DEMOCRACY IN THE SADC REGION
A comparative overview

Willie Breytenbach1

Willie Breytenbach is professor of Political Science at the University of Stellenbosch.

Published in African Security Review Vol 11 No 4, 2002


The purpose of this overview is to compare the state of democracy among SADC states. This is a region where electoral democracies increased significantly since 1989, but regular elections are necessary, not sufficient, to consolidate democracies. Socio-economic conditions as well as levels of freedom may be the missing links in this regard. Per capita incomes and human development indexes are combined to constitute an appropriate socio-economic development axis (quantified, then ranked) which is then juxtaposed with institutional criteria (autocracy, electoral democracy and consolidation zone), and then ranked according to surveys on political rights and civil liberties, constituting the liberty index (these measurements are ranked relatively). Factors such as electoral systems and deeply divided versus homogeneous societies are also taken account of. This makes for a model of development and democracy in the region with Mauritius at the most consolidated end, and the DRC the furthest away. This methodology can be repeated regularly, indicating comparative trends within the region, over periods of time.

Introduction

This paper focuses on democracy in the Southern African Development Community (SADC) region, which comprises 14 states including the island states of Mauritius and the Seychelles. The paper looks at ways of measuring democracy trends using neither a minimalist nor a maximalist approach; but rather an eclectic combination of quantitative and qualitative methodologies. It juxtaposes institutional and developmental indicators as two axes of the often dichotomous debate about the definition of democracy.

The paper therefore does not deal with the issues of regional co-operation or integration, except to speculate whether the deepening of co-operation and integration could have a beneficial impact on the development of democracy in the region.

For the definition of democracy, three assumptions were made. First, that without appropriate state institutions (and freedoms) democracy is not possible (“no state, no democracy”);2 second, without favourable socio-economic conditions, democratic institutions are unlikely to endure and consolidate (“once a country has a democratic regime, its level of economic development has a very strong effect on the probability that democracy will survive”);3 and third, that there are degrees of democracy (“Therefore it might be sensible to establish a category of semi-democracies to separate democracies from non-democracies’).4

This raises the issues of which came first: development or democracy; what kind of institutions are democracies made of; how should development be measured; what about the issue of deeply divided societies and democracy; and finally, does SADC have the kind of co-operation and integration rules that would be conducive to the strengthening of democracy in the region?

As explained above, it is the intention to design a methodology that could integrate these questions into a single matrix that would reflect comparable trends within the region.

Methodology and concepts

The minimalists use a minimum of independent variables in explaining democracies and their endurance. This has always to do with institutions,5 and sometimes only with regular elections. To be sure, for institutionalists, the starting point is the state. One is reminded of Linz and Stepan’s minimal conditions: “no state, no democracy”.6 For Robert Dahl, democratic states ought to have the following institutions: elected representation, free and fair elections, political parties, inclusive suffrage, the right to run for office, freedom of expression (i.e. a free press), associational autonomy, the rule of law, an efficient bureaucracy and development based on a market economy.7 More specifically, the state must be “subject to law”, according to Linz and Stepan.8

The institutional bottom line is probably to be found in Sartori’s definition:9
In a democracy no one can choose himself, no one can invest himself with power to rule, and therefore no one can abrogate himself unconditional and unlimited power.
This raises the issue of elections. Michael Bratton10 postulates that while you can have elections without democracy, you cannot have democracy without elections. Bratton says elections are necessary but not sufficient. What is sufficient, is not the quantity but the quality and meaning of elections. And this is essentially about free and fair elections. Freedoms are therefore part of the equation. On this point, an important insight is offered by Schedler11 who points out that regular elections may indeed be insufficient to establish liberal or advanced democracies. For him, political and civil freedoms must also be in place, must widen and must deepen. The point is as Crawford Young12 says: periodic elections are no guarantee against authoritarian manipulation.

Then there are those who believe that elections may, under certain conditions, indeed be sufficient to indicate that democracies have consolidated. Samuel P Huntington13 proposes the ‘two-turnover test’, that is, whenever the winners in a founding election are defeated in a subsequent election, and the new winners are also defeated later. In the 2000 study, Przeworski (et al)14 restates this position:
They established empirically that democracies consolidate where incumbent political parties actually lose elections. In this comparative overview, this conclusion cannot be verified, except to state that by those yardsticks mentioned above, only Mauritius would qualify as consolidated in the SADC region.
As the SADC region comprises a diverse collection of states, an attempt will be made to measure degrees of democracy. Based on institutional qualities, Schedler15 classifies these degrees. His classification is four-fold: authoritarian, electoral, liberal and advanced. The last three indicate degrees of democracy. For him, democracy and authoritarian systems are not an either/or dichotomy, as he distinguishes between at least three types of democracies as well as non-democracies. This is similar to the approaches of Bollen and Jackman16 and Vanhanen17 where distinctions are made between democracies, semi-democracies and non-democracies. Ours is also three-dimensional: autocracy, electoral and the consolidation zone. This allows one to register negative or positive trends, which is exactly what Schedler did. For Schedler, the crucial issue is whether human rights and liberties improve, or not. His electoral category can perhaps be subdivided further, providing for dictadura (elected dictatorship) as the most authoritarian form of an electoral democracy; and democradura (‘hard’ democracy) which is less authoritarian, but still lacking in full political and civil liberties.

Civil and political liberties as measured in the Annual Surveys of Freedom House are added to Dahl’s, Bratton’s and Schedler’s views. They are incorporated into our methodology. As Peter Meyns18 argued, this methodology can be adopted for pragmatic reasons: firstly, Freedom House covers all the states in the SADC region, and secondly its reports are published annually so that comparisons may be made that indicate advances or setbacks on the liberty scale. It rates countries as ‘free’ (1 to 2.5), ‘partly free’ (3 to 5) and ‘unfree’ (5.5 to 7).

The institutional index in this paper is therefore also a freedom index that takes account of political rights (the right to form political parties, voter choice, and free and fair elections) and civil liberties (religious and language rights, gender and family rights, and freedom of the press), according to the judgement of its Advisory Board panellists.

The institutional and liberty index used in this paper will therefore contain three categories as mentioned above: autocracy; electoral democracy (subdivided between dictadura and democradura); and then the consolidation zone, which can also include delegative democracies—i.e. where the executive branch of government is dominant and despite voting regularly the electorate does not really make an impact on policies, and ‘two-turnover test’ states that may be regarded not only as within this zone, but actually consolidated.

An autocracy is usually unfree; an electoral democracy (could be) partly free; and the consolidation zone (should be) free. Freedom House’s trends will be factored into the outcome, as the ratings for 1996 and 2000 are compared.

An important institutional feature in this paper relates to a point made by Przeworski et al quoting Linz,19 namely that the stakes are higher under presidential than under parliamentary executive institutions. In a presidential executive there is a single winner (the president) while in parliamentary systems the defeated candidate for the presidency (or usually the premiership) will become leader of the opposition. For example, in South Africa and Botswana, the heads of state are presidents, but their source of authority is parliamentary as in the English or Westminster system. They are leaders of the strongest parties in parliament as constituted by elections. In this way, South Africa and Botswana have presidents, but retained their English-inspired Westminster systems of parliamentary accountability for their presidents.

Another institutional feature is Arend Lijphart’s20 distinction between majoritarian and proportional or consensus democracies (1984 and 1999). For Lijphart, in deeply divided societies, the majoritarian model spells majority dictatorship and civil strife, rather than an inclusive democracy. G Bingham Powell21 is also convinced about the virtues of the consensus model in divided societies. However, Linz and Stepan22 point out that deep divisions are surmountable through constitutional devices such as decentralised federalism and consociationalism that could include any combination of grand coalition and segmental autonomy and proportional representation and minority veto. In the SADC region, however, there are no consociational democracies, although some states have proportional representation-types of electoral systems, such as South Africa, Namibia, Angola, Mozambique and even Mauritius where its ‘block’ system resembles proportional representation more than the majority-based first-past-the-post-system.

It is significant that Mauritius, arguably the region’s most consolidated democracy, falls in this category. Three other newly democratised states, Namibia (1989), South Africa and Mozambique (both since 1994), may have learnt from the lessons in other parts of Africa and opted, not for first-past-the-post-systems, but for list proportional representation.23 Since then, Lesotho may also have learnt this lesson, as major instability followed the election outcome in 1998 when the opposition rejected the outcome of a first-past-the-post election. In 2002, new elections were held, where electoral systems were combined: 80 seats according to the majority-based constituency system, and 40 seats allocated according to proportional principles. This system increased the participation ratio of smaller parties. Stability was restored. Under this system, opposition parties in deeply divided societies are assured of participation through contestation. In first-past-the-post systems such as in Botswana, Malawi, Tanzania, Zambia and Zimbabwe, and in Lesotho until 2002, strong (ethnic-based) parties can dominate electoral politics to the detriment of national unity, and legitimate opposition sidelined not by unfree elections, but by less representative electoral procedures.

On the relationship between development and democracy, there is a strong assumption that institutional survival depends very much on development, affluence and growth. These become powerful predictors of the likelihood of democratic success.24 One would therefore agree with Richard Joseph,25 who observes that the drive towards democratisation in the early 1990s in Africa seemed to override the concern for the assumed prerequisites of liberal democracy such as economic wealth, class and political history. The result of this omission according to him was the rapid emergence of democratic illiberalism soon afterwards or, put differently, the erosion of institutions.

But it would be wrong to be too dogmatic about this. René Lemarchand26 argues that although hunger and democracy do not mix, there are major difficulties with the social preconditions of democracy approach as this tends to confuse causes and effects. Our view on this point is the same as that of Plattner27 who observes that you cannot have liberalism or democracy or development without the other. It would therefore be difficult to have development without democracy, although there is evidence that authoritarian South Korea, Taiwan, Chile, Malaysia and Argentina opted (quite successfully) for economic development first,28 paving the way for their subsequent democratisation. Far from saying that the authoritarian model is superior for the development and democratisation of developing countries, Armijo et al29 point out that the Philippines under Ferdinand Marcos and Zaire (now the Democratic Republic of Congo—DRC) under Mobutu were authoritarian and made a mess of economic development. So, the lesson still remains: development under democratic conditions is more likely than under authoritarianism. Historically, however, democracies are unlikely to emerge without prior social (read classes) and political (read states) development.30

In 1959 Seymour Martin Lipset, following the thinking of Max Weber and Joseph Schumpeter31 linked democracy to economic development, showing that stable democracies scored on average higher than dictatorships in terms of wealth, industrialisation, urbanisation and education.32 This was a clear case of treating democracy as a dependent variable, therefore as an outcome of rather favourable socio-economic conditions.

In 1992, Larry Diamond33 indicated that the linkage between development and democracy was even stronger when human development indexes (HDIs) were used as an indicator of development. Likewise, in 1995, Adam Przeworski analysed data compiled between 1950 and 1990 and also found that:
… the secret of democratic durability seems to lie in economic development … under democracy with parliamentary institutions.34
Przeworski et al wrote in 199635 that “poverty is a trap”, and that “poor countries, those with per capita income under $1,000, do not develop”, hence poor democracies are unlikely to survive. Inequality must also be reduced. But more specifically, they found in their 2000 study that wealth does not particularly lead to democracy, though it sustains democracy once achieved.36

Why this is so has, according to others, something to do with the class debate—that is, which class (the middle class or the working class) is the more favourable for liberalisation and eventual democratisation?

Barrington Moore published his influential work on the role of classes in the making of the modern world and was dismissive about peasants as modernisers, but was convinced that the middle class is key not only to modernisation, but to democracy as well. And this class was synonymous with a failed peasant revolution in France37 as well as with the origins of capitalism in France, the UK and the US (1966, Part One). For him, democracy was also a dependent variable: it depended on the capitalist middle classes, hence his dictum, “No bourgeoisie, No democracy”.38

To be sure, Raynor39 analysed middle classes worldwide and found that the major component is the professionals (in the modern world, university-trained and propertied classes) with the business groups in second place (these groups can be assumed to be propertied as well). Then follows ‘white collar groups’, where again one may assume literacy and property levels lower than that of the middle classes, but higher than the proletariat and peasant classes which are often the have-nots. The super-rich upper classes as well as the underclasses and the peasants could therefore be seen as the enemies of democracy, according to this way of thinking.

This brings us back to Lipset and Przeworski who both argue that higher per capita incomes are favourable for democratic endurance and sustainability. In the 1996 study Przeworski et al40 found that the main reasons why democracies endure are affluence (in line with Schumpeter’s and Lipset’s thinking), but add that continued economic growth and decreasing income inequality (as implied above) are important determinants. They also mention rival variables such as a favourable international climate, as well as “good fortune”. In a continent such as Africa, where artificial boundaries are the rule rather than the exception, historical continuity and ethnic homogeneity could be regarded as “good fortune”.

These issues, together with leadership and electoral systems, will be excluded from this study although some thoughts about electoral systems will be discussed hereunder. Suffice to state that electoral systems in Namibia and South Africa are predominantly based on proportional representation. In the absence of the good fortune of homogeneity, proportional representation seems the more appropriate system for heterogeneity.

The exclusion of leadership (and political culture) are unfortunate, because these are important qualitative criteria. According to Linz and Stepan41 these are crucial for democratic consolidation, as the contrasts in leadership styles between Nelson Mandela and Robert Mugabe may suggest. But the measuring of these characteristics requires quantitative data (e.g. sampling) and supportive empirical evidence, which this study was not designed for. Bratton and Mattes42 did precisely such a study. They reported that while support for democracy is wide in the Southern African region, commitment may be shallow, perhaps because of dissatisfaction with the performance of their elected governments. Also, Decalo43 makes the point that there is no developed thesis on style and leadership, though there is adequate suggestive evidence of the significance of this variable.

After the role of the middle classes and affluence, the other significant variable in the endurance of democracy, is the reduction of income inequalities.

The measuring of socio-economic inequality through the Gini co-efficiency index—which measures societal inequalities within national populations—would have been an ideal tool. But Gini figures are not generally available for all African states. However, UN HDIs based on per capita purchasing power, life expectancy and adult literacy, are meaningful substitute measurements.44 Per capita figures are given. Although literacy figures are available for the region, there is no point in factoring them in separately as they are represented in HDI figures. Mentioning them separately would therefore be counting them twice. HDI figures are categorised into four groups: high, high medium, low medium and low.

In this way the institutional and freedom index will consist of three degrees of democracy as mentioned before: autocracy, electoral and consolidation zone; and the development index (really, the HDI and per capita rankings) will consist of four categories. This makes for three times four (or 12) juxtaposed boxes into which SADC states will be placed. Periodic analyses will highlight negative and positive trends.

Finally, in this study states will be compared relatively. But these comparisons will take place over a period of time so as to establish trends. The periods covered in this study relate to the 1990s with 1989 as the beginning of Huntington’s (1991) era of (re)democratisation (‘Third Wave’) in Africa.45 Depending on data, the cut-off date is 2000/2002.

Using the same methodology over a period of time could yield interesting patterns from which further prognoses may be derived (Figure 1).



One other issue that will be dealt with in the last sections of this paper is the issue of deeply divided societies and democratic endurance. It refers to racial or religious or ethnic divisions.46 These are often assumed to impact negatively on associational autonomy, party political support patterns (the choice of an electoral system is relevant here), or even on political stability in those cases where party political rivalry invokes identity struggles, intolerance or violence. The assumption could therefore be that the higher the heterogeneity the greater the chance for ethnic conflict, and the smaller the likelihood of stability which is generally seen as favourable for development.47 As mentioned above, homogeneity may be seen as ‘good fortune’. Freedom House48 states this correlation quite explicitly:
Indeed, democracy is, as a rule, significantly more successful in mono-ethnic societies (that is, societies in which there is a single dominant majority ethnic group representing more than two-thirds of the population) than in ethnically divided and multi-ethnic states.
Correlations for the SADC region will follow in the last section.

Institutional data for SADC member states


Comparing multiparty/single party/no-party systems during the period 1989 to 2002 makes for a positive picture of the SADC region (Table 1). On paper, parliaments have made a big comeback in Africa.49

In 1989 (when the redemocratisation trend emerged in Africa), the SADC region had only three multiparty states. By 2002 there were 12. In 2002, the free countries according to Freedom House were Mauritius, South Africa, Botswana and Namibia; the partly-free countries were Malawi, Mozambique, Seychelles, Lesotho, Tanzania and Zambia; and the unfree countries were the DRC and Swaziland as well as Angola and Zimbabwe, which had elections, but freedoms were eroded. The single party states declined from eight to zero during this period (excluding the DRC and Swaziland which may be regarded as no-party states). South Africa became a fully-fledged institutional democracy in 1994, whereas only the DRC and Swaziland remained non-electoral autocracies; dictaduras, i.e. elected governments, but with low levels of political and civil liberties included Angola, Zambia and Zimbabwe in 2002; and democraduras, i.e. elected governments but with middle levels of political and civil liberties included Lesotho, Malawi, Mozambique, Seychelles and Tanzania; and finally those in the higher consolidation zone (freedoms are adequately high, but HDIs fall short of high), included Mauritius, South Africa and Botswana.

In 2002 civil liberties in the region remained at a higher (that is, better) level than political rights when the DRC remained least free in terms of political rights, followed by Swaziland and Angola (an electoral democracy without a functioning multiparty parliament) in the second worst position. Angola was therefore rated at the same level as Swaziland, despite having had multiparty elections in 1992. No doubt, the civil war waged by the opposition after having lost the election explains this state of affairs.

The third worst political rights assessment obtained in Tanzania, Zambia and in Zimbabwe, despite the fact that Zimbabwe has had regular elections since 1980. According to Freedom House, political rights in Zimbabwe dropped to six in 2002, a score normally associated with autocracies. Its civil liberties are also scored at six. The electoral democracy in this country had clearly not prevented the erosion of civil and political liberties during this period, dropping to the level of dictaduras. As will be pointed out in the next section, Zimbabwe’s economy has shrunk, per capita has declined and its HDI ranking is down, suggesting that if affluence is needed for the consolidation of democracy—as many analysts maintain—then an increase in economic hardship may be seen as a prerequisite for the erosion not only of democracy, but of an autocracy as well. The strong performance of the opposition during the elections of 2000 and 2002, despite the erosion of the rule of law, may be proof hereof.

A significant correlation in this regard is that all these states with eroded political rights correlate with countries with presidential executive systems, that is, where presidents or premiers and their executives are not primarily accountable to parliament. The correlation goes even further: all states that scored either 1 or 2 (‘free’) in terms of political rights and civil liberties are consistently found in parliamentary and not in presidential systems (i.e., Mauritius, South Africa and Botswana), validating Linz’s views. (Lesotho has the other parliamentary system in the SADC region—it had free and fair elections in May 2002). As argued before, the positive aspects of list proportional representation electoral systems for political stability in deeply divided societies, are quite evident.

Economic and social data for member states in the SADC region

Of the 14 SADC member states only six had per capita incomes in 2000 of US$1,000, or higher. They are Seychelles, Mauritius, Botswana, South Africa, Namibia and Swaziland, and in that order. Except for South Africa and Swaziland where the trends were downwards, the trend for the remaining four were upwards. The regional average was also up slightly. Among the eight states with per capita incomes of US$1,000 or lower, only one was in the US$500–US$1,000 bracket, namely Lesotho, and here the trend was downwards. Among the seven with per capita incomes of US$500 or lower, the trend was also downwards, except for Tanzania and Mozambique where increases were quite significant, albeit from low bases. The rest drifted downwards, especially Zimbabwe, Angola and the DRC—all with serious internal conflicts (Table 2). It can therefore be concluded that conflicts are at odds with both development and democracy.

Other social indicators related to HDIs (Table 3) make up what is referred to as the development index. Not a single SADC state is in the ‘high’ HDI category. Seychelles and Mauritius are in the high medium category, with South Africa, Swaziland, Namibia, Botswana, Lesotho and Zimbabwe (in that order) in the low medium category; and with the rest, namely DRC, Zambia, Tanzania, Angola, Malawi and Mozambique (in that order) in the low HDI category.

The regional trend was downward. There were only two upward trends, namely Seychelles (the best trend) and Namibia. Still, relatively well-placed countries such as Mauritius, South Africa, Namibia and Botswana dropped relatively, which does not augur well for the region in terms of global standards.

In Table 4, the rankings in Tables 2 and 3 are added together making for an aggregate. This created combined regional rankings which may also serve as an aggregate development index. The weakness of this index is that it expresses a numerical index without weighting the development criteria. But the strength is that it does give a relative regional index, which is then (Figure 2) bracketed into the high, high medium, low medium and low HDI categories. Due to per capita incomes of higher than US$3,000, South Africa and Botswana join the high medium group, with Zimbabwe dropping out of low medium into the upper end of the low category.



Such an index is not perfect, but useful in understanding the general trends in the region. It serves as a timely reminder that democratisation processes in the region are fraught with obstacles. If it is true that one cannot have democracy without development, then the democratic future of the SADC region is in doubt, especially for the bottom seven which are all classified as countries with low human development. In those cases where rights and liberties have also dropped—Zimbabwe and Malawi—the prognosis is even worse.

The final ranking for the development index (Table 4), that combines the per capita ranking (Table 2) and HDI indicators (Table 3), is as follows.

High: None

High medium: Seychelles, Mauritius, South Africa and Botswana

Low medium: Swaziland, Namibia and Lesotho

Low: Zimbabwe, Zambia, Angola and Tanzania, DRC, Mozambique and Malawi jointly in the last position.

This ranking is now superimposed on to the freedom index in Figure 2, making the final democracy ‘mind map’ for the SADC region in 2000.

Ethnic structures of member states in the SADC region

Only three states in the SADC region are ethnically homogeneous (Lesotho, Seychelles and Swaziland)—the rest are either compact, that is a multi-ethnic country, but one group constitutes a numerical majority (Botswana, Mauritius, Namibia and Zimbabwe), or all groups are minorities, making the composition heterogeneous (Angola, DRC, Malawi, Mozambique, South Africa, Tanzania and Zambia) (Table 5). They are the typically deeply divided countries.



This multi-ethnic problem goes back to colonial times when boundaries were drawn artificially and heterogeneity became the standard. The outcome of this situation is that single nation, or mono-ethnic states in Africa—such as the majority of those historical pre-colonial states that had survived with their original boundaries more or less intact—became the exception rather than the rule. If homogeneity is preferable, as argued by Crawford Young50 and Freedom House,51 this may indicate good fortune as argued before. Then there are states with compact majorities where the composition is heterogeneous, but one particular ethnic group constitutes a numerical majority of 50% or more. This easily leads to dominant party rule where the ruling party has a strong ethnic basis as in Zimbabwe (Mashona) and Botswana (Tswana).

Then there is the dominant pattern in Africa, characterised by heterogeneity. It is mainly in those cases that state leaders often designed measures—commonly but misleadingly described as nation-building52—that aimed at the stifling of divisive elements within those states, creating the basis for authoritarian rule. As argued earlier, it would have been better for electoral democracies in deeply divided societies to have chosen list proportional representation systems, and not first-past-the-post systems. The former one-party states of the DRC, Malawi, Tanzania and Zambia are examples. In this way heterogeneity contributed towards the demise of the politics of contestation, and became the justification for non-contestational authoritarian rule.

The natural heterogeneity of typical colonies was exacerbated in settler societies such as in Algeria, Kenya, Zambia, Zimbabwe, Namibia and South Africa. This factor added race (for example Arabs, Asians and whites) to the complexities of nation-building in Africa. It also slowed down the tempo of decolonisation (Africa’s first liberation) as many settlers were often reluctant to embrace majoritarian (black) rule. This is why liberation wars were fought in Algeria, Kenya and Southern Africa’s settler societies where large numbers obtained. If consensus or consociational democracy were the alternative, political violence might have been reduced, even eliminated. But decolonisation as an extension of self-determination would not have been prevented. It was inevitable that all citizens would want to participate in public life and by means of free and fair elections in democracies that met the other institutional requirements as well. It is here that list proportional representation seems better than majority-based systems.

The correlations that emerged, albeit crudely, were that homogeneous societies and compact majorities (i.e. where one group is a numerical majority within a multi-ethnic state) produced greater stability and more regime continuity especially in list proportional electoral systems. However, Swaziland should serve as a reminder that homogeneity does not guarantee democracy. To the extent that elections were held, rough correlations between ethnic demography and party support patterns showed very little variation in years to come. To be sure, it is a typical feature of African politics that a ‘re-communalisation’ of politics takes place—that is, political parties are increasingly identified with ethnic groups,53 especially where economic opportunities become scarcer because of stagnant economies.

In African states with low levels of human development and with numerous ethnic minorities (typical heterogeneity), less stability prevailed as political contests easily turned into spoils for wealth and ethnic violence making for unstable states.54

Various authors began exploring these phenomena, seeking causality, going beyond theoretical propositions and empirical correlations (see, Horowitz;55 Cohen;56 Sklar;57 Rabushka & Shepsle;58 and Cross59). We shall not seek those types of explanations in this study (because of limitations explained earlier), only to register and compare the correlations and see whether and what kind of patterns emerged, from which further generalisations may be generated.

Given these observations, what are the correlations between Figure 2 and Table 5? First, with the exception of South Africa, all the other states classified as deeply divided are either autocracies or at best electoral democracies—that is, where elections take place but freedoms are compromised. Second, that none of the states in the consolidation zone are classified as homogeneous. This seems to suggest that while homogeneity is not a precondition for democracy, heterogeneity is a special problem that might call for special constitutional arrangements (Linz and Stepan), such as features of consociation (Lijphart 1999, and Powell 2000) or proportional representation. Third, not a single state in the consolidation zone—irrespective of ethnic structures—has low human development rankings. This reinforces the notion that poverty and democracy do not mix. From this point of view, the only candidates for elevation (with sufficient development and homogeneity) into the consolidation zone are Seychelles, and perhaps Swaziland. They are therefore underperforming democratically, based on potential. Whether they proceed to higher levels of democracy might depend less on development and more on qualitative aspects such as leadership and commitments to freedom, referred to by Sam Decalo and Andreas Schedler.

Regional integration and democracy: what relevance?

The co-operation and integration models applied to SADC are state-based, that is, only states can become members. From this perspective, SADC is an international organisation striving for functional co-operation in many development sectors. The end result is not one of supranational integration where national sovereignties are so diminished that a United States of Southern Africa can be foreseen. The immediate scenario is therefore not of a single market, single currency and supranational parliament. States should also meet certain convergence criteria for such integration.
In Europe, convergence criteria are mainly economical. In SADC there are no such criteria. Participation remains voluntary. National sovereignty therefore remains unaffected, except to the extent that multilateralism imposes itself in terms of common standards and practices on the participating states. But political integration remains a distant vision of Pan-Africanists and the African Union alike. Much more likely is that some members of the Southern African Customs Union, for example Lesotho, could become a tenth province of a greater South Africa. But this would rather be a case of incorporation and not integration.

One of the questions raised earlier was whether functional co-operation structures such as SADC, or even neo-functional integration structures such as the envisaged SADC free trade area, could serve to promote democracy in the region.

On this point it is easy to agree with Peter Meyns60 that there is apparently no causal link between the deepening of integration and the consolidation of democracy, unless, of course, good governance and the respect for human rights are made preconditions for participation, as is the case in the New Partnership for Africa’s Development (NEPAD). But if co-operation and integration do succeed in promoting development of the kind that reduces socio-economic inequalities among and within the participating states, then on the basis of the assumptions made in this paper, democracy may in fact be advanced. Conversely, the lack of meaningful co-operation and integration could hinder (erode) the promotion of democracy, especially in those cases where nations do not meet the good governance criteria for participation in NEPAD. This could be the case if they are guilty of deviant behaviour, or are too pre-occupied with domestic or international conflicts making for a redirection of their public resources away from warfare to welfare too problematic.

Peace and stability are therefore prerequisites for sustainable development and democracy. Conflict prevention becomes such a precondition. A functioning regional security organ that would prevent conflicts could therefore be a greater prerequisite for the deepening of democracy than some of the purely economic objectives of the older SADC, that had little, if anything, to do with successful conflict prevention, peacemaking or peacekeeping. Of late, the NEPAD initiative known as the African Peer Review Mechanism could establish convergence criteria related to the rule of law, the respect for human rights and the democratic accountability of its governments that SADC becomes not only more integrated, but also more democratic. In this way, compliance with good governance could promote development and democracy and regional co-operation, and in this sequence. But it is a long-term project.

Conclusion

The institutional environment among SADC member states is much more conducive to democratic endurance and even consolidation now, than in 1989. Parliaments have made a comeback, except in Swaziland where the monarchy resists democracy, and in the DRC where warfare until recently consumed everything. The South Africa–led peace process is, however, a step in the right direction. Although the outlook is mixed in the rest of the region, with Malawi and Zimbabwe deteriorating, the outlook improved for Tanzania and Mozambique.

But the socio-economic conditions necessary to make electoral democracies as well as those in the consolidation zone stronger, remain rather weak. The country with the best potential to improve its freedom index in order to move into the consolidation zone, is Seychelles. Namibia and Lesotho are also close to this zone. Swaziland must democratise first. Here the limitations are mainly institutional. Seychelles, Swaziland and Lesotho also have the good fortune of homogeneity.

For the DRC and Angola, the outlook is more daunting. They need peace first, then development, and maybe democracy could become possible thereafter. Although the outlook for Tanzania and Mozambique is improving, even they are as yet simply too poorly developed to make them strong candidates for the consolidation zone. In Zimbabwe, general lawlessness must come to an end. The institutions are in place. This implies a vast improvement in its performance on civil liberties and political rights.

Of the three states in the consolidation zone, Mauritius is the only one that has already achieved Huntington’s two-turnover test, which is similar to Przeworski’s (2000) minimalist requirement that incumbent parties actually lose elections and is, together with its high ranking on the human development index, well-placed to become SADC’s first consolidated democracy. For the time being, South Africa and Botswana are best described in O’Donnell’s terms61 as ‘delegated democracies’—not consolidated, but enduring.62 What they need to do most is to uphold their liberties and reduce inequality through economic development plus the strengthening of the middle class, and job-creating growth. That is what the theories suggest, and we agree.

Notes

  1. An earlier version of this article entitled Development and democracy in the SADC region, was published in D Hansohn (ed), 2002. Monitoring regional integration yearbook, 2, 2002. Nepru (Namibia) and Konrad Adenauer Foundation, Windhoek. This latest version benefits from the work done by Vanhanen (1997), Lijphart (1999), Powell (2000), Przeworski, et al (2000), Human Development Indexes (2000) and Freedom House (2002).

  2. J J Linz & A Stepan, Toward consolidated democracies, Journal of Democracy, 7(2), April 1996, p 14.

  3. A Przeworski, M Alvarez, J A Cheibub & F Limongi, What makes democracies endure?, Journal of Democracy, 7(1), January 1996, pp 40-41.

  4. T Vanhanen, Prospects of democracy: A study of 172 countries, London, Routledge, p 41.

  5. R A Dahl, Polyarchy: Participation and opposition, New Haven, Yale University, 1971.

  6. Linz & Stepan, op cit, p 15.

  7. Dahl, op cit.

  8. Linz & Stepan, op cit, p 17.

  9. G Sartori, The theory of democracy revisited, Chatham, Chatham House, 1987, p 13.

  10. M Bratton, Second elections in Africa, Journal of Democracy, 9(3), 1998, p 52.

  11. A Schedler, What is democratic consolidation?, Journal of Democracy, 9(2), 1998, pp 93-96.

  12. C Young, Africa: An interim balance sheet, Journal of Democracy, 7(7), July 1996, p 254.

  13. S P Huntington, The third wave: Democratisation in the late twentieth century, Oklahoma, University Press, 1991, pp 266-267.

  14. A Przeworski, M Alvarez, J A Cheibub & F Mongi, Democracy and development: Political institutions and well-being in the world, New York, Cambridge University Press, 2000.

  15. Schedler, op cit.

  16. K Bollen & R Jackman, Democracy, stability and dichotomies, American Sociological Review, 54, 1989.

  17. Vanhanen, op cit, p 41.

  18. P Meyns, Political integration – whither Southern Africa, in C Peters-Berries & M Marx (eds), Monitoring the process of regional integration in SADC. Harare, Konrad Adenauer Foundation, 2000, p 86.

  19. J Linz, The perils of presidentialism, Journal of Democracy, 1(1), 1990.

  20. A Lijphart, Democracies: Patterns of majoritarian and consensus government in 21 countries, New Haven, Connecticut, 1984; and A Lijphart, Patterns of democracy: Government forms and performance in 36 countries, New Haven, Yale University Press, 1999.

  21. G Bingham Powell, Elections as instruments of democracy: Majoritarian and proportional views, New York, Yale University Press.

  22. Linz & Stepan, op cit, pp 24-27.

  23. D Pottie, Electoral management and democratic governance in Southern Africa, Politikon, 28(2), November 2001.

  24. L Diamond, Economic development and democracy reconsidered, in L Diamond & G Marks (eds), Re-examining democracy: Essays in honour of Seymour Martin Lipset, London, Sage, pp 125-128.

  25. R Joseph, Africa 1990-1997: From abertura to closure, Journal of Democracy, 9(2), April 1998, pp 3 & 14.

  26. R Lemarchard, Africa’s troubled transitions, Journal of Democracy, 3(4), 1992, p 100.

  27. M F Plattner, Liberalism and democracy. Can’t have one without the other, Foreign Affairs, 77(2), March/April 1998, pp 171-180.

  28. Przeworski, et al, op cit, 1996, p 40; and L Armjijo, T Biersteker & A F Lowenthal, The problems of simultaneous transitions, Journal of Democracy, 5(4), October 1994, pp 162-168.

  29. Ibid, p 166.

  30. B Moore, Social origins of dictatorship and democracy: Lord and peasant in the making of the modern world, Boston, Beacon Press, 1996; and K Good, Development and democracies: Liberal versus popular, African Insight, 27(4),1997, p 253.

  31. See, J Schumpeter, Capitalism, socialism and democracy, New York, Harper, 1947.

  32. S M Lipset, Political man: The social bases of democracy, Garden City, Double Day, 1959.

  33. Diamond, op cit, pp 125-128.

  34. Przeworski, Consolidating third world democracies, in K Cullinan, Durable democracies, Democracy in Action, 9(5), October 1995.

  35. Przeworski, et al, op cit, 1996, p 41.

  36. Przeworski, et al, op cit, 2000.

  37. Moore, op cit, p 453.

  38. Ibid, p 418.

  39. J Raynor, The middle class, London, Longmans, 1967, pp 27-30.

  40. Przeworski, et al, op cit, 1996, pp 42-43.

  41. Linz & Stepan, op cit.

  42. M Bratton and R Mattes, Africa’s surprising universalism, Journal of Democracy, 12(1), January 2001, p 120.

  43. S Decalo, On statistical correlations of democratisation and prospects of democratisation in Africa: Some issues of construction, inference and prediction, in Tatu Vanhanen (Ed). Prospects of democracy: A study of 172 countries, London, Routledge, 1997, p 312.

  44. Diamond, op cit.

  45. E Rijnierse, Democratisation in sub-Saharan Africa? Literature overview, Third World Quarterly, 14(3), 1993, pp 647-653.

  46. Young, op cit, p 11.

  47. Ibid, p 13.

  48. Freedom House, Freedom in the World 2002: The Democracy Gap, <www.freedomhouse. org/research/ survey2002.htm>

  49. E Gyimah-Boadi, The rebirth of African liberalism, Journal of Democracy, 9(2), 1998, pp 25-27.

  50. Young, op cit.

  51. Freedom House, op cit.

  52. C Clapham, Democratisation in Africa: Obstacles and prospects, Third World Quarterly, 14(3), 1993, p 428.

  53. G Hyden, Democratisation and the liberal paradigm in Africa, Africa Insight, 27(3), 1997, p 163.

  54. Joseph, op cit.

  55. D L Horowitz, Ethnic groups in Conflict, Berkeley, University of California, 1985, pp 681-684.

  56. A Cohen, Custom and politics in urban Africa, Berkeley, University of California, 1969.

  57. R Sklar, Ethnic relations and social class, in A O Sanda (ed), Ethnic relations in Nigeria. New York, 1976.

  58. A Rabushka & K A Shepsle, Politics in plural societies: A theory of democratic instability, Columbus, Merrill, 1972.

  59. M Cross, Colonialism and ethnicity: A theory comparative study, Ethnic and Race Studies, 1(1), 1978.

  60. Meyns, op cit, p 85.

  61. G O’Donnell, Delegative democracy, Journal of Democracy, 5(1), January 1994.

  62. A Habib, Myth of the Rainbow Nation: Prospects for the consolidation of democracy in South Africa, African Security Review, 5(6), 1996.