Public Opinion and Democratic Consolidation
More than 20 years have passed since the Berlin Wall came down, an event followed in sub-Saharan Africa by pressures for political liberalization and transitions to multi-party rule. In addition, more than 10 years have elapsed since the Afrobarometer embarked (in Ghana in 1999) on a pioneering effort to conduct surveys of public opinion about these changes. The Afrobarometer Network – an international consortium of researchers1 – has since accumulated interviews with over 100,000 Africans in four rounds of surveys in up to 20 countries.2
The time is ripe, therefore, to assess the current state of political development in these countries from a citizen’s perspective and to track changes in public attitudes that have occurred over the first decade of this century (1999–2008).3 The central question concerns the fate of democracy. Do Africans say they want democracy, a preference that we call the ‘popular demand’ for democracy? Do they think they are getting it – that is, do they perceive that their leaders are providing a supply of democracy? Moreover, if citizens see evidence of democratic development in Africa, to what extent are these regimes established, stable, or consolidated?
Of course, it would be a mistake to view African politics exclusively through the lens of democracy. After all, most African countries have had limited experience with a form of political regime more commonly associated with the mature polities of the West. For this reason, we wish to keep an open mind about whether political regimes in Africa are best characterized as democracies or as some other regime configuration. After all, regimes may survive in various forms, including as autocracies or as hybrid regimes that are neither fully authoritarian nor completely democratic.
With reference to particular countries, this chapter arrives at three conclusions:
Evidence in this chapter is derived from face-to-face interviews in the languages of the respondents’ choice with more than 25,000 Africans in 2008 and 2009. This unique dataset gives voice to ordinary people whose views about democracy and development are all too often overlooked by ruling political elites. In focusing analysis on individual attitudes and behaviours, the Afrobarometer seeks both to provide a counterweight to the elite-centred focus of many political studies of Africa and systematically to achieve a measure of scientific precision and generalization that is unavailable from most anthropological and historical case studies. In the decade that Afrobarometer data have been available, scholars have generated scores of articles, working papers and monographs on popular attitudes towards democracy, governance, leadership, and corruption, as well as on everyday practices of economic livelihood and political participation, and many other topics (see the bibliography).
The Meaning of ‘Democracy’
Before attempting to measure popular attitudes to democracy, we must ensure that respondents to Afrobarometer surveys have a similar object in mind. We offer evidence to suggest that Africans regard democracy in reasonably standard fashion. When asked in 1999 and 2005, ‘what, if anything, does democracy mean to you?’ three-quarters of those surveyed could offer a definition. Respondents in almost all countries ranked personal liberty in first place, with a consistent average of 41 per cent mentioning liberty in both surveys. The next most common substantive meaning was ‘government by the people’, which was mentioned by 16 per cent and 10 per cent, respectively, in 1999 and 2005. Significantly, the rank order of meanings was virtually identical across all countries.4
To test further for shared meanings in 2008, we asked respondents to compare two hypothetical African regimes. Is Country A (with free speech, multi-party competition, and electoral turnover of leaders) more or less democratic than Country C (with restricted speech, a dominant party and regular re-election of incumbents)? Some 71 per cent of respondents ranked Country A as more democratic and only 3 per cent as less democratic than Country C.5 To be sure, popular understandings of democracy are not yet universal, but these results suggest that Africans share enough of an emergent consensus on what democracy is to allow valid and reliable comparisons of mass attitudes to political regimes.
Popular Demand for Democracy
Do Africans say they want democracy? To start investigating this question the Afrobarometer asks a standard question, the wording of which is given at the bottom of Figure 22.1. Respondents who say that ‘democracy is preferable to any other kind of government’ are deemed to support democracy.
Figure 22.1Support for democracy, 2008
* Mean excludes Zimbabwe
“Which of these three statements is closest to your own opinion?
A. Democracy is preferable to any other kind of government;
B. In some circumstances a non-democratic government can be preferable;
C. For someone like me, it doesn’t matter what form of government we have.”
In 2008, an average of 70 per cent of Africans interviewed in 19 countries expressed overt support for democracy.6 However, there was considerable cross-national variation around this mean value, ranging from 85 per cent support in Botswana to 39 per cent support in Madagascar. This distribution of opinion accords with each population’s political experience: Batswana apparently derive their regime preferences from more than 40 years of stable multi-party rule, whereas Malagasy convey concern over a recent history of irregular elections, mass protests and non-constitutional power grabs.
A critic might complain that public opinions are unreliable because citizens lack knowledge about democracy or easily acquiesce to a socially approved symbol. After all, it is as easy to express a favourable opinion about motherhood as it is to claim support for democracy. However, the substance of democratic preferences is affirmed by the fact that less than 7 per cent of Afrobarometer respondents in 2005 said they support democracy without being able to offer a definition of its meaning.7
Moreover, we probe whether professed democrats are willing to countenance alternative authoritarian regimes. Figure 22.2 compares the proportions of those interviewed across Afrobarometer countries that express disapproval of military rule. A higher mean proportion of respondents state opposition to this form of authoritarian governance (75 per cent) than profess pro-democratic sentiments (70 per cent), suggesting that Africans remain clearer about the kind of government they don’t want than the kind they affirmatively desire. Reflecting strong commitments to civilian rule, Kenyans display the most vigorous rejection of an army takeover (94 per cent). Burkinabe, however, whose leader first rose to power by means of a military coup, are more ambivalent (50 per cent). Majorities of national populations also roundly reject two other authoritarian regimes common to Africa: one-party rule and personal rule by a strongman (not shown). Note, however, that a majority of Mozambicans barely rejects one-party rule (51 per cent), suggesting either popular wistfulness for a pre-democratic past or a genuine, if mistaken, belief that one-party arrangements are democratic.
Figure 22.2Rejection of military rule, 2008
* Mean excludes Zimbabwe
“There are many ways to govern a country. Would you disapprove or approve of the following alternative: The army comes in to govern the country?”
A tougher test of deep popular commitment to democracy, however, is whether citizens both support democracy and reject a range of authoritarian alternatives (including one-party and one-man rule). We measure this composite attitude with an index of demand for democracy. Figure 22.3 indicates that in 2008 almost all Africans interviewed rejected at least one form of autocracy (90 per cent). However, far fewer rejected two or three such alternatives, and less than half of all respondents (45 per cent) expressed a robust demand for democracy that was unqualified by any kind of authoritarian nostalgia. In other words, the commitments of most Africans to democratic ideals remain quite shallow.
What has happened to demand for democracy over time? Figure 22.4 shows trend data for the 11 countries for which we have four survey observations between 1999 and 2008.8 On average, popular support for democracy was initially quite high in the aftermath of the regime transitions of the 1990s (68 per cent), though it began to dissipate by 2005 (62 per cent). In the last three years of the period covered, however, we discern a sharp upsurge in the mean level of expressed support for democracy (to 72 per cent). The reasons remain unclear. Perhaps economic recovery spurred by a commodity price boom has encouraged growing confidence in the political regime. However, support for democracy in 2008 is only weakly correlated with perceptions of improvements in macroeconomic conditions over the previous 12 months. More likely, citizens perceive that political reforms are taking root; after three or four rounds of competitive elections, including several turnovers of ruling parties, they are gaining confidence in the institutionalization of their right to choose leaders (see Lindberg, this volume).9
We also discern a steady upward trajectory of anti-authoritarian sentiments. Each time an Afrobarometer survey has been conducted, more and more Africans report that they reject all three forms of authoritarian regime, which suggests that one-party, military, and oneman rule are becoming less appealing over time. This trend is consistent with the diminished frequency of military coups in Africa over the past 20 years and the growing unwillingness of the international community, including even the African Union, to recognize such illegal transfers of power.
As expected, demand for democracy echoes the arcs of its component attitudes, moving up from 39 per cent to 45 per cent between 1999 and 2008. However, before we prematurely celebrate the consolidation of democracy in Africa, we must note that on average robust demand for democracy remains a minority sentiment in the countries we have studied.
The Perceived Supply of Democracy
On the supply side, the key question we need to ask is, do ordinary people think that they are getting democracy? One way to generate an answer is to ask everyone, ‘how much of a democracy is (this country) today?’ Response categories for this item range on a four-point scale from ‘a full democracy’, through ‘a democracy with minor problems’ and ‘a democracy with major problems’,to ‘not a democracy’.
Sceptics might again argue that non-literate people in the developing world are insufficiently knowledgeable or experienced about democracy to offer meaningful responses. However, in 2005 less than 10 per cent of respondents who had an opinion on the extent of democracy were unable to define the term. While this uninformed group was slightly more prone innocently to perceive ‘a full democracy’, their views on the extent of democracy otherwise resembled the opinions of those who better understood the nature of the regime.
By 2008, an average of 59 per cent of all Africans interviewed considered that they lived in a full or almost full democracy. The reader should recall, however, that because Afrobarometer surveys are conducted in Africa’s more open societies, this figure does not necessarily apply to the continent as a whole. The range of responses in Figure 22.5 is wider for the perceived extent of democracy than for any other item of opinion considered here.
Figure 22.5Perceived extent of democracy, 2008
* Mean excludes Zimbabwe
“In your opinion, how much of a democracy is (your country) today?”
While the citizens of Botswana again lead the pack (at 91 per cent), with Ghanaians close behind (at 83 per cent), Zimbabweans trail far below (at a dismal 14 per cent in 2005). With regard to the perceived extent of democracy, countries can be roughly divided into three groups of roughly equal size:
It is within this last group that we discern regimes at risk, for example in Kenya, Nigeria, Senegal, Madagascar, and Zimbabwe.
To check our measurements we offer two tests. First, to be externally valid, African public opinion should be consistent with judgements about the level of democracy made independently by international experts. Figure 22.6 compares the Afrobarometer’s extent of democracy with the well-known status of freedom score published annually by Freedom House. The scatter-plot shows that in 2008 experts and citizens converged on assessments of the level of democracy for countries like Kenya, Mozambique, Mali, and Benin. To be sure, the experts think that South Africa is more democratic than do its citizens, and Tanzanians think they have more democracy than professionals would grant. However, all told, the fit of the country cases to the shared regression line is good.
Second, as a test of internal validity, we would expect popular perceptions of the extent of democracy to be closely correlated with expressed satisfaction with ‘the way democracy works’. Figure 22.7 shows that, on average, less than half (49 per cent) of all Africans interviewed were satisfied (either ‘fairly’ or ‘very’) with democracy in 2008. However, the cross-country distribution for satisfaction closely resembles that for the extent of democracy; for example, democracy is again at risk in exactly the same group of countries. Moreover, satisfaction with democracy is closely associated with the perceived extent of democracy, not only at the country level, but also at the individual level.
Figure 22.7Satisfaction with democracy, 2008
* Mean excludes Zimbabwe
“Overall, how satisfied are you with the way democracy works in (your country)?”
As Figure 22.8 shows, fully 81 per cent of those who are satisfied with democracy also perceive extensive democracy.
Because of the coincidence of these measures, we combine the indicators for satisfaction with democracy and perceptions of the extent of democracy into an additive construct of the supply of democracy. It measures whether individual Africans are both satisfied with democracy and perceive it to be extensive in their country. The evolution of this indicator, along with its component attitudes, is shown in Figure 22.9, which illustrates several noteworthy trends:
It would appear, therefore, that the Africans we interviewed think that they have more democracy today than earlier in the decade. Most importantly, we note that the perceived extent of democracy has risen even as satisfaction has dropped. This unexpected result suggests an adjustment of mass expectations. People are recognizing a measure of democratic progress even as they realistically conclude that actual democratic practice is falling short of their dreams.11
A Model of Regime Consolidation
So how many African regimes are democracies, at least in the eyes of their citizens? Which regimes in Africa are consolidating? In Figure 22.10 we suggest a schema for addressing both the nature and the stability of various political regimes. It builds upon the political indicators of supply and demand that we have already described. The logic of the model is that in consolidated regimes demand and supply are in equilibrium (or balance), a condition represented by the diagonal intercept that dissects the space in Figure 22.10.
Figure 22.10Consolidation of political regimes
Assumptions:
1. Intercept line represents consolidated regime (equilibrium of demand and supply)
2. Dotted lines represent margin of sampling error around survey point estimates
3. Equilibrium at 70 per cent or higher represents consolidated democracy
4. Equilibrium at 30 per cent or lower represents consolidated autocracy
5. Points off intercept line represent unconsolidated regimes
What happens to a new democracy is the outcome of a continuing process of interaction between what elites supply and what the populace demands … Uninterrupted progress towards the completion of a new democracy will occur if popular demands for reforms to improve the regime are met by political elites. This positive equilibrium is often described as a stable or established democracy.
(Rose et al. 1998: 14, emphasis added)
We extend this argument about the consolidation of new democracies to claim that a sustained balance between mass demands and institutional supply signals the consolidation of any type of regime. Whereas a high-level political equilibrium connotes the consolidation of democracy, a low-level equilibrium signals the consolidation of autocracy.
For democracies, how high must a political equilibrium be? Let us assume that the consolidation of democracy minimally requires that for a sustained period of time, 70 per cent or more of the adult population wants this type of political regime and a similar proportion thinks they are getting it (Diamond 1999: 68). When these circumstances occur, the probability that democracy will break down declines considerably. In Figure 22.10, democracies are found in the upper-right bloc and are consolidated to the extent that they approach, or lie on, the equilibrium line within this space.
At the opposite end of the spectrum (in the lower-left segment) lie autocracies. In these regimes the populace neither demands democracy nor perceives its supply by the state elite (scoring 30 per cent or lower in each case). Because strong initiatives for democratization do not emanate from above or below, the regime is caught in a low-level equilibrium trap. The closer that actual regimes approach the equilibrium line, and do so over successive surveys, the more that autocracy is consolidated.
Regimes that lie elsewhere in the property space are hybrid regimes; that is, regardless of their formal characteristics, their citizens perceive neither democracy nor autocracy but something in between. Hybrid regimes may consolidate at intermediate levels, lending permanence to forms like electoral democracy, electoral autocracy, or other semi-formed systems (see van de Walle, this volume).12 Indeed, the greatest risk to the consolidation of new democracies in Africa is that the architecture of the regime hardens prematurely – that is, before democratic institutions or beliefs have had a chance to take root.
Regimes that fall far from the equilibrium line can be considered unconsolidated. Where demand exceeds supply, citizens will pressure their governments for ongoing democratic reforms. Where supply exceeds demand, we should suspect that elites have room to manipulate the rules of the democratic game. Moreover, because of the imbalance between political demand and supply, hybrid regimes are prone to instability.
Regime Consolidation in Africa
We can now enter Afrobarometer data for 2008 into the regime consolidation model. Figure 22.11 reveals several interesting results.
First, by our standards, there are no consolidated democracies among Afrobarometer countries. Botswana comes closest, where 65 per cent of adults demand democracy and 80 per cent think that democracy is institutionalized. To the extent that supply exceeds demand, however, Botswana’s democracy is not yet consolidated. Second, and by contrast, Lesotho is apparently caught in a low-level trap: in 2008, it could even be classified as a consolidated autocracy by our standards.13 Only 23 per cent of Basotho demand democracy and 18 per cent perceive a supply. Despite a façade of parliamentary institutions, the country’s political culture seemingly still manifests monarchical and military legacies inherited from the past.
Third, all other countries possess hybrid political regimes. Democratic demand and the perceived supply of democracy reach intermediate levels in Benin, Malawi, Mali, and South Africa. The coordinates for these cases fall close to the equilibrium line. We infer from their spatial location that their regimes are consolidating as hybrids that fall short of full democracy. Moreover, with supply and demand in balance, there is no political force propelling major political changes in these countries in the foreseeable future.
Fourth, most hybrid regimes are unconsolidated, but in distinctive ways. On the one hand, citizens demand more democracy than elites are willing to supply in places like Kenya and Zambia. Indeed, Zimbabwe in 2005 displayed high demand (56 per cent) but the lowest supply (10 per cent) of any Afrobarometer country, which suggests that Zimbabwe could register quick democratic progress if ever a new government were installed as a result of a free and fair election. Because demand is twice as high as supply in Nigeria and Senegal, one can also predict continued popular pressures from below for further democratic reforms in these countries.
Conversely, political regimes like those in Tanzania and Namibia are unconsolidated for other reasons. Citizens in these places report that their governments provide more democracy than they really want. In Burkina Faso, the supply of democracy (36 per cent) exceeds demand (18 per cent) by a factor of two to one. With this profile of opinion, the general public in this group of countries is predisposed to easily acquiesce to strong leaders who control dominant political parties or originate from the armed forces. Unless these autocratic elites reform themselves, any change in these countries is likely to move away from, rather than toward, consolidated democracy.
Diverse Trajectories
A major result of this analysis is that African political regimes are diverse along at least two dimensions. With regard to the nature of the regime, some are nearly democratic, a few seem autocratic (at least in the eyes of the citizenry), while most fall into an intermediate hybrid category. A common African regime type is electoral democracy, meaning democratic in institutional form (e.g. elections), but lacking essential attributes of rights availability or leadership accountability. With regard to the likelihood of change, some regimes – including some hybrids – have attained a stable equilibrium, but more are unconsolidated.
With Afrobarometer data it is possible to track the political evolution of particular African countries over recent years. Selecting the countries that have undergone the most change, we distinguish autocratizing from democratizing regimes according to their movement away from or toward consolidated democracy. In the interests of brevity, we simply report trends and leave the informed reader to interpret them and identify causal events. Again, the main theme is the diversity of regime trajectories.
Figure 22.12 illustrates four autocratizing regimes:
By contrast, Figure 22.13 selects four democratizing regimes:
Conclusion
At the risk of obscuring the above diversity of regime trajectories, we end with general observations. Figure 22.14 traces the overall trends in popular demand for democracy and the perceived supply of democracy from 1999 to 2008. The referent is a ‘typical’ Afrobarometer polity (as if there were such an abstract entity).
At the outset, in 1999, the Africans we interviewed perceived that on average, the provision of democracy by new multi-party regimes exceeded the level of popular demand. While 48 per cent felt supplied, demand stood at just 39 per cent. At this time, people were unsure about the virtues and vices of a democratic dispensation and therefore hedged their bets. They offered tentative support to democracy but at the same time retained residual loyalties to older, authoritarian alternatives.
By 2005, however, demand for democracy (which rose to an average 44 per cent) had outstripped the perceived supply (which by then had fallen to 40 per cent). We interpret this juxtaposition to mean that Africans began to appreciate the virtues of democracy, especially civil liberties and electoral participation, and so began to put authoritarian attachments behind them. Simultaneously, the experience of living under democratic regimes that rarely performed up to expectations led to increasingly sceptical judgements about how much freely elected governments were likely to achieve.
By 2008, as political rights became institutionalized, demand and supply came into alignment, at least in terms of gross averages across more than 25,000 respondents in 19 countries. Some 45 per cent demanded democracy and 46 per cent perceived a supply. However, it would be a mistake to interpret this result as evidence of an underlying consolidation of Africa’s existing array of regimes. There is too much political diversity across African countries – not only of regime type, but also in terms of the degree of institutional stability – to allow any such conclusion.
We can note both good and bad news. The good news is that democratic attitudes are generally on the rise among the African populations we have surveyed. If sustained, this increase – measured prior to the onset of the global financial crisis in late 2008 – is a promising portent for further democratization. However, the bad news is that less than half of all Africans interviewed demand democracy and perceive its supply when these indicators are measured rigorously. As such, the project of democratization still has a long way to go.
Notes
1The Afrobarometer is a joint enterprise of the Centre for Democratic Development (CDD-Ghana), the Institute for Democracy in South Africa (Idasa) and the Institute for Empirical Research in Political Economy (IREEP, Benin). Fieldwork, data entry, preliminary analysis, and the dissemination of survey results are conducted by National Partner organizations in each African country. Michigan State University and the University of Cape Town provide technical and advisory support services. Several donors support Afrobarometer’s research, capacity building and outreach activities, including SIDA, the UK Department for International Development (DfID), CIDA, the US Agency for International Development (USAID) and the Royal Danish Ministry of Foreign Affairs. For more information see: www.afrobarometer.org.
2Fieldwork for Round 4 Afrobarometer surveys was conducted in 19 African countries between March and December 2008. Due to political unrest, a Round 4 survey could not be conducted in Zimbabwe during 2008; instead this paper refers to results from the Round 3 survey in Zimbabwe of October 2005.
3Afrobarometer surveys can only be conducted in the continent’s most open societies. Hence the results do not represent the continent – or all Africans – as a whole.
4Note, however, that the second most common response was ‘don’t know’, suggesting that Africans still have much to learn about this unfamiliar form of government.
5Some 11 per cent said ‘equally democratic’ and 14 per cent said ‘don’t know’. Country B was a middle category, not reported here.
6All mean scores for 2008 exclude Zimbabwe.
7Of those who could not offer a definition (the 28 per cent who said they ‘don’t know’ any meaning of democracy in 2005), more than half (52 per cent) also said they ‘don’t know’ whether they support democracy and a further fifth (20 per cent) said that the type of regime ‘doesn’t matter’ to them.
8Botswana, Ghana, Lesotho, Malawi, Mali, Namibia, Nigeria, South Africa, Tanzania, Uganda, and Zambia.
9Citizen preferences for elections as a means of choosing leaders and their judgements about the availability of civil and political rights are strongly and significantly related to demand for democracy.
10The position of South Africa, often considered a democratic success story, reflects the instrumental conception of democracy held by many of its citizens. More so than other Africans, they equate democracy with the delivery of social and economic goods rather than the availability of equal political opportunity. This orientation may well be a by-product of the totalizing deprivations of apartheid.
11To confirm, we note that the proportion of dissatisfied democrats (i.e. those who demand democracy but are dissatisfied with its performance) has risen from 34 per cent to 41 per cent. The latter figure would be even higher if Zimbabwe were included in the 11-country sample for which we have trend data.
12An electoral democracy holds competitive elections, but respect for civil liberties is incomplete. An electoral autocracy holds sham elections that the opposition can almost never win.
13Moreover, Lesotho has sustained this status since 1999 (see Bratton et al. 2005: ch. 13).
Bibliography
Bratton, M. and Mattes, R. (2001) ‘Support For Democracy in Africa: Intrinsic or Instrumental?’, British Journal of Political Science 31: 447–474.
Bratton, M., Mattes, R. and Gyimah-Boadi, E. (2005) Public Opinion, Democracy and Market Reform in Africa, New York, NY: Cambridge University Press.
Diamond, L. (1999) Developing Democracy, Baltimore: Johns Hopkins University Press.
Eifert, B., Miguel, E. and Posner, D. (2010) ‘Political Competition and Ethnic Identification in Africa’, American Journal of Political Science 54(2): 494–510.
Evans, G. and Rose, P. (2007) ‘Support for Democracy in Malawi: Does Schooling Matter?’ World Development 35(5): 904–919.
Levi, M., Sacks, A. and Tyler, T. (2009) ‘Conceptualizing Legitimacy, Measuring Legitimating Beliefs’, American Behavioral Scientist 53(3): 354–375.
Moehler, D. (2009) ‘Critical Citizens and Submissive Subjects: Election Losers and Winners in Africa’, British Journal of Political Science 39: 345–366.
Rose, R., Mishler, W. and Haerpfer, C. (1998) ‘Democracy and Its Alternatives: Understanding Post-Communist Societies’, Baltimore, MD: Johns Hopkins University Press.
Schedler, A. and Sarsfield, R. (2007) ‘Democrats with Adjectives: Linking Direct and Indirect Measures of Democratic Support’, European Journal of Political Research 46(5): 637–659.
Stasavage, D. (2005) ‘Democracy and Education Spending in Africa’, American Journal of Political Science 49 (2): 343–358.
