 reflects the economic and financial measures adopted by the government under the framework of the economic rehabilitation programme, whose aim is to reactivate production and the gradual reduction of the existing financial imbalances. It [the law] also reflects the concerns about introducing changes in the management mechanisms, particularly at companies and local levels, in order to make the management of the economy more streamlined and functional.

(Law No. 2/87, of 30 January)

With regard to decentralisation, it is worth mentioning Article 8 of Law No. 2/87, of 30 January, as it establishes the following:

With the aim of promoting greater involvement and accountability of people’s assemblies and executive councils at district level […]. During 1987, in accordance with the working methods to be defined by the council of ministers, working closely with the provincial governments, in each province, a district will be selected to act as a ‘piloť where tests will be carried out on the administrative and financial autonomy measures to be gradually introduced.

(Article 8 of Law No. 2/87, of 30 January)

At the institutional level, the law no. 2/87, of 30 January, in a way constitutes a break from the previous period in terms of the principle of administrative and financial management insofar as the aforementioned law uses expressions such as ‘involvemenť, ‘accountability’, ‘administrative and financial autonomy’, which have little to do with the state centralism established in the Constitution of 1975. It is true that the law no. 2/87, of 30 January, does not create autonomous structures separated from the central state for local management and administration. Even so, it advances towards the gradual transfer of certain central responsibilities to subnational levels. To this end, the law no. 2/87, of 30 January, marks an important step towards the decentralisation process, whose fundamental bases would be launched later with the approval of the Constitution of the Republic in November 1990.

In the literature, decentralisation processes frequently appear as the result of a combination of internal and external factors (Olowu and Wunsch 2004). This aspect makes it relevant to the analysis in that there is a tendency to reduce decentralisation processes in Africa to merely external factors, particularly the ‘demands’ coming from donors, especially the World Bank and the International Monetary Fund. It is true that the ‘good governance’ component has played an important role in the implementation of decentralisation reforms in Africa, including Mozambique. However, it is important not to lose sight of the fact that for certain African states, particularly Mozambique, decentralisation is a relevant process of conflict management and state legitimacy in the context of bringing peace and democracy to the country (Faria and Chichava 1999).

This does not mean that decentralisation reforms in Mozambique are a consensual and peaceful process. In fact, over the last thirty years, the implementation of the main legislative and administrative measures in the context of decentralisation have proven to be contentious, resulting in political divisions not only between the main political players, namely FRELIMO and the Mozambican National Resistance (RENAMO), but also in the very heart of the ruling party. Indeed, in all discussions that led to the approval of several laws on decentralisation reforms⁴, conflict has been one of the most significant features, playing an important role in the structuring of the results of the reforms, insofar as party political interests end up putting issues of consolidating democracy and local development on the back burner. We will return to this aspect below when discussing how institutional dynamics shape the results of the reforms in the context of decentralisation.

III Decentralisation Reforms in Mozambique: Their Results in Local Development and Strengthening Democracy

As mentioned above, the results of the decentralisation reforms in Mozambique are modest, as is the case in other countries in sub-Saharan Africa. An important part of this outcome can be explained by the context, that is the dynamics of the institutions in the process of implementation of the reforms. According to North, ‘institutions are the rules of the game in a society or, more formally, are the humanly devised constrains that shape human interaction. Consequently, they structure the incentives in human exchange, whether political, social, or economic. Institutional change shapes the way societies evolve through time and hence is the key to understanding historical change’ (North 1990: 3). Drawing on North’s definition of institutions and the rich literature on institutional theory (March & Olsen 1989; Skocpol 1995; Peters 1996; Fusarelli 2003), we argue that institutions are fundamental elements that shape the incentives of the actors and policy outcomes, including decentralisation reforms. To analyse how institutions shape the results of the decentralisation reforms, we will focus on two key aspects: (a) institutional dimensions, in which there are different institutional dynamics that may shape the results of the reforms and (b) three important reforms: municipalisation, district development fund (the ‘7 million’), and decentralised provincial governance.

A Dimensions of Institutions

Institutional dynamics are analysed based on five dimensions of institutions against the backdrop of the nature and the functioning of the Mozambican political system. These dimensions of institutions are rule of law; voice, participation, and political accountability; political stability, violence, and state legitimacy; state capacity and independence from private interests; and sovereignty and independence. Below, we will be analysing some elements for each one of the mentioned dimensions of institutions. The dynamics, which exist in each of the five dimensions of institutions, will allow us to discuss the decentralisation reforms we chose for analysis in this chapter.

1 Rule of Law

Mozambique is, formally, a democracy that calls for the separation of powers and respect for the rule of law. However, political practices since the adoption of the democratic constitution in 1990 are still heavily influenced by the historical trajectory of the functioning of the institutions in the one-party state. In fact, the principle of separation of powers, although enshrined in the constitution, has been a challenge in the process of building democracy in Mozambique, where, in practice, the executive power has a strong influence on legislative and judicial powers, with a significant impact on the design, approval, and implementation of policies, including in the area of decentralisation. An example of this is the definition and creation of the local authority package and, more recently, the so-called decentralisation package, where the parliamentary debate did not significantly alter the initial proposals from the government on the subject, despite the controversies in many of the aspects approved.

2 Voice, Participation, and Political Accountability

Decentralisation is one of the most important aspects of the policy reforms in Mozambique in the last twenty years, insofar as not only did it significantly change the set-up of the state with the creation of autonomous political entities through the municipalisation process, but it also made political participation at local level more dynamic, holding regular local elections and lending viability to local political agendas.

Despite participation in local elections tending to increase, in comparison to what happens with general elections (Forquilha 2015), the decentralisation experience over the last twenty years has not necessarily extended the participation arenas. In fact, not only are the participation fora, particularly local councils and development observatories, still merely consultative in nature, where participation is more of a formality than a reality, but the debate on local issues at municipal assemblies has also been a limited, and often politicised, exercise (ANAMM and World Bank 2009). In addition, the experience of municipalising the country shows that the connection between the residents and the municipal representatives is tenuous (Forquilha et al. 2018, 2019), which negatively affects accountability. This tenuous connection comes mainly as a result of the proportional representation system in force in Mozambique. As pointed out by Brito:

The proportional representation system devalues the ties between the elected representatives and the voters. […]. This is due to the fact that the election of representatives is not based on who they are but rather, in the Mozambican case, according to closed party lists. Therefore, the representatives are not individually dependent on the trust of the voters, but mainly, their possible re-election is dependent on their good relationship with the party structures or their party colleagues, who, in the midst of the party apparatus, can influence their inclusion on the lists and in a suitable position. To sum up, serving the interests of the party and its leaders prevails over any interests of the voters.

(Brito 2009: 25)

In this context, municipal representatives are accountable more to party leadership than to the local residents. This explains, for example, the supposed resignation of the Mayors of Quelimane, Pemba, and Cuamba in 2011, in a process where FRELIMO played a crucial role in the decision of these mayors (Forquilha 2015). Indeed, the internal pressure coming from the FRELIMO party structures led the mayors of these municipalities to publicly resign from their functions.

3 Political Stability, Violence, and the Legitimacy of the State

Political violence has been a recurring factor in the last fifty years of the history of Mozambique (fight for national liberation; the war of the 1980s; post-election conflicts and, more recently, the conflict in Cabo Delgado Province). Since national independence in 1975, the nature of the state (the fact that the state is constitutionally defined as secular and not religious) has not been seriously questioned by any particular group, despite its absence in certain parts of the country, particularly the most remote rural areas. However, the fact that the Mozambican political system is founded on a democracy based on the ‘winner takes all’ principle, allied to the poor performance of the state in terms of providing basic services in a context of a divided society (Lijphart 2008), puts the country in a situation of high risk of the occurrence of politically motivated violence and political instability. Arguably, decentralisation could play an important role in integrating differences as a way of reducing the potential for conflict.

4 State Capacity and Independence from Private Interests

In Mozambique, the poor capacity of the state and the lack of independence from private interests have been some of the greatest challenges when it comes to the process of policy design and implementation. With regard to decentralisation, for example, state capture by the political interests of the ruling party has delayed the implementation of reforms and, in some cases, made it impossible. This is the case, for example, in the adoption of the principle of gradualism in terms of the municipalisation of the country, that is the gradual municipalisation of the entire country and the gradual transfer of functions and powers from the central state to local authorities. This explains why, twenty years after the beginning of municipalisation, the country still has only fifty-three local authorities, and many functions and powers have yet to be transferred from the central state to the municipalities.

5 Sovereignty and Independence

One of the recurring themes in the debates that led to the decentralisation reforms in Mozambique is the affirmation of the unitary nature of the Mozambican state and the idea of strengthening national unity, aimed at not calling the sovereignty⁵ and independence of the country into question. In this context, for the more conservative circles in FRELIMO, the idea of creating autonomous entities through the municipalisation process should be approached cautiously. This partly explains the adoption of the principle of gradualism, its incorporation into the legislation on decentralisation, and the creation of the controversial figure of state secretary in the context of decentralised provincial government. It is important to mention that the municipalisation experience in the country has shown that gradualism and the creation of the figure of the state representative in municipal areas and, later, the creation of districts in the provincial capitals brought conflicts, which had an impact on the process of providing public services at the local level.

Since the late 1990s, the process of implementation of decentralisation reforms has had strong technical and financial support from donors in interventions such as the national planning and decentralised finance programme and the different municipal support programmes (Weimer 2012; Bunk 2018). In these programmes, the government is seeking to play a delicate game aimed at ensuring the support of donors on the one hand while, on the other, having room to manoeuvre independently as a sovereign state. This delicate game has sometimes affected the process of implementation of the programmes, with cases of programmes ending before their time, as happened with the Municipal Development Programme (PRODEM).

B Municipalisation, District Development Fund (‘7 Million’)⁶ and Challenges of Decentralised Provincial Governance

In this section, three cases will be discussed to illustrate the institutional dynamics in the decentralisation reforms. These are the municipalisation, the district development fund (‘7 million’), and the challenges of decentralised provincial governance. By taking these three cases, the aim is to analyse how institutional dynamics shape the results of decentralisation reforms from a political and economic point of view.

1 The Municipalisation Process: Between Breakthroughs and Setbacks

The first attempt at municipalisation in Mozambique after independence came in 1994 with the approval of the law no. 3/94, of 13 September, on municipal districts. Approved by the then People’s Assembly, before the first multiparty election in the history of the country, the law no. 3/94, of 13 September, called for municipalisation of the entire country. In practice, it was a question of transforming all districts in the country into municipalised spaces with regular elections. But this law was repealed even before its implementation. In its place, three years later, the law no. 2/97, of 18 February, on local authorities, was approved, introducing functional and geographic gradualism into the process of municipalisation of the country. As a result, the first thirty-three local authorities were set up and the first local election was held in 1998. Obeying the principle of gradualism, twenty more local authorities were set up in the following years, making a total of fifty-three local authorities, with regular elections every five years. According to the legislation, more local authorities could be set up once the conditions for these were in place, based on the criteria established, namely:

1. a) geographic, demographic, economic, social, cultural, and administrative factors;

2. b) national interests or the interests of the place in question;

3. c) reasons of a historical and cultural nature; and

4. d) assessment of the financial capacity for engaging in the tasks assigned to them (article 5 of the law no. 2/97, of 18 February).

However, in practice, the criteria contained in the legislation for setting up new local authorities have, over the years, shown themselves to be vague, giving plenty of room for manoeuvre to the central government, which is responsible for proposing the list of new local authorities to parliament. In fact, a study of the ten years of municipalisation in Mozambique made reference to this aspect in the following terms:

The policy framework for decentralization has not been transparent. Despite gradualism being an essential principle of local authority reform in Mozambique, there has been little transparency in the analysis and discussion process for the selection of new municipalities. The technical basis for planning the continued expansion of the local governance system is not defined using a clear and repeatable methodology that establishes standards and goals for the creation of new municipal towns and introducing small, rural local authorities (autarquias de povoação).

(ANAMM and World Bank 2009: 11)

The process of setting up the local authorities has been marked by conflicts in the context of space occupation and political influence from the main political players, namely FRELIMO, RENAMO, and the Mozambique Democratic Movement (MDM). But it is not only in geographic gradualism that there is little transparency and, through this, conflicts between the main political players. The same is true for functional gradualism, that is the transfer of functions and powers from the central state to local authorities. In fact, in 2006, the Government of Mozambique approved the decree no. 33/2006, of 30 August, which calls for the gradual transfer of functions and powers from the central state to local authorities. Article 5 of the decree provides the following:

The transfer of functions and powers from state bodies to local authorities should be gradual in order to allow for the creation and consolidation of the necessary technical, human and financial capacities in the local bodies; 3. The set of functions and powers established in this decree shall be progressively transferred to the local authorities in the three years subsequent to its coming into force. This may be extended for a further two years.

(decree no. 33/2006, of 30 August)

More than ten years later, the implementation of this decree has been marked by huge conflicts insofar as the state has been reluctant to transfer functions and powers to the local authorities, arguing that the local authorities still do not have the capacity to receive the services to be transferred by the state, particularly in the areas of health and education.

Apart from territorial and functional gradualism, another aspect that significantly affects the implementation of the municipalisation process is funding for the local authorities. The historical trajectory of the process of political institution building in the country has been marked by a strong centralising rationale, erecting barriers to the funding mechanisms for the local authorities. In fact, the law no. 1/2008, of 16 January, which defines the financial, budget, and patrimonial arrangements for the local authorities and the local authority tax system, does not facilitate funding adjusted to the needs of the local authorities. Not only are the intergovernmental transfers provided for in the law a far cry from the reality. The local authority tax system also has severe limitations, which are a consequence of the highly centralised nature of the state tax system. To date, the main intergovernmental transfers to local authorities are: (a) the Municipal Compensation Fund (FCA); (b) the Investment Fund for Local Initiatives (FIIL); (c) the Fund for the Reduction of Urban Poverty (PERPU); and d) the Roads Fund (Weimer and Carrilho 2017; Schiller et al. 2018).

With regard to the FCA, article 43 of the law no. 1/2008, of 16 January, states that:

The Municipal Compensation Fund is a fund aimed at complementing the budgetary resources of local authorities; The amount of the Municipal Compensation Fund is subject to its own funding, which is presented in the state budget, and consists of 1.5% of the tax revenue provided for in the financial year in question.

(article 43 of the law no. 1/2008, of 16 January)

In practice, however, the FCA and other state transfers represent the revenue with the heaviest weight on funding budgets of local authorities, particularly small- and medium-sized local authorities, namely Type D municipalities and towns,⁷ which reveals high financial dependence of the local authorities on the central state. But it is important to mention that state transfers to local authorities have been marked by considerable variations over the years, as shown in Figure 7.1, which has put many local authorities into financial difficulties, given the low volume of own revenue in the funding of the budgets in question.

Figure 7.1 Transfers to local authorities as a percentage of national tax revenue (2009–2016)

Source: Authors’ calculation based on Ilal and Weimer (2018).

Contrary to what might have been expected, the local authority tax system has not been consistent and coherent enough to deal with the low volume of own revenue. In fact, when you look at the collection process for own revenue by local authorities, it becomes clear that:

The central powers in terms of taxation continue to lie with the central state. In certain cases, tax rates are set centrally, and the same is the case for the definition of the tax bases and other more general principles of taxation, established on the general principles and regulations of the Mozambican tax system. Along these lines, the degree of local autonomy in the definition of fiscal policy is much greater for tariffs and fees paid for services and for fines than for taxes.

(Schiller et al., 2018: 36–37)

In this context, the lack of local fiscal autonomy puts a financial stranglehold on the municipalities, and this has a significant impact on the funding of local basic services and weakens the decentralisation reforms, particularly with regard to the municipalisation process.

What institutional dynamics are present in the case of the municipalisation, and what dimensions of institutions are these dynamics linked to? How do these dynamics affect the outcomes of municipalisation? Based on the implementation process described above, there are three dimensions of institutions, whose dynamics shape the results of municipalisation reform. These dimensions of institutions are: (a) rule of law; (b) state capacity and independence from private interests; and (c) sovereignty and independence.

Concerning the first dimension of institutions – rule of law – the process of municipalisation brings to light institutional dynamics, which show the way difficulties linked to separation of powers hamper policy design and implementation. In fact, as happens with other reforms, the debate on relevant aspects of the municipalisation process in the country is structured and dominated by the positions of the executive power, which ultimately represents the interests of the ruling party – FRELIMO. Since then, and because of the proportional electoral system in force, where the representatives are elected from a party list,⁸ and the parliamentary majorities that FRELIMO has had since the first multiparty elections in 1994, parliament has taken on a minor role in the debate, very often limiting itself to ‘rubber-stamping’ the government positions in this matter. An example of this is the debate on the geographic and functional gradualism mentioned above and the process of creating new local authorities.

As to the second dimension of institutions – state capacity and independence from private interests – institutional dynamics in this dimension show how the process of implementation of the reforms, in the context of municipalisation, is captured by party political private interests and, in some cases, delays the implementation of the reforms. This is the case, for example, with the delay in implementing the decree no. 33/2006, of 30 August, on the gradual transfer of functions and powers from the central state to local authorities. As mentioned above, this process is marked by conflict, which has been most visible in the municipalities that are ruled by political parties other than FRELIMO, namely the MDM and RENAMO (the cases of Beira, Quelimane, and Nampula municipalities). In this context, the political interests of the ruling party (FRELIMO) significantly structure the functioning of the state in terms of transfer of functions and powers to the local authorities, as also happens with the process of setting up new local authorities. This makes the process of implementation of the reforms more difficult and shape the outcomes of municipalisation, particularly when it comes to improving basic services (education, health, sanitation, transport) and strengthening democracy.

Finally, the dimensions of sovereignty and independence are probably the ones that stand out the most in the decentralisation reforms, particularly when it comes to municipalisation. It should be reiterated that the concept of sovereignty in this chapter draws on political science literature, and refers to the holding of authority by a state within a territory and over its population (Caramani 2017; Poggi 2017; Sørensen 2017). From this perspective, sovereignty applies both domestically (the need for the state to exert authority within its territory and over its populations) and externally (in the context of international relations – non-acceptance of any interference from other states with regard to domestic affairs). In a decidedly centralised state, as was mentioned above, institutional dynamics linked to sovereignty and independence show how the debate on the municipalisation process was very much structured around the idea of the need to preserve the unitary nature of the state, in accordance with the Constitution of the Republic (article 8 of the Constitution of the Republic, as amended by the law no. 1/2018, of 12 June). In this sense, the choice of territorial and functional gradualism (gradual creation of the local authorities and gradual transfer of functions and powers from the central state to the local authorities) comes essentially as a result of institutional dynamics linked to sovereignty and independence. The same can be said about the low fiscal autonomy of the municipalities, the creation of the figure of a state representative in the municipalities and the establishment of districts in all provincial capitals, leading to territorial overlaps (municipalities/districts) and, in some cases, conflicts between the mayors and district administrators (Forquilha 2016).

In addition, decentralisation is one of the areas that has had significant support from international cooperation partners, both in the area of deconcentration and the area of devolution (Borowczak and Weimer 2012; Weimer and Carrilho 2017; Bunk 2018). In fact, since the late 1990s, bilateral and multilateral partners have been supporting decentralised planning and finance programmes, as well as municipal development. This is the case, for example, of the National Programme for Decentralized Planning and Finance (PNPFD); Thirteen-City Support Programme (P 13); Decentralization and Municipalization Support Programme (PADEM); Support Programme for Districts and Municipalities (PADM), Local Authority Development Programme (PDA); PRODEM. In all of these decentralisation support programmes/projects, the government has been a key partner in the implementation process, playing a delicate game while at the same time ensuring technical and financial support and maximising any room for manoeuvre with a view to affirming country’s sovereignty. In some cases, this delicate game has resulted in conflict between the government and the cooperation partners, significantly affecting the implementation process and the goals of the decentralisation support programmes.⁹

2 The District Development Fund (‘7 Million’): Between Discourse and Practice

Introduced into governance practice two years after Armando Guebuza took office in his first term (2005–09), the district development fund is one of the most important reforms at district level in the context of administrative decentralisation. In fact, for the first time, with the approval of the law no. 12/2005, of 23 December, on the state budget for 2006, each district was given a district investment budget of 7 million meticais, the equivalent at the time of around US$250,000. This decision by the government of Armando Guebuza was warmly welcomed initially, particularly by the district governments, cooperation partners, civil society, and the local populations in general, insofar as the resources made available could be used for putting the District Strategic Development Plans (PEDD) and the District Economic and Social Plans and Budget (PESOD) into practice in the context of the decentralisation reforms, with the involvement of local councils – an institution set up under the scope of the Local Government Act (LOLE), approved in 2003 (Law No. 8/2003, of 19 May). Initially called the Local Initiative Investment Budget (OIIL), the 7 million later became known as the District Development Fund (FDD), with the approval of decree no. 90/2009, of 31 December.

For many stakeholders, the availability of a specific sum, the use of which was completely dependent on the decision of the districts, was an important step in making one of the main aims of decentralisation a reality, local development. However, in many cases, the lack of clarity as to the criteria for using these resources was visible. In fact, the first year of implementation of the reform was marked by disparity in terms of projects funded by the district governments, insofar as ‘in some cases, there were administrators who renovated district palaces […] and other cases where local police stations and Frelimo party headquarters were renovated’ (Forquilha 2010b: 37). Given this lack of clarity in the use of the resources provided under the scope of the ‘7 million’, President Guebuza began giving concrete guidelines in the country’s provinces, underlining that the money had to be used for income generation, job creation, and food production. At the meeting of the council of ministers, extended to provincial governors, district administrators and FRELIMO officials, which took place in August 2006, President Guebuza once again insisted on the idea of creating wealth locally. After this, the guidelines for the preparation of PESOD for 2008, issued jointly by the Ministry of Planning and Development (MPD) and the Ministry of Finance (MF), gave instructions as to how the ‘7 million’ was to be used, incorporating the will of the President of the Republic – income generation, job creation, and food production. Although these instructions were in line with the idea of local development, for many administrators this change was hard to manage, insofar as an important part of the PESOD planning activities was left without funding. This difficulty was plain to be seen, for example, in the words of the former administrator of Metarica, in Niassa Province. He said:

[…] now that the ‘7 million’ is for food, income generation and jobs, I don’t know how the district government will be able to pay for holding district consultative council sessions or how the projects approved by the consultative council for this year will be funded … I think we’ll have to go back to planning like we used to, without consulting the people, without the consultative councils […].

(Forquilha 2010b: 37)

With the change in the guidelines on the use of the money, the ‘7 million’ began operating along microcredit lines, managed by the district governments. Funding would be allocated to borrowers who presented projects in line with the goals of the fund, namely income generation, job creation, and food production. The borrowers would repay the amount allocated with interest. Initially, the interest rate varied from district to district, before the regulation introduced by the Ministry of State Administration (MAE) in 2011, when interest rates varied ‘between 3% and 7% per month, i.e. 36%–84% per year’ (Weimer and Carrilho 2017: 85). In theory, the money repaid should have gone back into the fund, with the aim of extending access to it to more citizens locally. But, in the early years of the initiative (i.e. 2006–08), the money repaid fell very short of contributing to the revolving nature of the fund (see Table 7.1).

Table 7.1 OIIL repayments up to the first half of 2008

N/O	Provinces	Ceiling in MZM	Paid in MZM	Repaid in MZM	%
1	Niassa	110,632,870.00	81,452,030.00	1,782.67	2.2
2	Cabo Delgado	133,890,820.00	111,287,000.00	1,409.48	1.3
3	Nampula	166,679,180.00	139,834.980.00	3,533.27	2.5
4	Zambezia	172,272,600.00	130,640.800.00	4,394.25	3.4
5	Tete	139,696,240.00	73,903.000.00	1,500	0.002
6	Manica	96,058,850.00	62,898.500.00	199.2	0.3
7	Sofala	89,842,920.00	89,842.920.00	4,038.95	4.5
8	Inhambane	99,563,240.00	61,744.980.00	1,540.89	2.5
9	Gaza	85,308,880.00	85,308.880.00	1090.94	1.3
10	Maputo	53,817,000.00	27,127.300.00	15.75	0.6
Total		1,147,772.600.00	864,040,390	19,524.65	1.8

Source: MPD (2008).

Note: The decimal separators used in the original have been maintained.

In practice, there were enormous difficulties in that not only did the borrowers have trouble repaying as scheduled the amounts allocated, but also many of the projects approved and funded were not economically viable and did not have any clear connection to the district plans, namely the PEDD and the PESOD. In its debriefing regarding the early years of the implementation of the OIIL, the Mozambique government acknowledged that the process was facing constraints and challenges, namely:

1. (1) The non-formalisation of the procedures for the implementation of the OIIL is pointed to by almost all the participants as one of the main constraints to this process;

2. (2) The lack of full compliance with the decisions made at the 1st session of the council of ministers, extended to provincial governors, district administrators, and other state and government officials, held on 11, 12, and 13 August 2006, considering that a large part of the process of prioritising and allocating resources had been concluded, particularly in 2006;

3. (3) The absence of systematic monitoring and following up on the projects funded, throughout the country, at central, provincial, and even district level;

4. (4) Lack of business management experience of the beneficiaries, linked to the poor design and feasibility of the projects submitted for approval;

5. (5) Limited capacity of the bodies involved in the process of selection and approval of projects;

6. (6) Absence of contracts between the districts and the borrowers, resulting in the absence of a repayment plan and disparity in the setting of interest rates and, where these exist, they are unclear as to the obligations of the borrowers;

7. (7) Difficulties with general repayments considering the nature (unprecedented and innovative) of the OIIL, allied to the local authorities’ lack of experience in dealing with activities involving credit systems;

8. (8) Non-observance of connection elements, synergies, and consistency between projects, resulting in a weak production value chain and the subsequent underutilisation of the existing potential;

9. (9) Little or no involvement of local councils in the monitoring carried out in their areas of jurisdiction (MPD 2008: 14–16).

These constraints and challenges, on the one hand, and the administrators’ difficulties in managing the change introduced by the President of the Republic, on the other, clearly show how institutional dynamics shape the results of the reforms. What institutional dynamics are present in the case of the ‘7 million’, and what dimensions of institutions are these dynamics linked to? How do these dynamics affect the results of the ‘7 million’? Based on the implementation process described above, there are three institutional dimensions, whose dynamics shape the results of the ‘7 million’ reform: (a) rule of law; (b) voice, participation, and political accountability; and (c) state capacity and independence from private interests.

With regard to the rule of law, institutional dynamics from this dimension show that although the separation of powers is enshrined in the constitution, there is a clear dominance of the executive power over the other two, particularly when it comes to designing and implementing reforms and policies. In many cases, the reforms are the sole initiative of the executive and, sometimes, of the actual President of the Republic, as is the case of the ‘7 million’, without consultation and much less involvement of relevant actors. In this context, the ministers, who have to offer technical support to the reforms, limit themselves to finding mechanisms for making the reforms feasible, in the sense of making mere political guidelines a reality, entirely dependent on the President of the Republic. As a result, not only do the reforms become difficult to implement. They also become a mere political instrument for gaining ground and extending political influence, relegating issues of local development to second place. The ‘7 million’ is a clear example of this. In fact, different research carried out on the subject shows that the ‘7 million’ became an instrument for gaining ground locally, to the detriment of giving impetus to the local development process (Forquilha 2010b; Orre and Forquilha 2012; Sande 2011; Weimer and Carrilho 2017). In addition, the rules and regulations produced under the scope of the operationalisation of the ‘7 million’ were not duly observed. Indeed, not only were there many cases where the allocation of money was not governed by any contractual ties between the state and the borrowers, but many of the procedures linked to management of the fund lacked formalisation, particularly in the early years of implementation of the initiative (MPD 2008). Consequently, the process of implementation was marked by differences in criteria and the treatment of the borrowers, particularly in terms of fund repayment, insofar as, throughout the country, there were differences in the interest rates set for repayments and few borrowers were able to repay the state for the funds received (MPD 2008), a fact that contradicted the goals and the role of the ‘7 million’ in the context of local development.

With regard to the second dimension of institutions – voice, participation, and political accountability – the ‘7 million’ implementation process brings to light institutional dynamics that show poor citizen involvement in the process of designing and implementing policies. Acclaimed as an important instrument for giving impetus to community participation in the local development process through the so-called local councils,¹⁰ the ‘7 million’ did not in fact broaden the local participation base. Indeed, not only were the local councils politically ‘captured’, reinforcing political and economic exclusion locally. They also failed to properly monitor the fund allocation process and the implementation of funded projects (MPD 2008). Moreover, the accountability dimension was practically absent from the process of implementation of the ‘7 million’. This was partly a result of the poor regulation and observance of the rules regarding the operationalisation of the fund, as mentioned above when speaking of the rule of law.

As to the third dimension of institutions – state capacity and independence from private interests – the case of the ‘7 million’ brings to the fore institutional dynamics that make clear that poor state capacity, particularly when it comes to its poor institutionalisation and capture by party political interests, turned the reforms into an instrument for broadening FRELIMO’s base of influence, particularly in the rural context (Orre and Forquilha 2012). Interviews held in Chimbunila District in Niassa Province in 2018, in a research project on social cohesion and political violence, for example, show how FRELIMO used the ‘7 million’ to broaden their local base, in a clientelism rationale, setting issues of local development aside (Forquilha and Goncalves 2019).

3 The Challenges of Decentralised Provincial Governance

Following the results of the general election in 2014, where RENAMO won in some provinces in the centre and north, the country entered another political and military crisis and a lengthy negotiation process insofar as RENAMO claimed its right to govern in those provinces. In this context, decentralisation was one of the most important issues on the negotiating table. After attempts made by negotiating teams involving national and foreign personalities, President Filipe Nyusi, representing the Mozambique government, and President Afonso Dhlakama, representing RENAMO, decided to make more direct contact to negotiate the crisis on the facilitation of the so-called contact group, made up of representatives of some countries who are cooperation partners of Mozambique. After a long period of negotiation, marked by a certain amount of secrecy regarding the details of the issues discussed, particularly decentralisation, in February 2018 an understanding was reached between the government and RENAMO, which included a constitutional amendment and the subsequent production of legislation aimed at the introduction of Decentralized Provincial Governance Bodies (OGDPs).

In a process where parliament played only a minor role, insofar as it did not produce any profound or significant debate on the matter, limiting itself to ‘rubber-stamping’ the understandings that had been reached between the leaders (government and RENAMO), the new decentralisation package was approved before the October 2019 election. These are the Laws Nos. 3/2019, 4/2019, 5/2019, 6/2019, and 7/2019, all of 31 May, and the decree no. 2/2020, of 8 January, as mentioned above. In this context, the decentralisation package brings with it some important challenges, namely:

1. a) Little clarity regarding fiscal decentralisation: how can OGDPs be assured of a tax base consistent and coherent with their duties and responsibilities in the context of a highly centralised state taxation system? How can it be assured that the allocation of state resources via intergovernmental transfers to OGDPs are in line with the decentralisation rationale and it does not reproduce what happens already in the municipalities and districts (few resources channelled locally)¹¹?

2. b) Overlapping of some duties between the provincial directorates (OGDPs) and the state provincial services (OREPs);

3. c) Potential for conflict in the operation of OGDPs and OREPs;

4. d) Increase in the weight of the party machines on the functioning of OGDPs, making these more accountable to their political parties than to the voters.

Despite the challenges mentioned above, the new decentralisation package could have been not only an important means for improving public services and reinforcing democracy, but also a fundamental way of ensuring that the political institutions better reflected the heterogeneous nature of the country, its differences, thus reducing the potential for violent political conflict.

What institutional dynamics are at play in the case of decentralised provincial governance, and what dimensions of institutions are these dynamics linked to? How do these dynamics affect the outcomes of the reforms? Although the reforms regarding decentralised provincial governance in the country are recent, their implementation process brings to light two dimensions of institutions: (a) state capacity and independence from private interests and (b) political stability, violence, and the legitimacy of the state.

With regard to state capacity and independence from private interests, the process that led to the approval of the decentralisation package in 2019 shows once again how group interests, in this case political groups (FRELIMO and RENAMO), supersede the interests of the state in terms of development and the consolidation of democracy. In fact, although the issue of the election of provincial governors was raised following the claims from RENAMO in the context of the 2014 election results, it ended up becoming a potential instrument for pandering to the elite. By making the issue of the appointment of provincial governors a demand for its acceptance of the 2014 election results, RENAMO did not necessarily have in mind decentralisation as a mechanism for reinforcing local democracy and improving the provision of public services. Indeed, it is important to remember that before the demand for the appointment of provincial governors, one of RENAMO’s conditions for accepting the election result was the formation of a ‘caretaker governmenť, a request that was rejected by the Mozambican parliament in November 2014, with the majority vote of FRELIMO (DW 2014).

It was after this rejection that RENAMO began demanding the appointment of governors in the provinces where it had won the election. Soon afterwards, this demand would be transformed into a draft bill on provincial authorities, rejected by the Mozambican parliament in April 2015 (RTP 2015). The months following the rejection of the draft bill on provincial authorities were marked by episodes of military tension, culminating in a return to armed violence, which continued until an understanding was reached between the leaders of the government and of RENAMO in February 2018, followed by the amendment to the constitution and the approval of the decentralisation package in 2019. In turn, FRELIMO, through the government, tried to ensure the reinforcement of its control and political influence locally, introducing provincial state representation bodies (OREPs) into the decentralisation package. In practical terms, these operate like parallel structures to the decentralised provincial governance bodies (OGDPs). In this way, the interests of RENAMO and FRELIMO ended up significantly conditioning not only how the debate on decentralised provincial governance was conducted, but also the process of implementation of the actual reforms, subordinating the issues of funding improvements in the provision of public services and local development associated with the decentralisation process.

As to political stability, violence, and the legitimacy of the state, institutional dynamics show that the historical trajectory of the process of political institution building in Mozambique has been marked by armed violence contesting the legitimacy of the state. This has significantly affected the process of designing and implementing policies, as is the case of the decentralisation package. As mentioned above, the debate and approval of the reforms regarding decentralised provincial governance took place in the context of finding a solution for the conflict between RENAMO and the Government of Mozambique following the 2014 election results. In this sense, rather than being a mechanism aimed at improving the provision of services and strengthening democracy, the decentralisation package acted as a way of accommodating political, social, and economic differences, minimising the potential for conflict and thus contributing to the stability of the country and the legitimacy of the state. However, the reforms approved and particularly their implementation process in the context of the decentralisation package fall very short of minimising conflict.

In fact, apart from being costly, insofar as the state budget will now be funding parallel provincial governance structures (OREPs and OGDPs), the decentralisation package contains contradictions and major potential for conflict, as mentioned above when speaking of some of the challenges of decentralised provincial governance. Indeed, the early days of operationalisation of this model, after the elected provincial governors and the provincial secretaries of state, appointed by the President of the Republic, had taken office, were marked by confusion regarding the operational scope of the elected provincial governors and the appointed provincial secretaries of state. Following this, on 29 January 2020, the Ministry of State Administration and Public Service found itself obliged to publish a circular entitled ‘operationalization of state representation in the province and of the decentralized provincial governance bodies’ (circular 9/MAEFP/GM-DNAL/214/2020), with a view to clarifying the operational scope of OREPs and OGDPs. This conflict would have been even greater if the 2019 election results had dictated victory for the opposition parties (RENAMO and MDM) in some provinces.

IV Conclusion

More than twenty years on, how can the modest impact of the decentralisation reforms on the promotion of local development and the strengthening and consolidation of democracy in Mozambique be explained? Throughout this chapter, we have sought to answer this question based on institutional dynamics linked to the main dimensions of institutions identified above. The chapter underlines the idea according to which the results of the decentralisation reforms are shaped by the nature of the political system and the way it operates, in the context of institutional dynamics. From the cases of the district development fund (‘7 million’ fund), the municipalisation process and the challenges of decentralised provincial governance, on the one hand and, on the other, exploring institutional dynamics linked to five dimensions of institutions, with a particular emphasis on the rule of law, state capacity and autonomy from group interests, sovereignty and independence, the chapter sought to show how the institutions structure and shape the results of decentralisation reforms.

The analysis developed throughout the chapter showed that the five dimensions of institutions do not generate institutional dynamics that affect in the same way the results of the three reforms discussed. In fact, for the case of municipalisation, the process of implementation of the reforms in this area shows that the institutional dynamics that shape the results of the reforms are linked to three main dimensions of institutions. The first dimension is the rule of law, seen in the poor functioning of the system of checks and balances, placing the executive power (government) in a dominant position to the detriment of the legislative power (parliament) in the context of policies design and implementation. An example of this is the case of the adoption and implementation of territorial and functional gradualism, where parliament played a clearly minor role.

The second dimension of institutions, whose dynamics shape municipalisation reforms is state capacity and independence from private interests, which can be seen in the capture of the reforms with a view to promoting the interests of a political group. The process of implementation of the transfer of functions and powers from the central state to local authorities clearly shows how private interests of a political nature delay and, in a way, block the implementation of reforms. The third dimension of institutions with dynamics that affect the results of municipalisation is sovereignty and independence. In fact, the fear of autonomy of the local authorities that would affect the ideal of the unitary state significantly conditioned how the process of municipalisation is being implemented – gradually, from a territorial point of view and the transfer of functions and powers from the central state to the local authorities.

With regard to the district development fund (‘7 million’ fund), there are three dimensions of institutions, whose institutional dynamics are shown to be relevant. The first is the rule of law with an institutional dynamic marked by weak enforcement. As a result, the reforms became difficult to implement and were turned into a mere political instrument for gaining ground and extending political influence, relegating issues of local development to second place. The second dimension is voice, participation, and political accountability. Contrary to what might be thought, institutional dynamics show that the ‘7 million’ did not broaden the participation base in local communities, due to the political capture of the local councils, which reinforced political exclusion. Not only did this fail to reinforce democracy. Issues of local development also became secondary. Finally, the third dimension is state capacity and independence from private interests. The dynamics analysed in this chapter show how the weak institutionalisation of the state led to the capture of the ‘7 million’ by political party interests, which allowed FRELIMO to broaden its base of influence, particularly in rural areas.

Finally, concerning the case of decentralised provincial governance, the analysis suggests that there are two main dimensions underlying institutional dynamics that come from the implementation process. The first is state capacity and independence from private interests, which can be seen in the way the reforms were negotiated, approved and are being implemented. The second dimension is political stability, violence, and the legitimacy of the state. Indeed, the whole process of design and approval of the decentralisation package, which introduced decentralised provincial governance reforms, is deeply linked to the need of accommodating political differences, minimising the potential for violent conflict and increasing the legitimacy of the state.

Of the institutional dimensions mentioned above, state capacity and independence from private interests, particularly political groups, stand out in the three reforms analysed throughout this chapter, namely municipalisation, the district development fund (‘7 million’), and decentralized provincial government. In this context, the reforms develop according to group interests, particularly political party interests, which capture the state and use the reforms as a mechanism for maintaining and bolstering political power. In this regard, rather than being a means of improving the provision of public services and strengthening democracy, decentralisation works more as an instrument for reinforcing state control and pandering to the elite. This is probably the biggest challenge decentralisation is facing in Mozambique, therefore making it a fundamental issue to be taken into account in any reform in this area, within the context of strengthening democracy and promoting local development.