Abstract
This article analyses state formation in Botswana and Rhodesia/Zimbabwe in the light of general theories of state formation. It seeks to explain why Botswana became a strong state, while state power in Rhodesia/Zimbabwe has eroded. Three main arguments are made. First, Zimbabwe’s experience shows that security threats and compulsion to collect revenue are not sufficient conditions for establishing a strong state. Second, the Botswana case indicates that the formation of a strong state may be possible even in the absence of such compulsions. Third, both cases reveal that patterns of state formation ultimately depend on the interests of ruling regimes, and on their ability to promote those interests. The creation of a strong state requires that regimes must have an interest in a strong state as well as the power to carry out a project of state building. This depends on the relationship between regimes and their constituencies. The article concludes by pointing out the need for expanding the framework described by security-based and fiscal theories, to include considerations about how state formation is related to the underlying social forces on which state power is based.
Introduction
In 1966, Botswana gained independence from Britain. In the previous year, the white minority regime in Rhodesia had declared itself independent from Britain. At that time, it would have appeared extremely unlikely that, 40 years later, the state in Botswana would be widely recognized as one of the strongest and most effective states in the developing world, while the state in Zimbabwe is on the brink of collapse.
This article focuses on the process of state formation in Botswana and Rhodesia/Zimbabwe. The main question posed is this: how can the divergent trajectories of the state in the two countries be explained? Why did a strong state emerge in Botswana, while state power in Rhodesia/Zimbabwe has eroded dramatically?
These two countries appear to contradict two of the most influential approaches to state formation. One claims that state formation is the result of warfare and geopolitical competition (Mann, 1993; Tilly, 1985). The second approach argues that state formation is linked to the extraction of revenue (Moore, 2004). Drawing upon the tradition of historical sociology from Barrington Moore to Rueschemeyer, Stephens and Stephens, I argue that the geopolitical and the fiscal perspectives cannot explain state formation in Zimbabwe and Botswana, and that the nature of the ruling regimes’ social basis has defined the constraints and opportunities faced by regimes. The strategies chosen by the regime and powerful groups to adapt to these constraints have determined the pattern of state formation.
The article has the following structure. First, I briefly discuss the key concepts of the state and state formation, and present a conceptual framework explaining the formation of strong and weak states. Second, I provide a short description of the historical development of the state in the two countries. Third, state formation in Botswana and Rhodesia/Zimbabwe is analysed and discussed in relation to the conceptual framework outlined in the next section.
State formation: concepts and theories
State formation may be divided into two components. 1 First, and most fundamentally, it implies that the state must establish itself as the highest political authority within its territory, with the authority to make binding decisions for society as a whole. This requires that the state maintains an effective territorial monopoly of violence. Moreover, the state must have ‘direct access’ to its population (Taylor, 1998: 39). Rival centres of authority are subordinated to the state, and exercise political influence through state institutions rather than by challenging the state. I refer to this dimension as domestic sovereignty. I empirically assess the state’s domestic sovereignty first in terms of its ability to maintain an effective monopoly of violence throughout its territory, and second in terms of its ability to subject social forces such as chieftaincies, tribes and ethnic associations to its rule.
Second, state formation requires that the state has the administrative capacity to effectively enforce its rules and decisions. This requires, first, that the state must be able to reach local communities to implement centrally conceived policies (Mann, 1993; Migdal, 1988). Second, there must be a cohesive state organization with a well-defined division of labour and lines of authority to develop and implement government policies in a consistent way. This organization should be insulated against influences from unauthorized interest groups over its staffing, policy decisions, and resources (Weber, 1968). I use two main indicators of administrative capacity, related to ‘output’ and ‘process’ respectively. In terms of ‘output’, capacity is assessed by the state’s ability to implement policies and provide services. In terms of process, I focus on the extent of patronage and clientelism. A state with both domestic sovereignty and high administrative capacity is a strong state, while a state that has neither is a weak one.
Writers such as Mann (1993) and Tilly (1985) have argued that state formation in Europe was intimately linked to warfare. There were two ways in which inter-state war (or threats of it) contributed to state building. First, geopolitical competition ensured that only those states that were able to defend themselves were able to survive. Others were simply conquered and incorporated into other states. Second, in order to cover the costs of war, governments were forced to increase revenue collection. This in turn forced them to improve their administrative capacity in order to be able to tax their population.
Others have argued that the need to raise revenue can compel governments to strengthen the state even in the absence of security threats. According to the fiscal theory of state formation, strong states emerge when the ruling regime is compelled to expand its revenue base (Moore, 2004). If states rely on domestic taxation, such as personal income tax, property tax and corporate tax, they are compelled to develop their capacity for tax collection. By contrast, states that have access to unearned revenue (‘rents’) are not compelled to do this. The two main sources of rents have been control over natural resources (resource rents) and sovereignty (strategic rents). Minerals and diamonds have been important resource rents for many states. Strategic rents are based on the regime’s ability to use its sovereignty to extract economic surpluses from international networks (Jackson, 1990; Moore, 2004). 2
Both these theories focus on how state formation is affected by the interests of rulers. Thus, according to both perspectives, the formation of a strong state is possible when the interests of rulers are served by state strengthening. These interests, in turn, are seen as structurally determined, either in the form of geopolitical threats or in the form of a compulsion to generate economic resources. If regime interests are best served by strengthening state institutions, it is likely that such a policy will be attempted. 3 Conversely, if there is a real or perceived contradiction between regime interests and state building, it will not.
However, if a project of state building is to succeed, it is not enough that the ruling regime sees it as being in its interest, or that it is committed by norms, values and ideologies to engage in state strengthening. It must also have the power and resources that make it possible to carry out such a project, which, in turn, depends on the relationship between regimes and their constituencies. If groups on whose support the regime depends do not consider a strong state as serving their interests, it may be impossible for the regime to garner sufficient support for state strengthening. This may be illuminated by Shefter’s (1994: 28–9) distinction between ‘constituencies for bureaucratic autonomy’ and ‘constituencies for patronage policy’. The former are coalitions with an interest in insulating the administration from outside interference, to ensure that positions and resources are allocated according to bureaucratic procedure. The latter are groups who rely on patronage. Constituencies for state formation may consist of both state officials (political leaders and civil servants) and members of groups outside the state. The form of state that emerges will to a large extent depend on the power balance between these groups, and which of them the regime least can afford to alienate (Shefter, 1994).
In this article, I argue that rulers’ interest in creating a strong state does not depend on security threats or compulsion to raise revenue. I investigate how state formation in Zimbabwe and Botswana has been shaped by the interests and power of ruling regimes’ constituencies and not by security threats and a compulsion to raise revenue. The availability of allies for state formation, in turn, depends on the nature of a country’s class structure. When key social groups are market dependent, they also have an interest in the establishment of an effective state, since the functioning of the market depends on it (Khan, 2005). Thus, while capitalists may want to limit the size of the state, they nevertheless depend on a well functioning and effective state to protect property rights, enforce the law, provide infrastructure etc. If key groups are not market dependent, state effectiveness matters less for their welfare, either because their material well-being is not affected by the state, or because they benefit from an ineffective state. In such conditions, it will be difficult for regimes to find allies who will support state building.
Capitalists (both agricultural and industrial) are by definition market dependent. They therefore have a long-term interest in state building, since a reasonably effective state is a precondition for overall economic growth in a market economy, from which capitalists will benefit (Khan, 2005). Likewise, the working class is market dependent, and could therefore in principle serve as a constituency for state strengthening. If the state can be relied upon to provide security, limit exploitation and perhaps provide some basic services, the working class may have an interest in state building. However, both capitalists’ and workers’ support for state building is likely to be contingent. 4 If they have (actual or potential) influence on state policies, or if they can count on the state to protect their basic interests, they may support a strengthening of the state. By contrast, if the state is seen as repressive, and as a threat to their property rights, security and well-being, they may not.
The position of rural classes (other than agricultural capitalists) towards state building is likely to vary, depending on rural class structure and on the nature of rural classes’ links with the state (Boone, 2003). If rural elites depend on the state in economic terms, it would be relatively easy for the ruling regime to forge an alliance with them. Thus, if there are strong rural classes with economic interests that are served by a strong state, and if the ruling regime is closely linked to these classes, the formation of a strong state could be possible.
State formation in Rhodesia/Zimbabwe
The state in Rhodesia was a state built for and run by the white settlers. After the transfer of power from the British South Africa Company to the colonial government in 1923, the settlers constructed an interventionist state (Herbst, 1990; Sylvester, 1991). Something resembling a corporatist system was established, in which interest groups worked in close collaboration with the government. 5 Military force was an important feature of this state from the outset (Cliffe, 1981; Good, 1974).
Because Britain was not willing to grant the country formal independence unless steps were taken towards majority rule, Rhodesia unilaterally declared its independence in 1965. This led to international sanctions during the next 15 years. As the guerrilla war intensified from the mid-1970s, power was centralized, and military spending increased. As a result, the form of rule became more authoritarian. Moreover, the economic base of the state started to erode and industry began to struggle (Skålnes, 1995). Nevertheless, even in this period, the state showed considerable capacity. Thus, it was able to relocate up to half a million people to ‘protective villages’ in order to cut them off from the guerrillas (Cliffe, 1981: 27).
The main limitation of state power during minority rule was that it relied on governing rural areas reserved for the African population (the Native Reserves) through traditional leaders (Alexander, 2006; Munro, 1998). From the early 1960s, the government sought to ‘disengage the state from rural development and to refract accountability through traditional authorities by focusing local conceptions of rights and civic responsibilities on those leaders’ (Munro, 1998: 144). However, this policy was not successful, and the rural population increasingly resisted and evaded state regulations. As a result, the state’s domestic sovereignty was incomplete, since it could only govern the Reserves indirectly.
In the first elections after independence in 1980, ZANU-PF, the liberation movement led by Robert Mugabe, won a clear majority. The new government did not seek to demolish the existing state, and by and large retained the state apparatus it inherited. Thus, white civil servants were encouraged to stay, and blacks entered the civil service through an overall expansion. By 1983, 86 per cent of the established officers were black (Murapa, 1984). In addition to teachers and health workers, state administration expanded in areas such as agricultural extension services. In the first decade of independence, therefore, both the size and the reach of the state was expanded.
Instead of risking undermining agricultural production through radical land reform, the regime sought to satisfy rural expectations by providing social services and subsidized inputs for agriculture (fertilizers, pesticides, water). This led to increased agricultural production (especially in the former Native Reserves), improved services and consolidated rural support for ZANU-PF (Dashwood, 2000).
The new government also took over the authoritarian elements of the former regime. Spending on defence, para-military forces and police remained high, and in 1983, the country’s defence budget stood at 6.5 per cent of GDP, compared with a 3.5 per cent average for developing countries (Weitzer, 1984). This was seen as necessary with contain the threat from South Africa and quell internal dissent.
At the same time, the new government sought to curtail the power of the chiefs, partly because it saw them as collaborators with the minority regime, and partly because they constituted a potential rival authority in the countryside (Alexander, 2003; Munro, 1998; Raftopoulos, 2003). ‘Development Committees’ were introduced at village, ward and district levels. However, the government was never able to bypass traditional leaders. Thus, the attempts to establish domestic sovereignty did not succeed. Instead, the state retreated from the task of establishing itself as the highest political authority in the countryside.
There was also a growing tendency of ‘patrimonialization’, as access to economic resources became dependent on political connections. This was especially clear in relation to the policies of ‘indigenization’, which supported the black business class through preferential treatment in the award of government contracts, licences etc. (Dashwood, 2000; Davies, 2004; Moyo and Yeros, 2007). As argued by Scott Taylor, ‘indigenization as currently practised in Zimbabwe is a neo-patrimonial game that rewards inefficiency and depresses productivity’ (1999: 260). Such practices mostly benefited those with links to the ruling party.
With the approval of the programme of structural adjustment in 1990, there was a clear move towards a more ‘liberal’ economic policy. However, economic performance in the post-liberalization period has been far worse than in the 1980s, in terms of growth, employment and wages (Brett, 2006).
Structural adjustment also contributed to a weakening of state capacity, as a result of the decline in economic resources and cuts in public administration (Alexander, 2003; Moyo and Yeros, 2007). The impressive expansion of government services (especially in rural areas) proved impossible to sustain, as expenditures were cut and growth rates declined.
By the mid-1990s therefore, state capacity in terms of both ‘output’ and ‘process’ had declined. Service delivery had eroded and there had been a growth of patronage. At the same time, the state did not have direct access to rural communities, since traditional leaders had not been subordinated to the state. Thus, the state had become significantly weakened.
In 1997, a series of government decisions sparked a crisis the country has yet to resolve. Large payments to war veterans, the engagement in the war in Congo and the decision to appropriate a large number of white-owned commercial farms, seriously undermined state finances (Bond and Manyanya, 2002). Initially, the government tried to cover the cost of these decisions by raising taxes, but this only contributed to further popular resentment.
According to Alexander, subsequent developments have reinforced the decline: ‘By the year 2000, the state was much weakened, and was becoming increasingly politicised . . . The [land] occupations produced a strange spectacle: a government effectively unravelling its own state with great vigour’ (2003: 113–14).
State formation in Botswana
When Botswana achieved independence in 1966 there was little reason to believe that it would become the most successful state on the African continent. The country had ‘no army, no strong bureaucracy and a weakling middle class’ (Molutsi, 1989: 104). It had only 22 km of paved road, 22 university graduates and a mere 100 graduates from secondary school (Acemoglu et al., 2003). Financially, more than 50 per cent of its revenue was aid from the UK (Samatar, 1999).
The central government also had limited ability to govern. The bureaucracy left behind by the British was clearly incapable of promoting economic development. In 1964, two years before independence, the entire civil service consisted of 607 officers (Picard, 1987). Of these, 502 were expatriates. This meant that ‘the Tswana educated elite was so small that it ended up cooperating with the colonial state, the chiefs and European settlers to form the new ruling class at independence’ (Molutsi, 1989: 104). Moreover, while there were no armed insurgents, powerful chiefs managed the country’s eight tribal reserves (Samatar, 1999). Thus, the state was in a disadvantaged position in terms of both administrative capacity and political authority.
The new leadership was determined to promote economic growth. 6 Its strategy was to pursue market-based development, in which the state had a key role: ‘In the absence of a vibrant local private market economy, political leaders and bureaucrats shared a view of development where the state played a key role in jump-starting the economy’ (Samatar, 2002: 19–20). This required strengthening the state’s administrative capacity. First, the government gave high priority to strengthening the capacity to coordinate and manage economic development. It strengthened the Economic Planning Unit and set up a Central Statistics Office. Key posts in these units were held by expatriates (Picard, 1987). Second, it was decided that the pressure for rapid Africanization (or ‘localization’) of the civil service was to be resisted. Seretse Khama, the first President, was ‘deeply conscious of the dangers inherent in localising the public service too quickly’ (quoted from Parsons et al., 1995), and made it clear that he would ‘never sacrifice efficiency on the altar of localisation’ (Picard, 1987: 205). While localization was considered desirable, the government recognized that proceeding too quickly would undermine efficiency as long as there was a shortage of qualified Africans.
The administration was given a high degree of independence (Leith, 2005; Samatar, 1999). 7 At the same time, politicians defined the policy framework within which the administration had to work, by (a) ‘resisting the pressure to quickly localize technical and professional levels of the public service; (b) insulating the public service from political intervention; (c) clearly demarcating the political and economic boundaries within which policy making must operate’ (Samatar, 2002: 34–5).
In the first decade after independence, public sector employment grew rapidly, reaching 21,675 in 1976. During the same period, the central government civil service grew from 607 to 2611 in 1977 (Parson, 1984). In addition to the strong Ministry of Finance and Development Planning, the Botswana Meat Commission (BMC) and the Botswana Development Corporation (BDC) were particularly important. The BMC became the sole agency responsible for the slaughter, marketing and export of meat and meat products. It has ensured high standards of production and relatively high producer prices, and surpluses have been returned to the producers (Samatar, 1999).
The BDC was established in 1970 to provide capital for commercial agents. It is fully state owned, and its purpose is to serve as the principal national organization to promote industrial development (Leith, 2005; Sebudubudu, 2005). To ensure that the BDC would not become a ‘constant financial drain on the public sector’ (Samatar, 1999: 155), it was made clear that it would operate on a commercial basis. Although many of its investments in industry and agriculture have been unprofitable, its success in other areas, especially finance and real estate, has more than made up for these losses.
The strengthening of administrative capacity is seen in terms of both ‘output’ and ‘process’. In terms of output, infrastructure and service provision were radically expanded. The state increased the country’s capacity in the cattle industry, by building abbatoirs and providing veterinary services and vaccines and by funding the construction of fences. Moreover, it greatly expanded the provision of services in areas such as health and water supply. Thus, by 1998, 88 per cent of the population lived within 8 km of a government health facility, and 97 per cent had access to safe drinking water (UNDP, 2005).
In terms of process, the state in Botswana is characterized by a high degree of coordination of policies and, most notably, by the relative absence of clientelism. That Botswana has been able to contain clientelism is well documented. First, in comparative rankings of ‘good governance’ (such as those compiled by the World Bank, UNDP and Transparency International) Botswana is consistently given high rankings (Robinson and Parsons, 2006; World Bank, 1984). Second, using three empirical indicators of patrimonialism (concentration of power with the leader, provision of personal favours and misuse of state resources), von Soest (2009) has shown that although elements of patrimonialism can be found in Botswana, it has been restricted and less widespread than in other African countries. Further, the number of ministries and turnover of ministers and senior administrative staff have been limited. This indicates that appointment to high level positions has not been used to buy support or co-opt potential opponents.
Further, by bringing traditional leaders, a potential source of opposition, under state control, the state was able to establish domestic sovereignty (DuToit, 1996; Taylor, 2002). 8 The Chieftainship Act of 1965 gave the President the power to recognize traditional rulers. Thus, rather than constituting an alternative power centre, chiefs have ‘served as facilitators for the implementation of government policy, particularly in rural areas. Whilst accorded respect and status, their role within Botswana was reinvented, and chiefs became agents of the government at the grass-root level’ (Taylor, 2002: 9). Moreover, new institutions were established (land boards, councils, development committees) that replaced traditional leaders and transferred power to the state. This meant that the state’s position as the highest political authority was consolidated.
Thus, a weakening of the power of traditional leaders and the establishment of domestic sovereignty went hand in hand with an expansion of the state apparatus and a strengthening of administrative capacity.
The politics of state formation: security, revenue and political constituencies
This section seeks to explain the pattern of state formation in the two countries. Returning to the two frameworks for analysing state formation described earlier (the security based and the fiscal), I first argue that neither of them fit with the experience of Zimbabwe and Botswana. Next, I outline an alternative explanation, focusing on the nature of regime power and interests in the two countries, and on the character of ruling regimes’ allies and political constituencies.
Rhodesia/Zimbabwe
In Rhodesia, the ruling regime confronted a growing mobilization from the national liberation movement, which directly threatened its hold on power. Moreover, after the unilateral declaration of independence in 1965, the regime was faced with sanctions, which forced it to carry out a policy of self-reliance. Because of these threats, a strengthening of the state was seen as an imperative. And as expected from security-based theories, this led to a strengthening of state institutions. The external environment strengthened the government’s support among the white population, and legitimized appropriation of economic resources. Further, average annual growth rates of 8.9 per cent between 1967 and 1974 created the economic foundations for doing so (Skålnes, 1995). Paradoxically, therefore, enforced import substitution resulting from sanctions had the unintended effect of strengthening the economy, making it possible to create a stronger state administration.
After independence, the state was faced by new security threats. First, the apartheid government in South Africa was hostile to the new government (Hanlon, 1988). It tried to undermine it by imposing sanctions (Sylvester, 1991). It also carried out several bomb attacks. Second, the government feared that a rebellion could break out in Matabeleland, and that rebels there could align with South Africa against the government (Weitzer, 1984). Although the threat was probably exaggerated, it served to justify high military expenditure and repressive policies. From the late 1980s however, both these threats subsided. As it started moving towards majority rule, South Africa gave up its strategy of subverting Zimbabwe, and the situation in Matabeleland stabilized after 1987, when ZANU-PF was merged with ZAPU, the regional party from Matabeleland.
According to the fiscal theory of state formation, strong states emerge when regimes depend on domestic taxation, such as personal income tax, property tax and taxes on corporate profits. In such situations, states are compelled to develop their capacity for tax collection and their administrative capacity. The state in Rhodesia/Zimbabwe always relied primarily on earned income. The country does not have large mineral resources and although aid increased significantly after 1980, it never constituted more than 5 per cent of the government’s total revenue (Jenkins, 1997; World Bank, 1995).
Thus, in Rhodesia/Zimbabwe, both the security situation and the structure of state revenue pushed the regime towards strengthening the state. The country was faced by security threats both before and after independence. It also relied mainly on earned income. Nevertheless, the state was never able to establish itself firmly as the highest political authority in the countryside. Moreover, the administrative capacity that had been built up before the 1990s has eroded in recent years.
To explain this outcome, it is necessary to analyse the nature of the state’s and the ruling regime’s ties to different social groups. Analysts of Rhodesian politics agree that the state was dominated by a class of commercial farmers, with the support of the white working class (Arrighi, 1966; Cliffe, 1981). However, the regime was able to gather support in other sections of the white population. The white working class depended on the state for the preservation of its privileged position as a labour aristocracy. Further, a collective fear resulting from sanctions and guerrilla mobilization contributed to creating a broad white coalition in support of the ruling Rhodesian Front (RF).
The RF saw the strengthening of the state as an imperative for its own political survival. And in some ways, it succeeded. The state apparatus – both military and civilian – was expanded and strengthened. The absence of foreign aid and large mineral resources made such an expansion of the state apparatus dependent on the ability to raise domestic revenue. Thus, the social composition of the Rhodesian regime meant that the class interests of its main constituencies were served by an effective state administration.
At the same time, the racial dualism of the Rhodesian state made it impossible for the regime to establish full domestic sovereignty in the Native Reserves. The regime had no legitimacy in these areas, and residents were only partly integrated into the market. First, ownership of land in the reserves was communal rather than private. Second, only a small part of the crops produced here were sold on the market. Third, residents were only partly employed in wage labour (as casual workers in agriculture or in urban areas). Consequently, residents in the reserves did not ‘objectively’ depend on a well functioning state, and most likely considered the state as a force to be avoided, rather than a solution to their problems.
While the RF regime tried to build support in rural areas by aligning itself with traditional leaders, it did not succeed. Instead, the efforts at shifting the responsibility of development on to local communities, while maintaining state control, proved contradictory, and led to a fragmentation of authority in the countryside (Munro, 1998). The problem was that the efforts at enhancing the formal, institutional authority of traditional leaders ‘undermined the moral authority of the chiefs and reduced the state’s capacity to regulate authority in the countryside’ (Munro, 1998: 205). The government’s policy of ‘indirect rule’ meant that it retreated from the task of establishing the state as the highest political authority in the countryside. Thus, ‘by the early 1970s, the Rhodesian state had become dependent on chiefs who could only sustain their moral authority if they defied the state’ (Munro, 1998: 208).
After independence, the social basis of state power changed. The new regime was initially a broad cross-class alliance, held together by opposition to minority rule and mobilized on a national-democratic basis. White capitalist farmers (and other whites), who had close ties to the old regime, stood in an antagonistic relationship to the new regime. At the same time, the regime continued to rely on taxation of whites. Majority rule, together with the moderate approach to economic reform and redistribution, meant that a separation between economic and political power emerged. ZANU-PF controlled the state, but whites continued to be economically dominant. This meant that regime power was no longer based on an alliance with a strong capitalist class.
While the regime was dominated by the former liberation movement, urban professionals, the labour movement and peasants initially supported it. It could also count on the support of those sections of the black capitalist class that benefited from policies of ‘indigenization’. 9 A state-dependent bourgeoisie of this kind, which relies on the state’s distribution of ‘rents’, is not likely to act as a pressure group for the maintenance of an effective state.
After the adoption of the programme of structural adjustment in 1990, urban support for the regime eroded (Bond and Manyanya, 2002; Dashwood, 2000). The labour movement, which had been loyal supporters of ZANU-PF, gradually become more critical of the regime. Similarly, the (predominantly white) business community, while supportive of liberalization, had always been suspicious of the regime. Thus, the ruling regime lost credibility in urban areas. As urban hostility to the regime grew, it was left with the support of the peasantry and the party elite and groups linked to it – none of whom saw their interests best served by a strong state.
In terms of votes, the main constituency of ZANU-PF has been residents of communal areas. Thus, the impressive expansion of government services after independence (social services and subsidized inputs for agriculture such as fertilizers, pesticides and water) was mainly concentrated in these areas (Herbst, 1990; Skålnes, 1995). However, by the late 1990s, the standard of government services had declined, as social services and agricultural subsidies were reduced. This undermined the policy of building legitimacy through service provision. As a result, rural support of the regime also weakened (Alexander, 2003; Raftopoulos, 2003).
Like many African states, Zimbabwe has had an ambiguous attitude towards traditional leaders, seeing them variously as a threat to the authority of the state and as allies to be used to establish state hegemony at the local level. After independence, the government sought to replace the authority of traditional leaders to administer local communities directly. However, the government was not able to bypass traditional leaders, and had to depend on their cooperation for key government tasks such as tax collection and land allocation (Munro, 1998).
In the late 1990s, the state sought to recreate an alliance with traditional leaders in order to draw upon their legitimacy (Alexander, 2003, 2006; Munro, 1998). With the approval of the Traditional Leaders Act in 1998, chiefs and headmen were ‘granted leadership of development institutions below the District Council, control over land allocation in communal areas and a greater role in that one-time bastion of state directed modernisation, the resettlement areas’ (Alexander, 2006: 183). They were also given responsibility for local tax collection. Thus, a version of indirect rule was once again introduced, with traditional leaders acting as intermediaries between the state and the local community.
Taken together, this meant that the position of the ruling regime had become increasingly fragile. Both its political basis in urban areas and its core rural support had been weakened. In an attempt to revive its legitimacy, the government decided to carry out large-scale land reform by taking over the majority of the white-owned farms. Redistribution of land had been a key demand in the independence struggle, and while progress in this area had been limited in the 1990s, the demand for land was never in doubt.
Both the radicalized land policy and the accommodation of traditional leaders must be seen as attempts to shore up rural support in a situation where regime legitimacy had eroded. In this situation, the struggle for political survival compelled the regime to choose policies that could – at least in the short to medium term – bolster its legitimacy. And in spite of their chaotic implementation, land reforms probably increased regime support in rural areas. However, the government had no funds to pay for land. Moreover, land reform led to a dramatic fall in agricultural production and export earnings. Faced by a situation where its own position was under threat, the regime chose policies which could strengthen its own legitimacy, even when these policies also undermined economic growth and the basis of state power itself. As a result, the regime found itself in a downward spiral, in which both the welfare of the population and the power of the state were undermined.
Botswana
In Botswana, the regime did not consider itself to be under threat from its neighbours. This is revealed by the fact that the country did not even have an army until 1977, when the Botswana Defence Force was established as the result of tension between Botswana and Rhodesia at the height of the Rhodesian civil war (Molomo, 2005). However, this threat cannot explain the emergence of a strong state, since, by then, the formation of a strong state had by and large already been accomplished. The regime also maintained a working relationship with the regimes in South Africa and Rhodesia. They did this through economic cooperation and by maintaining a balanced foreign policy, opposing minority rule in neighbouring countries while at the same time forbidding rebel groups to operate in Botswana (Dale, 1995; Niemann, 1993). Further, there were no internal security threats, and the regime faced no challenges from domestic opponents.
The Botswana case cannot be explained by the fiscal theory of state formation either. Botswana has been dependent on both strategic rents and national resource rents. In the first years after independence, it relied heavily on aid from the UK. Then, after the successful renegotiation of the Southern African Customs Union in 1969, its income from customs increased significantly. Later, as the diamond sector expanded, mineral exports became its main source of income. The proportion of state revenue derived from ‘earned income’ remained between 10 and 15 per cent (Government of Botswana, various years). Thus, the state has depended mainly on unearned rents. It has gone from being dependent on strategic rent (aid) to dependence on customs to dependence on rents from export of natural resources (diamonds). While revenue from both income tax and corporate taxation has increased, they have remained minor in comparison with various forms of unearned rents.
Thus, an alternative explanation is needed for the emergence of a strong state in Botswana. I propose that this outcome must be seen as the result of the character of relations between the ruling regime and its constituencies. The ruling regime in Botswana has been described as an elite alliance, composed of political leaders, civil servants, businessmen, traditional leaders and cattle farmers (Molutsi, 1989; Samatar, 1999; Taylor, 2002). Further, most of these categories were overlapping. Cattle farmers and civil servants were the most important sections, who dominated policy formulation and decision making.
A key factor that made the formation of a strong state feasible in Botswana was the fact that one of the key groups in the country’s ruling coalition, the cattle farmers, was market dependent. 10 The role of cattle farmers was crucial for the formation of a strong state (Acemoglu et al., 2003; Parson, 1984; Samatar, 1999; Taylor, 2005). Their dominance is revealed by the composition of the national parliament. Of the total 42 MPs in the first three parliaments, no less than 27 were large or medium cattle owners (Samatar, 1999). This meant that positions in the state were not their main source of income; their economic interests were linked to the cattle sector rather than to their positions in the state (Acemoglu et al., 2003; Tsie, 1998). And since an effective state was seen as a condition for economic development, the creation of such a state became a priority.
In order to be able to create the kind of state desired by the dominant class, the ruling regime needed political support from other groups, beyond the members of the regime. This meant that the regime needed allies who supported a strengthening of the state, or at least did not undermine it. One such group was the civil service. In addition, support had to be found in rural areas, among peasants and local elites.
Most prominent among rural elites were the traditional leaders. In fact, while also a large-scale cattle farmer, President Khama was the country’s most prominent traditional leader. But in spite of his chiefly origin, Khama decisively undercut the power of chiefs. The Chieftainship Act made chiefs subordinate to the state. However, chiefs did not react to this erosion of their power by becoming hostile to the regime. Rather, most of them remained supportive of the ruling party, the Botswana Democratic Party (BDP). The fact that Khama was himself a chief probably made his policies more acceptable to them. Another reason for the success of this strategy was that many chiefs were cattle farmers, and had economic interests which were served by the creation of a strong state. Crucially, having made traditional leaders subordinate to the state, the regime was not compelled to use patronage to buy their support.
Peasants, the largest class in the country, have also by and large supported the BDP. Several reasons have been suggested for this support. First, at the material level, peasants were tied to large cattle owners through the mafisa system, in which cattle were farmed out to poorer kinsmen. This created a material dependence of peasants on cattle owners, but also a perception of the relationship as one of mutual benefit (Parson, 1984). Moreover, since boreholes were privately owned by cattle farmers, peasants’ access to water for their cattle depended on their relationship to borehole owners. Finally, cattle owners occupied ‘strategic positions in the Land Boards, which have been responsible for the allocation of tribal land to tribesmen since 1970’ (Tsie, 1996: 605). This material dependence was reinforced at the ideological level by the fact that Khama and other BDP leaders could draw on their status as traditional leaders to enhance their legitimacy: ‘The prevalence of party members of high ethnic and traditional status and the association of the BDP with the maintenance of traditional authority provided a legitimacy that is not easily undermined’ (Parson, 1984: 51).
The nature of rural social relations both enabled and constrained the regime. On the one hand, the overlap between traditional leaders and cattle farmers and the paternalist ties between cattle farmers and peasants strengthened the regime’s political position. Both chiefs and peasants became supporters of a regime dominated by cattle farmers and civil servants. This meant that there were no social groups that could form the base of a strong opposition. On the other hand, the regime was constrained by aspects of the rural social order. Notably, village assemblies (kgotlas) were preserved, and incorporated into the state (Acemoglu et al., 2003). This fusion of the old and the new made the establishment of state control possible. At the same time, it created a set of constraints, which ensured that the regime and its local representatives were held accountable. Paradoxically, therefore, the existence of non-capitalist social ties enabled a modernizing elite to pursue a policy of state building and capitalist development.
Peasant support for the BDP has been reinforced by government policies. Key policies included service provision and infrastructural development that benefited the rural poor. Substantial funding was committed to rural development, providing schools, clinics, water and roads. These programmes ‘not only cushioned the poorer section of the peasantry from proletarianisation and the effects of drought but also went a long way to demonstrating the concern of the government for people’s welfare’ (Tsie, 1996: 605). This demonstrated that the ruling regime made sure that all sections of the population would benefit (at least to some extent) from state policies. Thus, peasants did not see economic growth coming at their expense, and the regime was able to promote the class interests of members of the ruling regime while at the same time presenting its policies as serving the interests of society as a whole. Moreover, since service provision was distributed to all sections of the population on a non-ethnic and non-regional basis, opposition along ethnic or regional lines was avoided (DuToit, 1996).
Thus, while the new elite clearly pursued its own class interests, they simultaneously pursued ‘a general developmental project that transcended narrow class interests’ (Taylor, 2002: 12). This meant that the economic interests of the dominant coalition of cattle owners, civil servants and domestic capitalists in the creation of a strong state were compatible with their political interests (Acemoglu et al., 2003). A policy of state building did not threaten the regime. It had the support of cattle farmers (the dominant group), chiefs (the main potential rivals) and peasants, in addition to civil servants, and was not faced by any strong challenges from oppositional groups.
The inclusiveness of the regime, together with the weakness of the opposition, meant that there were no rival political groupings who could threaten the regime’s hold on power. The co-existence of a class of market dependent cattle farmers and paternalist ties between this class and the peasantry meant that the ruling coalition had both an interest in the creation of a strong state and the ability to build it. The regime was not compelled to reproduce its power through patronage, and was able to pursue its policies without fearing that those policies could undermine its hold on power. 11
The conditions of state formation
To sum up, both Zimbabwe and Botswana appear to contradict the security-based and the fiscal theory of state formation. While both theories would lead us to expect a strong state to emerge in Zimbabwe, but not in Botswana, the actual outcome was the opposite. A strong state in terms of both sovereignty and capacity emerged in Botswana, while the state in Zimbabwe has gone from a situation with high capacity and limited sovereignty to one where both capacity and sovereignty have been severely weakened.
In Zimbabwe, domestic sovereignty was never established, since the state was unable to bypass traditional leaders. And while administrative capacity was relatively high in terms of both output and process until the 1990s, subsequent developments have led to increased patrimonialization and erosion of services. By contrast, the state in Botswana was able to make traditional leaders subordinate to the state while at the same time containing patrimonialism and improving administrative capacity.
I have sought to explain these outcomes on the basis of the relationship between ruling regimes and their constituencies. In Zimbabwe, the erosion of state power in recent years must be explained by the changing nature of the ruling regime’s constituencies. After independence, it was impossible to create an alliance for state building with the white capitalists, who had controlled the state during minority rule. As for the emerging black bourgeoisie, they were dependent on the state for access to economic resources and privileges. Their interests were therefore well served by their incorporation into state-based patronage networks. Further, the regime gradually lost support among workers, the urban poor and capitalists not integrated into the regime’s patron–client networks. The single most decisive event in this regard was the adoption of the structural adjustment programme in 1990, which eroded the support base of the ruling regime (working class, urban middle class), and over time undermined the regime’s support. This made the regime dependent on the support of a state bourgeoisie and residents of communal areas. And when rural support began to erode in the 1990s, the regime responded by carrying out large-scale land reform and by recreating an alliance with traditional leaders. Thus, the regime became dependent on constituencies without a strong interest in an effective state. The absence of a strong political alliance that pushed for the formation of a strong state is the most important reason for the crisis of the state in Zimbabwe.
In Botswana, a strong state, with a secure monopoly of violence and an effective administration was established, in spite of the adverse conditions faced by the country at independence. In political terms, the state has been able to incorporate potential rival power centres (the chiefs), construct an alliance that includes the majority of the population and maintain it without relying on the distribution of spoils. In administrative terms, the bureaucracy is effective and competent. This has been accomplished in the absence of security threats or a compulsion to collect revenue. The key factors that made this possible were the compatibility of the ruling coalition’s economic and political interests and the broad social basis of the regime. Before the emergence of diamond revenue, both the economic and the political interests of the regime were served by an effective state. Later, when the regime had access to diamond revenue, they were less dependent on the cattle sector. However, by that time, a reasonably effective state had already been created. Moreover, the regime’s interest in political survival was still served by such a state, since it was not compelled to use patronage to maintain its power. There was no strong domestic political opposition that threatened the regime’s control over the state. And because the economic interests of the regime were served by a strong state, while their political interests were not threatened by it, the regime was able to strengthen the state.
Conclusion
Three conclusions may be drawn on the basis of this analysis. First, Zimbabwe’s experience shows that security threats and compulsion to collect revenue may not be sufficient conditions for establishing a strong and sovereign state. While Zimbabwe was faced with security threats and depended on earned revenue, the state has been fundamentally weakened. Second, the Botswana case indicates that the formation of a strong state may be possible even in the absence of such compulsions. Third, the experience of both countries suggests that patterns of state formation ultimately depend on the interests of ruling regimes and their ability to promote those interests. A one-sided focus on incentives and interests is voluntarist, since it presupposes that regimes can build a strong state if such a state serves their interests. This ignores the power constraints faced by many regimes. If state building is to succeed, it is not enough that the ruling regime sees it as being in its interest. It must also have the power that makes it possible to carry out such a project, which, in turn, depends on the relationship between regimes and their constituencies. Otherwise, it may be impossible to garner sufficient support for state strengthening.
At a theoretical level, this points to the need for expanding the ‘Weberian’ framework of the security-based and the fiscal theories, to include considerations about how state formation is related to the underlying social forces and classes on which state power is based. Such an approach incorporates the insights of those who emphasize the importance of security threats and sources of revenue, while integrating these concerns with an analysis of how state formation is shaped by the political foundations of ruling regimes and the distribution of power and resources in society.
The constituencies of a regime are fundamentally conditioned by the nature of national class structures. Thus, this analysis supports the argument that the analysis of class-conflicts and class-coalitions is ‘a master key to understanding the social structuring of interests and power in society . . . because the organization of class-interest is constitutive of major collective actors’ (Rueschemeyer et al., 1992: 5). Most importantly, class structure affects state formation by defining the opportunities and constraints within which the formation of political alliances takes place. At the same time, the impact of class structure on political development is always mediated and modified by other factors, including class organization, geopolitics, security and the state’s sources of revenue.
Placing the relationship between regimes and constituencies at the centre of analysis does not imply a rejection of the security- and revenue-based theories. Instead, it makes it possible to treat these theories as specific instances of a more general logic of state formation. The existence of security threats and the compulsion to raise earned revenue are two mechanisms through which regime interest in state building and political support for it may be generated. However, security threats and dependence on earned income are not necessary conditions of the formation of strong states. Instead, they affect the process of state formation by shaping the interests of ruling regimes and their allies. This may make it possible to create a strong alliance for state building, but a politically strong constituency for state building may also emerge in other ways. There may be situations, as in Botswana, where the distribution of power and resources in a country is such that powerful constituencies for state building emerge even in the absence of security threats or a compulsion to raise revenue.