lived poverty and the fate of african democracies: insights from the...

26
1 Lived poverty and the fate of African democracies: insights from the Afrobarometer (This is a draft of a work in progress. Please, do not cite without the authors’ permission) Boniface Dulani (University of Malawi) 1 (Corresponding author) Blessings Chinsinga (University of Malawi) 2 "From Aristotle down to the present, men have argued that only in a wealthy society in which relatively few citizens lived at the level of real poverty could there be a situation in which the mass of the population intelligently participate in politics and develop the self-restraint necessary to avoid succumbing to the appeals of irresponsible demagogues." Lipset, 1959:75 Abstract The recent availability of survey data from the Afrobarometer has provided a great opportunity to re-examine the long- held thesis that more well to do nations have the greatest chance of sustaining democracy than poorer ones, with respect to Africa. Given that the poor make up the biggest proportion of the populations of most African countries, a number of studies speculated that Africa’s democratic experiments at the end of the 20 th century were doomed with very high likelihood of reversal. Yet, given that it has been two decades since democracy’s third wave hit the African shores and the fears of full-fledged reversal are yet to be realized, it begets asking the question again as to whether poor African peoples are lesser democrats than their richer counterparts. We replicate the study carried out by Bratton (2005) to assess the effects of poverty on democratic attitudes and political behaviour, using survey data from three Afrobarometer rounds. Our findings corroborate those of Bratton and paint a rather mixed picture on the relationship between poverty and democratic attributes. On one hand, poorer Africans are more likely than their richer colleagues to demand democracy and generally express lower levels of rejection for non-democratic forms of government such as one man, one party and military rule. However, at the same time, the evidence suggests that African peoples that experience extreme forms of poverty have higher expectations than the non-poor about the quality of democracy and therefore express less satisfaction with the performance of their democracies and are less likely to rate their countries as full democracies. While the higher levels of support for non-democratic forms of government among the poor might point towards a higher risk of democratic regression, the fact that Africa’s poor demonstrate little contentment with the quality of democracy on 1 Department of Political Science, 347 South Kedzie Hall, East Lansing, MI 48823, USA. Tel.+ 1 517-353-1686 (O), + 517-488-4825 (Mobile). E-mail: [email protected] 2 Department of Political and Administrative Studies, Chancellor College, P.O. Box 280, Zomba, Malawi. Tel. +265 999-836680, e-mail: [email protected]

Upload: nguyennga

Post on 12-Apr-2018

215 views

Category:

Documents


2 download

TRANSCRIPT

1

Lived poverty and the fate of African democracies: insights from the Afrobarometer

(This is a draft of a work in progress. Please, do not cite without the authors’ permission)

Boniface Dulani (University of Malawi)1 (Corresponding author)

Blessings Chinsinga (University of Malawi)2

"From Aristotle down to the present, men have argued that only in a wealthy society in which

relatively few citizens lived at the level of real poverty could there be a situation in which the mass of

the population intelligently participate in politics and develop the self-restraint necessary to avoid

succumbing to the appeals of irresponsible demagogues." Lipset, 1959:75

Abstract

The recent availability of survey data from the Afrobarometer has provided a great opportunity to re-examine the long-

held thesis that more well to do nations have the greatest chance of sustaining democracy than poorer ones, with respect

to Africa. Given that the poor make up the biggest proportion of the populations of most African countries, a number

of studies speculated that Africa’s democratic experiments at the end of the 20th century were doomed with very high

likelihood of reversal. Yet, given that it has been two decades since democracy’s third wave hit the African shores and

the fears of full-fledged reversal are yet to be realized, it begets asking the question again as to whether poor African

peoples are lesser democrats than their richer counterparts. We replicate the study carried out by Bratton (2005) to

assess the effects of poverty on democratic attitudes and political behaviour, using survey data from three Afrobarometer

rounds. Our findings corroborate those of Bratton and paint a rather mixed picture on the relationship between poverty

and democratic attributes. On one hand, poorer Africans are more likely than their richer colleagues to demand

democracy and generally express lower levels of rejection for non-democratic forms of government such as one man, one

party and military rule. However, at the same time, the evidence suggests that African peoples that experience extreme

forms of poverty have higher expectations than the non-poor about the quality of democracy and therefore express less

satisfaction with the performance of their democracies and are less likely to rate their countries as full democracies.

While the higher levels of support for non-democratic forms of government among the poor might point towards a higher

risk of democratic regression, the fact that Africa’s poor demonstrate little contentment with the quality of democracy on

1 Department of Political Science, 347 South Kedzie Hall, East Lansing, MI 48823, USA. Tel.+ 1 517-353-1686

(O), + 517-488-4825 (Mobile). E-mail: [email protected] 2 Department of Political and Administrative Studies, Chancellor College, P.O. Box 280, Zomba, Malawi. Tel. +265

999-836680, e-mail: [email protected]

2

offer while expressing their support for it, might offset the non-democratic tendencies . The pressure for further

democratic reforms in Africa is thus likely to come from the poor rather than from the contented rich. This scenario can

lead to an uneasy equilibrium, where the political elites are pressurized to maintain certain attributes of democracy

while at the same time being given some leeway to exhibit traits of authoritarianism. This might thus help to shed more

light as to why other studies have come to the conclusion that most African regimes are consolidating as hybrid regimes,

that is neither moving towards full democracies nor regressing to autocracy.

1.0 Introduction

The relationship between democracy and development has been a matter of continuous debate.

Over the last half century, numerous studies have sought to test and establish the relationship

between the two. The evidence adduced from these studies has however led to a rather mixed

picture, where “a number of different predictions about the relationship between economic growth

and democracy…are split sufficiently evenly between positive and negative correlations that no

overall theoretical presumption maybe claimed,” (Minier, 1998:241. See also Remmer, 1995). The

predominant school of thought in this debate has claimed that the existence of mass poverty poses

substantial challenges to democratic survival. The poor, in other words, are considered to make poor

democrats. This proposition is best summed up by Seymour Lipset’s (1963:31) seminal thesis that

contended that “only in a wealthy society in which relative few citizens live at the level of real

poverty could there be a situation in which the mass of the population intelligently participate in

politics and develop the self-restraint necessary to avoid succumbing to the appeals of irresponsible

demagogues.”

The argument that poverty is ill-suited to sustaining democratic politics has led many scholars to

predict that the extension of the third wave of democratization to the developing world in the 1990s,

particularly in Africa, was doomed to failure as many of the countries on the continent are

characterized by widespread and deep-rooted poverty. Magnusson and Clark (2005:553), for

example, contend that the first decade of the third wave democracies in Africa confirm the

proposition that “impoverished states do not consolidate democratic regimes easily that democracy

in experience of democracy.”

While the prevalent wisdom has been that poor people make poor democrats, a new and emerging

literature is beginning to challenge these arguments by proffering evidence that poor people are “no

3

more or less likely than their wealthier counterparts to hold democratic values or to prefer

democracy above other political regimes,” (Bratton, 2008:29). Taking advantage of the expansion of

public attitude survey research that now covers more developing countries, some of this

bourgeoning literature has questioned the continued validity of the old school of thought by

adducing evidence that suggests that poor people in Africa, Latin America and South Asia, do not

value democracy any less than their richer counterparts. As Krishna (2008:9) surmises, these new

works are increasingly demonstrating that the conventional view that poor people make poor

democrats is “wrong, or at least, if ever correct, no longer true”. Instead, the poor in these regions

exhibit as much faith, or greater, in democracy as he non-poor and that poor people participate in

democratic activities no less, and sometimes more, than other citizens.

In his contribution in the Krishna volume, Bratton (2008) has examined how poverty interacts with

various dimensions of democratic citizenship in 15 African countries. Among several findings,

Bratton finds that poorer Africans are less politically tolerant than their wealthier colleagues

although they are more likely to favor political equality; that higher levels of lived poverty are

associated with lower levels of both demand for, and satisfaction with democracy. One of Bratton’s

more interesting finding however is that poverty in Africa is actually associated with higher levels of

political participation, demonstrated by higher propensities to vote among Africa’s poor majority

than the wealthier citizenry.

Despite Bratton’s positive findings, however, questions still remain on whether the widespread and

deep-rooted poverty in Africa can explain a number of challenges with the continent’s democracies

over the last two decades, particularly with regard to the limited numbers of partisan turnovers in

elections as well as evidence suggesting that several of the new regimes are consolidating, not as

democracies, but rather, as hybrid regimes - that is neither democracies nor autocracies (van de

Walle , 2002; van de Walle, 2007; Barkan, 2008; Bratton and Mattes, 2009). The limited number of

partisan turnovers in Africa, described by Lindberg (2006:42) as “the ultimate indicator of

competition,” is reflected by the fact that in the first 15 years after the onset of the onset of

democratic transitions in Africa, there were only 9 turnovers from a total of 44 elections held

between 1989 and 2001 that were deemed free and fair (Lindberg, 2005). Thus despite evidence

suggesting Africa’s majority poor are supportive of democracy and take part in elections more than

the well-to-do, they are not utilizing their electoral clout to vote out parties that are failing to uplift

4

their lives by pursuing policies that can result in substantial reductions in the levels of poverty. Using

Afrobarometer survey data from 2008, Bratton and Mattes (2009:14) meanwhile adduce evidence

that suggests that there are “no consolidated democracies among [19] Afrobarometer countries.”

Instead, the two authors further observe that “almost all other countries possess hybrid political

regimes,” with a number consolidating as hybrids that fall short of full democracy.

Such evidence begets the question: why, if Africa’s majority poor are better democrats than

conventional wisdom would suggest, are African democracies failing to consolidate and exhibit

characteristics that would suggest consolidation? What type of “democrats” are Africa’s poor and

what does this entail about the fate of democracy on the continent? Drawing from Afrobarometer3

survey data collected in three rounds in 2002, 2005 and 2008, we revisit the question of poverty and

democracy in Africa. We use Afrobarometer’s concept of the “Lived Poverty Index,” a composite

measure that examines the experience of poverty from the perspective of the poor themselves. This

indicator measures poverty from four dimensions: experience of going without (a) enough food; (b);

enough medicines or medical treatment; (c) enough fuel for cooking; and (d) a cash income.

Our findings reaffirm those of Bratton (2005; 2008), namely, that Africa’s poor demand less

democracy by holding lower democratic values than their richer counterparts. For example, the well-

to-do Africans tend to reject non democratic alternatives more than the poor and also tend to

support democracy and multi-party competition more than the poorest citizens. The poor,

meanwhile, perceive less supply of democracy than the richer folk. Interestingly though, using the

larger dataset, we find, just like Bratton, that Africa’s poor tend to vote more in elections than the

non-poor citizens.

The results lead us to similar conclusions as those adduced by Bratton (2005, 2008), namely that

poverty does not pose a threat to African democracies any more than wealthy: while the tendency of

poor people to demand less democracy might encourage ruling elites not to adopt stronger

3 The Afrobarometer is a comparative series of public attitudes on democracy, market reforms and civil

society in Africa. The first Round of surveys took place between July 1999 and October 2001, and covered more than 21,000 adult citizens in 12 countries. The second round of surveys reached more than 23,000 persons in 16 countries between June 2002 and October 2003. Rounds three and four took place in 2005 and 2008-09 respectively, covering a total of 18 and 20 countries. The total N in R3 and R4 was about 25,000 and 27,000 respectively. In each round, trained fieldworkers conducted face-to-face interviews in a local language of the respondent’s choice.

5

democratic reforms, the higher levels of dissatisfaction among the majority poor with the democracy

on offer might serve as a latent, if dormant, force against regression to authoritarianism. In the short

to medium term, the demand for the supply of better democracy might at the minimum pressure

ruling elites against abandoning the democratic experiments that were ushered in through the efforts

of the poor in the early 1990s. In a nutshell, the paradoxical democratic attributes of Africa’s poor

might actually help to explain the hybridization of political regimes on the continent in ways that

prevent consolidation towards democracy or regression towards authoritarianism.

2.0 How is the experience of poverty impacting on demand and supply of democracy in

Africa?

Over the years, the proposition that democracies will more likely exist and survive in richer countries

has been substantiated by analyses of cross-country data that have shown democracy to be more

prevalent and more stable in countries that have higher per capita incomes (Diamond, 1992; Lipset,

1994; Przeworski et al., 1996; 2000; Posner, 1997; Sklar and Whitaker, 1997; Krishna, 2008). In one

such recent study, Przerworski et al. (2000:273) have gone further to place a monetary premium on

the prospects of democracy’s endurance, arguing that the probability of the survival of democracy in

any country are greater when the country is richer, and that prospects of survival “increase steeply

and monotonically as per capita incomes get larger. Indeed, democracy is almost certain to survive in

countries with per capita incomes above $4,000.”

While there is very little agreement as to why higher per capita incomes should increase the

prospects of democracy survival, one of the main arguments is that poor people are more

preoccupied with survival and therefore have very little interest in taking part in democratic political

activities. Others meanwhile contend that democracy is a luxury good, for which poor people have

less need (Przerworski et al. (1996). The poor, in other words, are more willing to trade off

democracy in favour of immediate consumption. Yet another school of thought contends that

democracy has an intrinsic value which is increasingly sought after as populations become better off

(Helliwell, 1994). We should thus expect support for democracy to rise as levels of poverty decline.

However, the majority of the studies that have informed the argument that poor people make poor

democrats have tended to draw the conclusions by comparing predominantly richer industrial, and

largely democratic, countries against the largely authoritarian regimes of the developing world prior

to the onset of the democracy third wave. The assumption underpinning these studies being that

6

aggregate poverty at the national level is what drove the adoption of authoritarian rule, even if a

number of these countries had never experienced any form of democracy in the first place. These

arguments also tend to overlook the fact that the protests that ushered in democratic rule in the

developing world in the early to mid-1990s were largely driven by the very poor people, who bore

the blunt of the failed authoritarian ‘experiments’ which had been justified as offering the best

prospects for democracy (Sandbrook, 1996; Joseph, 1997; Bratton and van de Walle, 1997). Similar

lessons can also be taken from the recent protests that have precipitated the Arab Spring in recent

years (Anderson, 2011). As with the democratic protests of the 1990s in Africa, the protests that

precipitated the Arab Spring were led by the poor, suggesting that the existence of authoritarianism

does not equate to its acceptance and approval by the poor.

Despite evidence showing that the majority of Africa’s new democracies have not regressed back to

fully-fledged authoritarianism, there is still doubt as to whether the poor majority on the continent

will continue to support the new regimes (Sandbrook, 1996; Magnusson and Clark, 2005; Lakner

and van de Walle, 2009). Bratton and Mattes (2009), meanwhile, have offered the most

comprehensive set of evidence to date that supports the view that a high proportion of the post

authoritarian regimes in Africa are consolidating, not as democracies, but rather, as hybrid regimes.

This is borne out for their findings which appear to show that low demand and low supply of

democracy are reaching equilibrium levels that suggest consolidation of the new hybrid regimes

instead of progressing towards fully fledged democracies. These new trends necessitate once more a

re-examination of the poverty-democracy debate and to re-examine whether poverty is factor.

3.0 Data

This paper draws upon Afrobarometer survey data collected in 2002, 2005 and 2008. The 2002

survey covered a total of 16 countries4 and had an N of just over 23,000 respondents. The 2005

surveys meanwhile added two more countries and had an N of slightly over 25,000 respondents. In

the 2008/09 surveys, an additional two countries were added to bring the total number to 20, with

an N of just under 28,000 respondents. In all surveys, respondents were selected using a random,

4 The 2002 list of countries included: Botswana, Cape Verde, Ghana, Kenya, Lesotho, Malawi, Mali, Mozambique,

Namibia, Nigeria, Senegal, South Africa, Tanzania, Uganda, Zambia and Zimbabwe. The two additions in 2005

were Benin and Madagascar while Burkina Faso and Liberia were added in 2008.

7

stratified, multistage, national probability sample representing adult citizens aged 18 years or older.

The sample size in each country ranged from 1,200 to 2,400.

3.1 Measuring poverty: The Lived Poverty Index

The conceptualization of poverty presents definitional and measurement challenges, especially when

comparing it across time and space (Hagenaars, 1988; Bourguignon and Chakravarty; 2003; Clark

and Hume, 2005, Bratton, 2005). While many authors have insisted on the necessity of defining

poverty as a multidimensional concept, very little has actually been done to include the various

dimensions of deprivation into the practical definition and measurement of the term. This is in large

part due to the absence of people-centric measurements of deprivation, as a result of which

traditional indicators of poverty have tended to rely on aggregate measurements such as per capita

income and proportions of people living below poverty lines that are often arbitrarily conceived.

The increasing availability of national survey data that examines the experiences of deprivation

among respondents makes it possible to fully capture the multidimensional nature of poverty in

ways that were not previously possible. As such, survey data serves a very important tool to measure

the experiential dynamics of poverty from the perspectives of the poor themselves instead of relying

solely on the perceptions of experts. Within the African context, the Afrobarometer surveys have,

included questions that provide both objective indicators of absolute poverty that measure the

standard of living people actually enjoy. The surveys have also included subjective measures of

poverty that focus on people’s access to, or lack thereof, different kinds of resources.

Specifically, survey respondents have been asked to indicate whether over the previous year they

themselves, or anyone in their family has gone without: (a) Enough food to eat; (b) enough clean

water for home use; (c) medicines or medical treatment; (d) Enough fuel to cook your food; and (e)

a cash income. A range of ordinal level responses scales are then offered with the options: “never,”

“just Once or twice,” “several times,” “many times,” or “always”.

Because the existential reports refer to both income flows and other felt needs, the Afrobarometer

data has made it possible to develop a hybrid Index of Lived Poverty. This Index starts by identifying

citizens who can be considered as non-poor, that is those who during the year prior to the survey,

never went without food, clean water, medicine, cooking fuel or cash income. The second category

is of the moderate poor, namely those individuals who report occasionally going without a number

8

of basic necessities. The third category is of the very poor, that is individuals who report

experiencing several shortages in the previous year. The fourth category is of the extremely poor –

that is individuals who suffer regular and persistent shortages across the various basic needs. Table 1

provides a summary of the proportions of individuals in each poverty category in the three survey

rounds that form the basis of this paper:

Table 1: Classifications of African survey respondents by poverty category, 2002-2008/09

Index range Circa 2002

(16 countries)

Circa 2005

(18 countries)

Circa 2008

(20 countries)

Non-Poor 0 12 13 13

Occasional Poor 0.2-0.8 27 26 34

Very Poor 2.0-2.8 36 35 39

Extremely Poor 3.0-4.0 25 27 14

Source: Afrobarometer, R 2, 3 and 4.

It is worth noting that while the proportions of the people in the top two poverty categories were

similar in Rounds 2 and 3, there is a marked difference in the proportions in R4, which records a

significant reduction in the proportion of people in the extreme poverty category by almost half.

Overall, the proportion of people living in extreme poverty had declined from a high of 27% in R3

to 14% in the 2008/2009 surveys. Meanwhile, the proportion of people experiencing occasional

poverty had increased by 7 percentage points between the 2002 and 2008 surveys while that of the

non-poor has remained the same. The same figures also obtain when the data for the 4 countries

that were added between 2002 and 2008 are excluded from the analysis. Excluding the four

countries added between the 2002 survey round and the 20085, the same results obtain, with all but

two (Namibia and Senegal) of the sixteen countries in the Round 2 surveys recording declines in the

proportions of people classified in the extreme poverty category. Among the sixteen countries that

were covered in the 2002 and 2008 surveys, the proportion of people falling in the extreme poverty

category declined by 11 percentage points.

In the analysis that follows, we compressed the poverty categories into three: the non-poor and the

occasional poor in one group; the poor in one group and an extremely poor category combining

5 The four countries added between 2002 and 2008 are: Benin, Burkina Faso, Liberia and Madagascar

9

those who experience frequent and continuous poverty. This was done to make the analysis,

especially presentation of graphs and tables, less cluttered.

4.1 Poverty and democratic attitudes in Africa

To gauge the effects of poverty on the democratic attitudes, we first examine how poverty relates to

influences attitudinal indicators that measure demand for democracy. These look at data on support

democracy as the best system of government and those who explicitly reject authoritarian

alternatives (military rule, one party rule and strongman rule).

4.1.1 Poverty and Support for democracy

Overall, the evidence appears to support the view that the poor tend to hold less democratic

attitudes compared to their lesser poor counterparts:

Figure 1: Support for Democracy: Proportion of those saying democracy is preferable by poverty category (2002-2008)

Question wording: Which of these three statements is closest to your own opinion? Statement 1: Democracy is preferable to any other kind of government. Statement 2: In some circumstances, a non-democratic government can be preferable. Statement 3: For someone like me, it doesn’t matter what kind of government we have. (Figures are for those who indicate support for democracy)

10

It is clear from figure 1 that the more severe the levels of poverty, the less the support for

democracy. In the three survey rounds, support for democracy has remained highest among the least

poor citizens, followed by the moderately poor. The extremely poor, on the other hand, have

consistently remained the least supportive of democracy over time. However, it is interesting to

observe that between 2002 and 2008; support for democracy has gone up the highest among

Africa’s extremely poor, increasing by 7 percentage points between 2002 and 2008. This might

suggest that even after some citizens have graduated from severe forms of poverty, those left behind

have gained more faith, rather than lost it, in democracy. This is in sharp contrast to the predictions

of doom that suggested that democracy in Africa would come under increasing threat from the

poor. What this suggests is that rather than consider the poor as less supportive of democracy, it

perhaps takes more time for them to warm up to the concept. Thus rather than read the low levels

of support for democracy among the poor when it was a relatively new concept, the evidence

suggests that over time, support might actually increase rather than decrease. Thus the high numbers

of the poor might conceivably provide the strongest basis for democratic support rather than

undermine it.

4.1.2 Rejection of non-democratic alternatives

We probe further the relationship between poverty and democratic attitudes by examining whether

poor people are willing to countenance alternative authoritarian regimes.

4.1.2.1 Poverty and the rejection of one party rule

Figure 2 compares the proportions across the four poverty categories who express disapproval of

one party rule in the 20 countries. Consistent with the results from the responses on support for

democracy, higher proportions of the non-poor and the less poor reject one party rule compared to

the moderate and extremely poor.

11

Fig. 2: Proportion of those saying they disapprove or strongly disapprove one party rule by

poverty category, 2002-2008

Question wording: There are many ways to govern a country. Would you disapprove or approve of the following

alternatives: Only one political party is allowed to stand for election and hold office? Figures are for respondents

who say “strongly disapprove” or “disapprove”

In all the three surveys, the less-poor express the highest disapproval levels for one party rule while

the poorest reject one-party rule the least. Only in two countries, Benin and Tanzania do more

extremely poor reject one party rule than the non-poor., with 11 and 6 percentage point differentials

between the two categories. However, it is worth pointing out that the low levels of support for

democracy among the poor in the three survey rounds do not necessarily imply acceptance of this

form of government either. Instead, poorer citizens often express a lack of knowledge than the

lesser poor. Additionally, there has also been a large increase among the poorest categories in

expressing support for multi-party competition, which has jumped from 52 per cent in 2002 to 64%

in 2008. This is despite pre-transition rhetoric by ruling elites claiming that multi-party competition

is ill-suited to Africa’s multi-ethnic societies and would increase prospects of civil war (Jackson and

Rosberg, 1982; Nyong'o, 1992). While such rhetoric might have led a number of the poor to be

cautious in their embrace of multi-party politics in the years immediately following the transitions

from authoritarian rule, the passage of time has made the poor to embrace democratic competition

as they become more familiar with the new system. .

12

4.1.2.2 Rejection of military rule (2002-2008)

Consistent with the previous findings, the poorest citizens reject military rule less than the less poor.

Indeed, while the rejection of military rule among the non-poor and moderately poor has remained

almost the same between 2002 and 2008, the proportion of the extremely poor rejecting military rule

has actually fallen by six percentage points (Figure 3).

Fig. 3: Proportion of respondents rejecting military rule by poverty category, 2002-2008

Question wording: There are many ways to govern a country. Would you disapprove or approve of the following

alternatives: The army comes in to govern the country? Figures are for respondents who say “strongly disapprove”

or “disapprove”

An analysis of the data at the country level however reveals some interesting insights that provide

useful insights into the poverty and democracy debate. In Zimbabwe, which in recent years has seen

an increasing militarization of the political space, rejection of military rule among has actually gone

up among the extremely poor citizenry, but has declined among the non-poor6. Contrary to the

sweeping generalization that the poor are more susceptible to authoritarian modes of governance,

the Zimbabwe case suggests that where the poor face the blunt of the militarization of politics the

most, they will actually reject authoritarianism more than their non-poor counterparts. This is

6 Among the non-poor and modertaely poor Zimbabweans, for example, rejection of military rule has

declined by 8 and 4 percetage points repesctively between 2002 and 2008, while going up by 3 percentage points among the extremely poor.

13

further supported by evidence which shows that rejection of military rule among the extremely poor

citizenry has gone up in seven of 16 countries while remaining static in an additional one country.7

There is nonetheless evidence that suggests that some of the rhetoric of authoritarianism in a

number of countries on the continent is having a negative effect on the rejection of military rule.

This is particularly true in a number of countries which Bratton and Mattes (2009) have identified as

some of the declining regimes in Africa. In Senegal and Lesotho for example, rejection of military

rule has remained steady and very high among the less poor but have declined among the moderate

and extreme poor groups. In Mali, which recently experienced a coup d’état, there were dramatic

declines in the rejection of military rule between 2002 and 2008, with the highest decline among the

extremely poor. What unites these countries however, is the very high numbers of people in the

moderate and extremely poor categories. Taken together, the mixed evidence adduced here suggests,

first, that the poor are not all commonly susceptible to the rhetoric of demagogues. Rather, when

the beneficiaries of state patronage are predominantly among the ruling elites, as has been the case

of Zimbabwe, the poor actually look at democracy as a system of government best placed to take

them out of poverty. Although such countries might be classified as declining regimes, this might

only be from the perspective of the supply of democracy rather than its demand, especially among

the poor.

4.1.2.3 Rejection of one man rule

The third type of non-democratic regime that Afrobarometer survey respondents are asked to rate is

on man rule. In the three survey rounds, all three poverty categories have exhibited high rejection

rates for this form of government. Although the poor are reject one-man rule the least, the

differences in all three survey rounds are very marginal and not statistically significant. This leads us

to conclude that it is possible that what is driving the low levels of demand for democracy among

the poor is not because they are less democratic than the non-poor, but rather, it might be because a

high proportion of them might not have had the actual experience of some of the non-democratic

regimes with which to make a comparison. Since many of the Afrobarometer countries experienced

at last one-man rule, it is possible for the poor to compare that type of government and opine that

democracy is a better system of government based on their actual lived experiences. It is thus

possible that the primary factor that explains the low levels of demand for democracy among the

7 Viz;: Mozambique; Cape Verde ; Zimbabwe; Tanzania ;Nigeria ;Namibia ; Botswana and Kenya

14

poor have very little to do with poverty itself, but rather, a lack of informed knowledge. Where

experience substitutes learned knowledge, the poor can demand democracy as much as their non-

poor counterparts. The case of Zimbabwe once more provides a very good illustration, as the poor,

who have suffered the most under Robert Mugabe’s autocratic rule, have been the most consistent

and strongest rejecters of consistently rejected one-man rule the strongest.

In sum, while the poorer African citizens exhibit low demand for democracy than the their less poor

counterparts, the evidence is still somewhat positive given that here is very little evidence to suggest

that the poor are actually gaining more faith in democracy nor are they warming up to non-

democratic alternatives.

4.2 Poverty and the Supply of democracy

To measure supply for democracy, we use Bratton and Mattes’s measures which combine those who

say that they think their country is a democracy with those who say they are either ‘fairly’ or very

satisfied with the way democracy works in their country.

4.2.1 Poverty and perceptions of the extent of democracy

The first measure of the supply of democracy draws from the question that asks respondents to

place their country on a continuum of democracy ranging from a full democracy to not a

democracy. Figure 4 provides a summary of the perceptions of democracy in the three well-being

categories:

15

Figure 4: Extent of democracy by poverty category (per cent saying country is a full

democracy or a democracy with minor problems, 2002-2008)

Question wording: “In your opinion how much of a democracy is [name of country] today? Figures are for

respondents who say country is “A democracy, but with minor problems,” or “full democracy”.

The non-poor, yet again, are more likely than the poorest citizens to rank their countries as full

democracies or democracies with minor problems. On one extreme, less than half of the extremely

poor have ever considered their countries as full democracies or democracies with minor problems.

In order to address the perception that the poor do not understand the meaning of democracy and

are therefore more likely to consider their countries as less democratic than they actually are, we

examined in greater detail the case of Zimbabwe, whose democratic scores have declined over the

last decade under President Robert Mugabe’s authoritarian rule. The results even in this case suggest

that the poorer citizens are more likely to judge their country as less democratic compared to their

non-poor counterparts (Figure 5)

16

Fig. 5: Extent of Democracy in Zimbabwe by poverty category (proportion saying

Zimbabwe is either a full democracy or a democracy with minor problems)

In all the three survey years, Zimbabwe was rated “free” by Freedom House, with combined scores

of 6 (2002); 6.5 (2005) and again 6.5 in 2009. Yet, while very few among the poor have rated their

country as democratic, a higher proportion of the non-poor, albeit less than half, have been rating

Zimbabwe as democratic. The Zimbabwe case shows, therefore, that being poor is not necessarily

always a predictor of a citizenship that will inflate the perceptions about their country’s democratic

credentials. Thus the evidence from the last two figures (Fig 4 and Fig 5) should be taken to imply

that the poor might actually apply more stringent measures of democracy than the non-poor.

Because the poor are less beholden to incumbent governments, they might apply tougher standards

to evaluate the quality of democracy compared to the non-poor. This is also borne out by the fact

that when presented with the statement about an individual who “lives in a country with regular

elections; that has one big political party and many small ones; where people are afraid to express

political opinions or to vote for the opposition; and where the opposition is so weak that it seems

that it can never win an election,” a slightly higher proportion (15%) of the non-poor in the 20

countries surveyed in 2008 described this country as a democracy while only 12 per cent of the

extremely poor considered them so. The non-poor, in other words, are more likely to inflate the

extent of democracy in their countries than the poorest citizens, thus contributing more to the

perception that more democracy is being supplied when in actual fact, it is not.

17

4.2.2 Poverty and satisfaction with democracy

How do poor people view democracy? The Afrobarometer surveys ask respondents to indicate their

levels of satisfaction in the way democracy works in their country. As with the previous dimensions

of democracy demand, the poor generally express low levels of satisfaction in the working of

democracy than the non-poor (Fig. 6)

Fig. 6: Poverty and satisfaction with democracy (2002-2008)

Question wording: Overall, how satisfied are you with the way democracy works in [name of country]? The figures

are the proportions of those who say they are either fairy satisfied” or “very satisfied”

Consistent with our previous findings, the poorest citizens express less satisfaction with democracy

compared to the non-poor. When taken together, the findings on the perceived extent of democracy

and satisfaction in it, it becomes clear that contrary to the school of thought that perceives Africa’s

poor as less democratic, they actually have higher standards in defining democracy and it takes more

to get them satisfied. This aspect is further reflected in the perception of the freeness and fairness of

elections in the Afrobarometer countries. Examining the characterization of the previous elections,

we find, once more, that the poorest citizens are less likely than the less-poor folk to describe the

previous elections as either “completely free and fair” or free and fair but with minor problems”

(Table 2)

18

Table 2: Proportion of respondents describing previous elections as free and fair by well-

being category (2005 and 2008)8

Non Poor/occasional

Poor

Poor Very poor/

Extremely Poor

circa 2005 (18

countries)

71 65 57

circa 2008 (20

countries)

66 62 56

Question wording: “On the whole, how would you rate the freeness and fairness of the last national elections, held in

[election year]?” Figures are for respondents who say it was “completely free and fair” or “Free and fair, but with

minor problems”

We can see yet again from the foregoing that the poor not only have higher expectations of

democracy, but they also employ higher standards in judging the freeness and fairness of elections.

Taken together, these findings suggest that the future of democracy in Africa, particularly its supply,

might be less threatened by the poor, but rather the non-poor. The poor have higher requirements

for describing their countries as full democracies; are more difficult to satisfy with the brand of

democracy on offer and are less satisfied with the quality of elections. By contrast, higher

proportions of the non-poor have rather lax standards in perceiving the extent of democracy; are

also easily satisfied with poor quality democracy and are quick to describe elections as free and fair.

In the end then, the future of democracy in Africa, might rest on the poor, who will demand better

quality democracy and press for more democratic reforms than the contend non-poor. Indeed,

anecdotal evidence already seems to suggest this is the trajectory on the continent, as a number of

the recent protests for democratic reforms in countries such as Malawi, Zimbabwe, Kenya, Senegal,

Uganda, Cote D’Ivoire, to mention but a few countries, much as in the Arab spring, have originated

from the disconcerted poor while the non-poor have largely fought for the retention of the status

quo.

In order to begin a more robust test of the relationship between poverty and democracy, we

undertook some bivariate correlations to test the direction and significance of the relationship that

poverty has with a number of the variables that measure demand and supply of democracy in Africa.

The findings are presented in Table 3:

8 This question was not asked in all countries in 2002

19

Table 3: The effects of poverty on demand and supply of democracy (20 countries, 2008)

Dependent variable Bivariate correlation9 (Pearson’s r)2

Poor are Less likely to support democracy (2008) -.042***

Less likely to reject one-party rule -.043***

Less likely to Reject military rule -.038***

Less likely to Reject one-man rule -.051***

thus

Less likely to

Demand democracy

Less likely to be satisfied with the way democracy works

-.032***

Less likely To perceive extensive democracy -.044***

Thus

Less likely

To perceive supply of democracy

Based on a four point scale from extremely poor to non-poor (***p=<.001). Question wordings for dependent variables given in text.

The findings in Table 3 mirror those of Bratton (2005), both in terms of demand and supply of

democracy. However, the weaker correlates in 2008 compared to the 2005 data on the supply of

democracy variables, suggest that the significance of poverty in reducing the perceptions on the

supply of democracy are weakening. This further underlines our argument that with the passage of

time, poor people’s understanding of democracy is increasing, thereby contributing to their

increased understanding of the concept and thus reducing the gap between the poor and the non-

poor.

4.3 Poverty and Political behaviour

In his analysis of poverty and democracy using R2 survey data, Bratton (2006 and 2008) has adduced

evidence that shows that while the political attitudes of the poor towards the supply and demand of

democracy are somewhat consistent, there is less consistency in poverty’s relationship with mass

action. We did a re-run of the same tests that Bratton undertook in 2005 using the survey data from

the 20 countries in R5. Our findings brought us to the same conclusions as Bratton’s.

9 The correlations are calculated using the R4 data

20

4.3.1 Poverty and Voting behaviour

The first major finding relates to voting trends. Across the 16 and 20 countries that were surveyed in

2005 and 2008/09, members of the extremely poor majority were somewhat, and significantly, more

likely to report having voted in the previous elections than their wealthier counterparts. In 2005, the

extremely poor were four percentage points more likely to have voted than the well-to do. In 2008,

when there were increases in the proportions of people reporting to have voted across all three

poverty categories, the extremely poor yet again reported having voted more than the non-poor by a

margin of 3 percentage points (Fig.7)

Fig. 7: Poverty and voting in elections (2005 and 2008)

.

Question wording: With regard to the [year] national elections, which statement is true for you? Figures are for

those who say they voted in the last elections

The findings on voting patterns are consistent with those adduced by Bratton in his earlier study.

The findings on the relationship between poverty and voting from the 2005 and 2008

Afrobarometer surveys contradict the conventional views about voting. While the conventional

literature has argued that poverty and low socio-economic status would depress political

participation, the Afrobarometer evidence suggests that the opposite is true in Africa. Here, rather

than depressing voter turn-out, poverty actually has a catalysing effect. Despite having low regards

about the nature of their country’s democracies, and expressing low satisfaction with the quality of

democracy on offer, Africa’s poor are showing their faith in democracy by turning out to vote in

21

larger numbers than the non-poor. That this is the case is even more surprising, considering the

findings that demonstrate that the poor in Africa demand less democracy on average compared to

the richer folk.

4.3.1 Poverty, partisanship and collective action

An interesting dimension that requires further analysis is whether the poor have a greater proclivity

to support the ruling party and thus contribute to the limited numbers of turnover on the continent.

The evidence however suggests that this is not the case. While a majority of respondents indicate

that they would vote for the ruling party in elections, the proportions are evenly split between the

poor and non-poor alike. A closer analysis of the national level data however reveals that in the only

two countries where elections have resulted in partisan turnover, namely, Ghana (2007) and Senegal

(2012), large majorities of respondents indicated that they would vote for the opposition party if an

election were held the next day. However, in both cases, there were no noticeable differences in the

voting intentions of the poor and non-poor.

The evidence on the influence of poverty on collective action in between elections is, however,

rather mixed. On one hand, the findings show that the poor are less likely than the well-off to take

political action by attending community meetings; joining others to raise an issue or attending a

demonstration or a protest march. This is somewhat surprising, especially considering the earlier

findings suggesting that the poor are less than pleased with the quality of democracy on offer.

Despite having such misgivings on the quality of democracy, it would appear Africa’s poor are

loathe to take action, either singularly or by joining others other than casting their vote in elections.

These findings are somewhat consistent with the conventional wisdom about the high opportunity

costs that democratic participation imposes especially on the poor. Collective political action, in

other words, is a luxury for Africa’s poor, a high proportion of whom also express fear of political

intimidation and violence more than the non-poor.

When it comes to contacting public officials, the poor tend to hold off from contacting formal

government officials such as Members of parliament, and elected councillors. The poor are also less

likely to contact religious leaders. However, reflecting perhaps the lingering importance of traditional

leadership in Africa’s settings, the poor are actually more likely than the well off to contact

traditional leaders. The failure of the poor to contact their elected representatives, including both

22

councillors and Parliamentarians in spite of the evidence suggesting the poor tend to vote more in

elections, suggests a rather problematic relationship between representatives and the represented. In

polities such as these where the poor form a clear majority, the absence of contacts with elected

representatives means that the interests of the poor are less likely to be reflected in the deliberations

in official public deliberations. It is perhaps not surprising that the poor are less supportive of

democracy given that official government decision making bodies are not likely to advocate their

interests in the absence of contacts between the official representatives and the poor. The limited

contacts between the poor and elected representatives can also suggest that the availability of

informal local institutions, such as traditional leaders and other local notables, makes the poor less

likely to use formal institutions to communicate their needs and wishes. Of course this might imply

that formal officials are not as readily accessible to the poor. While informal institutions offer a more

accessible alternative, the limited resources and power at these levels means that the poor are further

marginalized from the parts of the state where real decisions are made.

Table 4: The effects of poverty on demand for democracy (20 countries, 2008)

Dependent variable Bivariate correlation10 (Pearson’s r)2

Poor are More likely To have voted in elections (2008) +.041***

No more likely To vote for the ruling party -.003

Less likely to Attend community meetings -.105***

Less likely To join others to raise an issue -.085***

less likely To attend a demonstration or protest march

-.030***

Less likely To contact MP -.011*

less likely To contact local councillors -.029***

Less likely To contact religious leaders -.054***

more likely To contact traditional leaders +.013*

More likely to contact formal leader alone or with a group

+.024***

Based on a four point scale from extremely poor to non-poor (***p=<.001). Question wordings for dependent variables given in text.

10

The correlations are calculated using the R4 data

23

5.0 Discussion: Lived poverty and the fate of African democracies

In order to get a fuller understanding of the effects of poverty on democratic attributes, we used

multivariate regression analysis to test whether poverty’s political effects on a number of political

values and behaviour. The findings once more mirror those adduced by Bratton using data from an

earlier Afrobarometer survey round, with the evidence suggesting that many of poverty’s effects are

conditional on other aspects of social structure.

5.1 Conditional effects of lived poverty on support for, and satisfaction with, democracy

When controlling for other social factors such as age, gender, education, employment and rural

residency, the effect of lived poverty only retains its significance in impacting support for

democracy. However, satisfaction with democracy is largely driven by the levels of education while

lived poverty loses its significance (Table 5)

Table 5: the conditional effects of poverty on support for, and satisfaction with, democracy (2008)

Support for

democracy

Satisfaction with

democracy

Support opposition

to examine

government

Lived poverty -.049*** -.005 -.008

Gender (female) .064*** .068*** .052***

Age .059*** .066*** .060***

Rural .071*** .096*** .052***

Education .040*** .080*** .002

Employment -.007 -.013* -.001

Cell entries are standardized coefficients (beta).

***p=<.001, **, p=<.01, *p=<.05

The fact that education is the major driver of both the support for, and satisfaction with

democracy, once more reinforces our earlier argument that it is not poverty that undermines

democracy, but rather, the absence of education, with the majority of the poor lacking it. Of

course while lack of education is itself an indicator of poverty, what this evidence suggests is

that if African countries can strive to provide more education to their peoples, even if they do not

24

make significant strides in addressing the incidence of deprivation, it is possible to build a

stronger foundation for democracy by inculcating democratic attitudes in their peoples.

5.2 Conditional effects of lived poverty on political behaviour

In order to understand fully the effects of lived poverty on political behaviour, we run a regression

analysis that controlled for the same social aspects we controlled in examining the effects of lived

poverty on political attitudes. The results are presented in table 6:

Table 6: the conditional effects of poverty on political behaviour (2008)

Voted in last

election

(2008)

Vote for

the ruling

party

Attend

community

meeting

Attend

demonstration

or protest march

Contact

Traditional

leader

Lived poverty .046*** .002 -.088*** -.043*** .029***

Gender (female) .014* .078*** -.095*** -.022*** -.059***

Age .007 .012* .019** .008 -.070

Rural -.022*** .044*** .143*** .006 .016*

Education .014** -.002 .025*** .005 -.018**

Employment .004 -.006 .021 .023*** -.017**

Cell entries are standardized coefficients (beta).

***p=<.001, **, p=<.01, *p=<.05

The effects of lived poverty remain significant with respect to four of five aspects of political

behaviour. The only non-significant effect is on voting for the ruling party. While lived poverty has a

negative effect on the two dimensions of democratic attitudes discussed earlier, it has a positive

effect on two dimensions of political behaviour, namely, turning out to vote and contacting

traditional leaders. It however has a negative effect on attendance at community meetings and

attendance of demonstrations or protests marches. Poverty, in other words, increases the propensity

to vote while undermining other forms of political action. Thus despite having played a critical role

in helping to introduce the political transitions of the 1990s, the poor have since retreated and are

not taking part in the forms of activisms that could help transform politics in ways that could

potentially lead to the improvements of their lives. While this might suggest that the poor are less

engaging in critical action that could make African democracies more responsive to their needs, it

should be borne in mind that generally, the poor are risk averse, fearing the consequences of

25

political activism more than the non-poor. But this does not mean that they would always remain

passive. As the experience of the early 1990s showed, the poor do not remain passive forever. When

pushed to the brink, they will react and press for political reforms. The fact that they are

demonstrating an unwillingness to engage in political activism might therefore be an indicator that

despite their reservations about the quality of democracy on offer, they have not been pushed to the

limit where they would take action. It might also of course be a reflection of their trust in the power

of the vote over protests, seeing as it is that they take part more in voting than the non-poor. If this

is the case, then the future of democracy can be considered to be on somewhat firmer ground.

6.0 Tentative Conclusions

We have offered some tentative insights into the effects of poverty on the fate of democracy in

Africa. Although more remains to be done, our findings presented from three Afrobarometer

rounds reaffirm those of Bratton (2005, 2008), which were based on one round of Afrobarometer

survey data conducted in 2002. The evidence adduced in our analysis demonstrates that the

experience of poverty in Africa has a negative effect on demand and supply of democracy as well as

on a number of attributes of political participation. The fact that the poor demand democracy less

than the well-to do while at the same time perceiving limited supply of democracy, in our view

explains why more African regimes are consolidating as hybrid regimes instead of moving towards

democratic consolidation. This is in part because the majority poor are not as demanding of greater

democracy even if they consider the current supply of democracy to be inadequate.

Another interesting finding from this study has been the re-affirmation of the view that poverty in

Africa has a positive effect on the propensity to vote. This finding is in sharp contrast to the

conventional wisdom that holds that the struggles for survival would result in the poor looking at

democratic participation as a luxury good. On the contrary, our findings suggest that Africa’s

majority poor, have placed much faith in the power of the vote as a means to uplifting their lives.

While the vote has so far been used very little to vote out under-performing regimes, it is

encouraging nonetheless to note that the poor are no more predisposed to vote for the ruling party

than the rich. The vote thus remains a potent tool for which the poor can remove governments that

completely ignore their interests. This bodes well for the future of democracy on the continent, even

if poverty does undermine other aspects of democracy.

26

This is still a work in progress. We look forward to your suggestions on how best to take this project

further and in ways that compliments, rather than simply reaffirm, the findings of similar studies in

the past.

Bibliography

(to be added)