November 23, 2018

Understanding Decentralization in Mozambique with a Ziblatt-Mann Framework

By Coline Ferrant

Revisiting the institutionalization of the Mozambican state

In this article, I examine decentralization reforms in Mozambique since 1994, through an analysis of primary and secondary sources, and a framework that draws from the theses of political scientist Daniel Ziblatt and sociologist Michael Mann.

Ziblatt studies the formation of nation-states, specifically, the adoption of a unitary or federal structure. Federalism emerges when central authorities succeed in negotiating with national subunits that exert infrastructural power, that is, “the ability to regulate society, to tax, to maintain order” (Ziblatt, 2006: 16). In this article, I do not address the formation of the Mozambican state, but its institutionalization: its capacity to structure, centralize, and build resources, by negotiating with actors that exert infrastructural power.

Ziblatt borrows the concept of infrastructural power from Mann. In his four-volume “history and theory of power relations in human societies” (Mann, 1986: 1), Mann conceptualizes power as the capacity to control resources. He distinguishes four sources of power associated with four types of resources: economic, ideological, military, and political. Political power is divided into two types: despotic and infrastructural. An actor exerts infrastructural power when it needs routinized and institutionalized negotiation; conversely, it exerts despotic power when it does not need routinized and institutionalized negotiation. Infrastructural power is the negotiated capacity to control political resources. It does not only stem from the state to society, but also from society to the state, especially through parties and pressure groups (Mann, 1984, 1986, 2008).

Applying a Ziblatt-Mann framework to Mozambican decentralization reforms raises four issues: the difficult transposition of European frameworks, the importance of external actors, the limited autonomy of the state from the ruling party, and the weakness of society towards the state.

Numerous Africanists caution against reckless exportations of analytic frameworks elaborated for the singular European socio-historical trajectory (Herbst, 2000; Hyden, 2006; Young, 2012). Mann states that infrastructural power is characteristic of industrialized societies (Mann, 1984: 189) and, in African societies, faces two challenges: the difficult access to hinterlands and the preeminence of military power (Mann, 2008: 364-365). Ziblatt mentions that the formation of African and Latin American states is bound to external actors, namely, colonizers (Ziblatt, 2004: 97-98).

This last point echoes a frequent criticism of Ziblatt’s framework: its inattention to the international geopolitical context (Davidson-Schmich, 2007; Prasad, 2007). The Mozambican state is, indeed, deeply tied to external actors. Afflicted by a civil war from 1977 (i.e., two years after independence) to 1992 between ruling Frelimo (Frente de Libertação de Moçambique) and opposing Renamo (Resistência Nacional Moçambicana), the country is now touted as a success story of the reconstruction of a pacific, democratic state with the help of the international community (Alden, 2001; Olowu et Wunsch, 2004; Hyden, 2006; Manning, 2008).

International donors praise decentralization policies, which they frame as pacific, depoliticized antidotes for development and democratic local governance (Alexander, 1997; Olowu and Wunsch, 2004; Kyed and Buur, 2006; de Matos Fernandes, 2009). Following the 1992 peace settlement, several donors, including Denmark, Switzerland, the Netherlands, and the World Bank, made their funding conditional upon the implementation of a decentralization process (Kathyola and Job, 2011). The first decentralization reforms occurred in 1994, a little before the first post-conflict elections. Yet, such quick implementation does violence to “a lesson potentially relevant for any decentralization effort”; for this to be successful, state subunits should have sufficient infrastructural capacities (Ziblatt, 2004: 98).

This leads to a last puzzle: the relevance of the concept of civil society in Mozambique. A Ziblatt-Mann framework involves negotiations between the state and social actors that have infrastructural capacities. Yet, 17 years (1975-1992) of centralized and authoritarian Frelimo rule have shattered any social actor outside the party-state (Alden, 2001; Kyed and Buur, 2006). The state is not autonomous, because it is conflated with a social actor (i.e., Frelimo); barely any other social actor exerts infrastructural power.

I explore decentralization reforms in Mozambique since 1994 following an analytic framework that amends Ziblatt’s and Mann’s frameworks with the four aforementioned puzzles. As a prerequisite, I delineate the institutional and historical context of decentralization reforms. I show that formal decentralization obfuscates a process of state centralization, through the cooptation of the very actors decentralization reforms supposedly empower. I explain two mechanisms absent from a Ziblatt-Mann framework: the conflation of the state with a party and the negotiating capacity of the international community. I conclude by assessing the inputs and limits of a Ziblatt-Mann framework to make sense of decentralization in Mozambique.

Processes and actors of an ambiguous decentralization

Decentralization consists in transferring human, financial, and institutional resources from the central government to local administrations. When these local administrations are autonomous from the central government, it is devolution; when they are dependent, it is deconcentration.

Major administrative disorganization followed the 1992 peace settlement and the 1994 first democratic elections (Alexander, 1997). In 1993 and 1994, the Ministry of State Administration (Ministério da Administração Estatal) hosted brainstorming workshops on the country’s future territorial organization, in which civil servants, traditional authorities, Frelimo and Renamo officials, and international donors participated (Kyed and Buur, 2006; de Matos Fernandes, 2009). In September 1994, shortly before the elections, these actors concurred on a first decentralization law: Lei 3/94. It created municipal districts (distritos municipais) to which were devolved the elaboration of development, cultural, and social policies, and attributed the capacity to collect taxes and participate in the decisions of the central government that affect them. Traditional authorities (autoridades tradicionais), endowed with judiciary functions and traditional legitimacy based on clan ties, saw their role officially recognized by the state (West and Kloeck-Jenson, 1999; de Sousa Santos, 2006; Hyden, 2006).

In 1997, the Frelimo government imposed a second decentralization law (Lei 2/97), despite Renamo’s opposition. This law pushed devolution backward. Municipal districts were replaced with a system of cities (cidades) and villages (vilas); these statutes were granted to only 33 entities, mostly urban. This resulted in what scholars have called “dual decentralisation” (Kathyola and Job, 2011: 173) or “mixed/partial devolution” (Olowu and Wunsch, 2004: 39): devolution for these 33 cities and villages, deconcentration for the rest of the territory, mostly rural. In 2000, another decree (Decreto 15/2000) adopted by Frelimo without consultation granted two other actors the capacity to mediate between the state and local populations: neighborhood or village secretaries (secretários de bairro ou aldeia) and leaders legitimized as such by communities or social groups (líderes legitimados como tais pelas respectivas comunidades ou grupos sociais). Tax collection and policing were formally devolved to traditional authorities.

The infrastructural and despotic state co-optation of local actors

Yet, legal texts remain unclear as for the relations between the state and local administrations. Since 2000, the state has argued for the constitutional principles of “consolidation of national unity” and “promotion of the country’s balanced development” (art. 11) to reserve the right to censure local administrations’ decisions and maintain central state’s services even when they compete with local administrations’ services (de Matos Fernandes, 2009). The state organizes the institutional and financial deprivation of decentralized entities, both directly and indirectly. First, central government transfers are extremely weak: less than 1% of the state budget on average since 1994, compared to an average of 13.8% in all the developing world (Kathyola and Job, 2011). Second, as long as decentralized entities lack infrastructural capacities, taxation duties provided by law remain unenforceable. For instance, Manica municipal district, in the center of the country, has failed to collect taxes since 1994 due to the absence of cadastral information (Kathyola and Job, 2011). As for deconcentrated services, the state deliberately lengthens the deconcentration of administrative and financial resources. For example, the Ministry of Health has still not drafted a deconcentration plan, under the pretext that districts lack human resources and that centralization allows for economies of scale (Kathyola and Job, 2011).

Further, the state coopts traditional authorities through clientelist negotiations on the one hand, and by fueling conflicts between them on the other hand. Clientelism is a social relation between a patron and a client that involves dependence and reciprocity. The state-patron negotiates with the authorities-clients the provision of authorities’ infrastructural power in exchange for financial and political benefits. Since 2000, traditional authorities retain 10% of the taxes they collect (Kathyola and Job, 2011) and their members receive allowances (Alexander, 1997). As they exert state prerogatives, they can, in Weberian terms, claim a legal authority on top of their traditional authority (Weber, 1958 [1922]). Local chiefs seek political positions for their children (de Sousa Santos, 2006) and claim advantages proper to civil service: official housing, uniforms, and wages (Alexander, 1997). In exchange, the state captures traditional authorities’ infrastructural power, namely, their territorial anchorage. Mozambique is a geographically vast, sparsely populated country, which complicates state penetration in marginalized areas. The state can thus extend its control to places formerly out of its reach (Herbst, 2000; de Sousa Santos, 2006; Kyed and Buur, 2006; de Matos Fernandes, 2009). The state relies on traditional authorities’ infrastructural power for its own institutionalization, yet deprives them of any capacity for political control of the population. Decentralization reforms do not provide for any participatory democracy device or community representation at the state level (de Sousa Santos, 2006).

In addition, the state exerts despotic power by pitting traditional authorities against one another other. Laws are vague when it comes to modes of identification of traditional authorities, which generates conflicts between rival lineages. For instance, a leadership fight between three clans left three dead between 2000 and 2002 in the northern district of Dombe (Kyed and Buur, 2006).

The capture of state resources by Frelimo

In the Ziblatt-Mann framework, the state is an autonomous entity, distinct from society. Accordingly, until now, I have talked about the state without examining it. Still, despite the establishment of democratic pluralism, Mozambique has not known alternation since 1994: Frelimo has won all presidential and legislative elections. Besides being a more efficient organization and having a stronger internal discipline than its rival Renamo (Carbone, 2005; Manning, 2008; Young, 2012), Frelimo’s omnipotence arises from a more efficient capture of state resources. I can then complicate the aforementioned processes.

Frelimo strongholds are located in the predominantly urban south, while Renamo is most present in central and northern rural areas (Kathyola and Job, 2011; Carbone, 2005; Manning, 2008; Forquilha, 2010). Civil war occurred within the context of Cold War; it opposed Frelimo Marxist regime to Renamo guerilla, supported by Rhodesian secret services and then the South African apartheid regime (Manning, 2008; Young, 2012; Kathyola and Job, 2011). Renamo built ties with traditional authorities, which Frelimo fought because it deemed them obscurantist and tied to the colonial regime (Alexander, 1997; West and Kloeck-Jenson, 1999; Kyed and Buur, 2006; Forquilha, 2010).

The processes of creation of decentralized entities and identification of traditional authorities were marred with partisan interests. The fact that decentralized entities were mostly created in urban areas reflects a balance of power in favor of Frelimo (Kathyola and Job, 2011; Galli, 2003). The cooptation of traditional authorities is bound to political and electoral strategies. Frelimo’s sudden interest in traditional authorities on the threshold of the first free elections was driven by an awareness of the traditional authorities’ capacity to mobilize voters (Forquilha, 2009; West and Kloeck-Jenson, 2009). Traditional authorities practice “political nomadism” (Forquilha, 2010: 61): they sell themselves to the highest offering patron, either Frelimo or Renamo. Renamo is thereby losing its historical predominance among traditional authorities because the ruling Frelimo patron has more resources to give them, especially positions in local administrations (West and Kloeck-Jenson, 2009; Carbone, 2005; de Sousa Santos, 2006; Forquilha, 2009).

Lastly, despite the formal separation between parties and the state, the majority of civil servants in deconcentrated services are Frelimo members (Carbone, 2005). In addition, local secretaries and leaders provided by Decreto 15/2000 are for the most part Frelimo intermediaries: they serve to compete with traditional authorities when these are deemed too close to Renamo (de Matos Fernandes, 2009; Forquilha, 2009).

The international donors as safeguards against Frelimo

The international community, including the United Nations, the World Bank, the European Union, the United States, Denmark, the United Kingdom, Sweden, Norway, the Netherlands, and various NGOs, has granted unprecedented amounts of support to development and peacekeeping. The country receives one billion dollars per year on average since 1992 (Kathyola and Job, 2011); more than half of state expenses stem from this aid (Alden, 2001; Kathyola and Job, 2011; Young, 2012). Such manna allows donors to rein in Frelimo’s endeavor to control decentralization.

Development support includes subsidies for political parties in a bid to foster competition within the party system. Renamo, as a minority party, is the principal recipient. For instance, it was granted 17 million dollars to run for 1994 elections (Alden, 2001; Carbone, 2005; Manning, 2008). NGOs such as the NIMD (Netherlands Institute for Multiparty Democracy) and the American NDI (National Democratic Institute) fund the training of Renamo local officials. Political scientist Carrie Manning reckons that Renamo would have been unable to transition from a guerrilla movement to a political party without this external support. Further, Frelimo is compelled to maintain a constant dialogue with Renamo so as to show donors proof of sustainable democratization of political life (Manning, 2008).

The Frelimo-state associates donors to the implementation and monitoring of decentralization reforms. One of the most popular means of advancement among local civil servants is to be hired in international projects (Galli, 2003). Yet, Frelimo constantly criticizes the international community for interfering in the country’s internal affairs (Alden, 2001). It attempts to coopt this aid by focusing it on its southern urban strongholds, especially the capital Maputo (Kathyola and Job, 2011). For instance, in 1998, the city of Maputo received 207.6 million dollars from external support, while the northern region of Niassa, despite having the same population (approximately 1 million inhabitants), got only 16.6 million dollars (Alden, 2001).

Frelimo and the international community exert infrastructural power because they are able to control political resources. Yet, the state does not negotiate with these actors, as Ziblatt’s thesis would suppose. Rather, these actors constrain the state, by controlling its resources (Frelimo) and, so to say, by controlling whom controls its resources (the international community). Herein lie two shortcomings of the Ziblatt-Mann framework: the absence of state autonomy for being subject to predation by an internal actor exerting infrastructural power (i.e., Frelimo), and the crucial role of external actors that are able to rein in Frelimo predation.


The Ziblatt-Mann framework proves helpful when showing that the Mozambican central state becomes institutionalized by negotiating in a clientelist way with traditional authorities that exert infrastructural power. Yet, major amendments should be made. This institutionalization also relies on state despotic power, exerted on decentralized entities and deconcentrated services. The Mozambican state is not autonomous, but subject to predation by Frelimo, to the point that state institutionalization conflates with Frelimo institutionalization. External actors are crucial: the international community, by interfering in decentralization policy-making and supporting Renamo, reins in Frelimo’s attempts to control decentralization efforts.

My amended Ziblatt-Mann framework can still gain in analytic precision and elegance by unpacking two elements. First, I can elaborate on Frelimo’s and Renamo’s internal strifes. For instance, the recognition of traditional authorities in 1994 was hotly debated among Frelimo (Forquilha, 2009). Second, I can delve into other groups in Mozambican society, even if they barely exert any infrastructural power: religious leaders (Alexander, 1997; Alden, 2001; de Sousa Santos, 2006; de Matos Fernandes, 2009), organizations of women, youth, and workers affiliated with Frelimo (Kathyola and Job, 2011), and agricultural cooperatives (Alden, 2001; Galli, 2003).


  • Alden, Chris. 2001. Mozambique and the Construction of the New African State: From Negotiations to Nation Building. New York, NY: Palgrave.
  • Alexander, Jocelyn. 1997. “The Local State in Post-War Mozambique: Political Practice and Ideas about Authority,” Africa, 67(1): 1-26.
  • Carbone, Giovanni. 2005. “Continuidade na renovação? Ten Years of Multiparty Politics in Mozambique: Roots, Evolutionand Stabilisation of the Frelimo-Renamo Party System,” The Journal of Modern African Studies, 43(3): 417-42.
  • Davidson-Schmich, Louise. 2007. “Review: Structuring the State,” German Studies Review, 30(3): 640-41.
  • De Matos Fernandes, Tiago. 2009. O poder local em Moçambique: descentralização, pluralismo jurídico e legitimação. Porto, Portugal: Edições Afrontamento.
  • De Sousa Santos, Boaventura. 2006. “The Heterogeneous State and Legal Pluralism in Mozambique,” Law & Society Review, 40(1): 39-75.
  • Forquilha, Salvador. 2010. “Chefferie traditionnelle et décentralisation au Mozambique : discours, pratiques, dynamiques locales,” Politique africaine, 117: 45-61.
  • Galli, Rosemary Elizabeth. 2003. Peoples’ Spaces and State Spaces: Land and Governance in Mozambique. Oxford, England: Lexington Books.
  • Herbst, Jeffrey. 2000. States and Power in Africa: Comparative Lessons in Authority and Control. Princeton, NJ: Princeton University Press.
  • Hyden, Goran. 2006. African Politics in Comparative Perspective. Cambridge, England: Cambridge University Press.
  • Kathyola, Janet and Oluwatoyin Job. 2011. Decentralisation in Commonwealth Africa: Experiences from Botswana, Cameroon, Ghana, Mozambique and Tanzania. London, England: Commonwealth Secretariat.
  • Kyed, Helene Maria and Lars Buur. 2006. “New Sites of Citizenship: Recognition of Traditional Authority and Group-Based Citizenship in Mozambique,” Journal of Southern African Studies, 32(3): 563-81.
  • Mann, Michael. 2008. “Infrastructural Power Revisited,” Studies in Comparative International Development, 43: 355-365.
  • ―. 1986. The Sources of Social Power: Volume 1, A History of Power from the Beginning to AD 1760. Cambridge, England: Cambridge University Press.
  • ―. 1984. “The Autonomous Power of the State: Its Origins, Mechanisms and Results,” European Journal of Sociology, 25(2): 185-213.
  • Manning, Carrie. 2008. The Making of Democrats: Elections and Party Development in Postwar Bosnia, El Salvador and Mozambique. New York, NY: Palgrave.
  • Olowu, Dele and James Wunsch, eds. 2004. Local Governance in Africa: The Challenges of Democratic Decentralization. Boulder: Lynne Rienner.
  • Prasad, Monica. 2007. “Review: Structuring the State,” American Journal of Sociology, 113(2): 560-62.
  • Weber, Max. [1922] 1958. “The Three Types of Legitimate Rule,” Berkeley Publications in Society and Institutions, 4(1): 1-11.
  • West, Harry and Scott Kloeck-Jenson. 1999. “Betwixt and between: ‘Traditional Authority’ and Democratic Decentralization in Post-War Mozambique,” African Affairs, 98(393): 455-84.
  • Young, Crawford. 2012. The Postcolonial State in Africa: Fifty Years of Independence, 1960-2010. Madison: University of Wisconsin Press.
  • Ziblatt, Daniel. 2006. Structuring the State: The Formation of Italy and Germany and the Puzzle of Federalism. Princeton: Princeton University Press.
  • ―. 2004. “Rethinking the Origins of Federalism: Puzzle, Theory and Evidence from Nineteenth-Century Europe,” World Politics, 57(1): 70-98.



About the Author


Mathew Maavak

As the global COVID-19 scourge appears to recede, questions remain over the source and morphology of a virus that had locked down two-thirds of humanity over the first half of…

Read more

Dieter Boden

This article is a book review of Philipp Ammon, Georgien zwischen Eigenstaatlichkeit und russischer Okkupation. Die Wurzeln des Konflikts vom 18. Jh. bis 1924, Frankfurt/M 2020. This is the fascinating…

Read more

Esmira Jafarova

When Armenia occupied about 20 percent of Azerbaijani territories in the consequence of the military assault against the latter and the resulting active military hostilities at the beginning of the…

Read more