election quality, public trust are central issues for …...copyright © afrobarometer 2016 1...
TRANSCRIPT
Copyright © Afrobarometer 2016 0
Election quality,
public trust are central issues
for Africa’s upcoming contests By Peter Penar, Rose Aiko, Thomas Bentley, and Kangwook Han
Afrobarometer Policy Paper No. 35 | September 2016
Copyright © Afrobarometer 2016 1
Introduction
Nothing kindles democracy’s energies, anxieties, hopes, and frustrations like an election. The
quality of an election can spell the difference between a cooking fire and an explosion.
If a successful election can calm and focus a nation (e.g. Namibia 2015), a disputed election
can tear it apart (e.g. Burundi 2015, Côte d'Ivoire 2010, Kenya 2008).
With at least 25 African countries conducting national elections in 2016-2017,1 great attention
is focused on electoral management bodies – typically national electoral commissions – as
crucial players in electoral processes and in shaping public perceptions of how well
democracy is working. Poor electoral management can enable election fraud and, even if it
doesn’t swing an election, produce political alienation, public mistrust, protest, and violence.
In 2016, we have already seen examples of unrest in Kenya, where opposition calls for
electoral commission reforms using the hashtag #IEBCMustFall have sparked demonstrations
and a violent reaction from security forces; in the Republic of the Congo, where election
malpractices led to violent protests; and in Gabon, where bloody clashes erupted after
President Ali Bongo claimed a widely disputed re-election victory. In Ghana, pre-election
anxieties are high amid questions about the electoral commission’s revision of the voter roll
for December’s election.
Against the backdrop of history’s examples – in Africa and elsewhere – of tampering with
voter rolls, suppression of competition and voter turnout, ballot stuffing, vote-buying, multiple
voting, and manipulation of results, free and fair elections, agreed to in the African Union’s
Charter on Good Governance and Elections, depend on competent election management
supported by citizen sensitization efforts to build public confidence.
Using 2014/2015 Afrobarometer data from 36 African countries, this analysis examines public
perceptions of electoral management institutions and the quality of elections. Overall, public
trust in national electoral commissions is moderate at best. Although a majority of citizens say
their most recent elections were mostly free and fair, citizens express serious concerns about
the fairness of vote counts, corruption during elections, and the safety of voters during
campaigns and at the polls. Citizens’ views of electoral commission performance and
election quality generally mirror the opinions of country experts found in international
assessments.
More broadly, many citizens say elections are not working well as mechanisms t o ensure that
people’s views are represented and that voters can hold non-performing leaders
accountable. Few countries have achieved improvement in the perceived performance of
elections over the past decade.
Afrobarometer survey
Afrobarometer is a pan-African, non-partisan research network that conducts public attitude
surveys on democracy, governance, economic conditions, and related issues across more
than 30 countries in Africa. Five rounds of surveys were conducted between 1999 and 2013,
and findings from Round 6 surveys (2014/2015) are currently being released. Afrobarometer
conducts face-to-face interviews in the language of the respondent’s choice with nationally
representative samples that yield country-level results with margins of error of +/-2% (for
samples of 2,400) or +/3% (for samples of 1,200) at a 95% confidence level.
Round 6 interviews with almost 54,000 citizens in 36 countries (see list in Appendix Table A.1)
represent the views of more than three-fourths of the continent’s population.
1 According to the National Democratic Institute, national elections (not counting constitutional referenda) in
2016 include: Benin, Cape Verde, Central African Republic, Chad, Comoros, Côte d'Ivoire, Democratic Republic
of the Congo, Djibouti, Equatorial Guinea, Gabon, The Gambia, Ghana, Morocco, Niger, São Tomé & Príncipe, Seychelles, Somalia, Uganda, and Zambia. In 2017: Algeria, Kenya, Liberia, Rwanda, Senegal, Sierra Leone. See https://www.ndi.org/electionscalendar/.
Copyright © Afrobarometer 2016 2
Key findings
On average across 36 surveyed countries, just half (50%) of respondents say they trust
their electoral commission “somewhat” (25%) or “a lot” (25%). Some of the lowest
levels of trust are expressed in countries with closely contested elections in 2016
including Gabon (25%), São Tomé and Príncipe (31%), and Ghana (37%).
Two-thirds of Africans rate their most recent election as “completely free and fair”
(41%) or “free and fair, but with minor problems” (24%). Citizens’ views on election
quality are generally consistent with assessments by international experts.
Copyright © Afrobarometer 2016 3
But substantial proportions of the population are skeptical about the quality of their
elections. More than four in 10 Africans say that voters are at least “sometimes”
threatened with violence at the polls (44%); that opposition parties and candidates
are at least “sometimes” prevented from running (43%); that the news media “never”
or only “sometimes” provides fair coverage of all candidates (43%); and that voters
are “often” or “always” bribed (43%).
Only one-third (34%) of Africans think that votes are “always” counted fairly. Several
countries with a history of election-related violence express low levels of confidence
in the vote count, including Kenya (26%), Zimbabwe (22%), and Nigeria (6%).
Fully half of Africans say that elections do not work well as mechanisms to ensure that
people’s views are represented (50%) or that voters are able to remove non -
performing leaders from office (51%). Countries with the highest levels of
dissatisfaction with the representation and accountability performance of elections
include Gabon, Morocco, Sudan, Nigeria, Swaziland, and Madagascar.
Trust in national electoral commissions
Election management in Africa is the focus of considerable attention from the United Nations,
African regional organisations, and international entities such as the Electoral Institute for
Sustainable Democracy in Africa (EISA), the International Foundation for Electoral Systems
(IFES), the International Republican Institute (IRI), and the National Democratic Institute (NDI).
At the national level, electoral management bodies are engaged in ongoing review and
reform efforts to improve logistical management and transparency. Countries are auditing
their voter rolls, implementing biometric voter registration and identification, engaging
political stakeholders to monitor management procedures, and improving processes for
transmitting results. Some of the most visible reforms have centered on electoral commissions
themselves, with a view to improving performance on one important measure of election
management: public trust in electoral management bodies.
Across 35 African countries, just half of citizens say they trust their national electoral
commission “somewhat” (25%) or “a lot” (25%) (Figure 1). One in five respondents (21%) say
they have no trust whatsoever in their election management body. On a scale of 1 (not at
all) to 4 (a lot), the average extent of trust in electoral commissions in Africa is 2.6, which lies
between “just a little” and “somewhat.”
Three countries with elections this year express some of the lowest levels of trust: Gabon (25%
trust “somewhat” or ”a lot”), São Tomé and Príncipe (31%), and Ghana (37%). Other
countries showing remarkably low levels of trust in their electoral commissions include Sudan
(28%), Nigeria (31%), Algeria (31%), Liberia (33%), Togo (37%), and Sierra Leone (37%). At the
other extreme are Burundi (surveyed in September-October 2014), Niger, and Namibia (each
74%).
With regard to elections in 2017 and 2018, there are warning signs for Kenya and Zimbabwe,
as less than half (46%) of citizens trust their electoral commission in these countries, which
have experienced high levels of election-related violence in the past.
When considering public trust levels by country, it is important to note the timing of the
Afrobarometer Round 6 fieldwork and its proximity to elections (see Appendix Table A.1 for
fieldwork and election dates). In Sudan, for example, where recorded trust levels are low, the
survey was fielded shortly after the April 2015 elections, which were so flawed that the
African Union considered not sending observers and all major opposition parties boycotted
the polls. In Nigeria, the survey was conducted three months prior to the March 2015
elections, which were the object of widespread anxieties but turned out to be largely
peaceful and successful. In Namibia, the survey was completed during the run-up to the
November 2014 elections, a period marked by optimism about the implementation of
electronic voting machines. These particulars of the timing of the survey relative to the
Copyright © Afrobarometer 2016 4
national election calendar can have some effect on the findings and should be kept in mind
when making cross-country comparisons.
Figure 1: Trust in national electoral commissions | 35 countries* | 2014/2015
Respondents were asked: How m uch do you trust each of the following, or haven’t you heard enough about them to say: National Electoral Commission? (* Question was not asked in Morocco.)
Average levels of public trust in electoral commissions have not changed dramatically over
time. Across 32 countries where this question was asked in both Round 5 and Round 6, trust
levels have remained essentially unchanged (50% “somewhat” or “a lot” in 2011/2013 and
51% in 2014/2015).
8%
10%
7%
16%
10%
13%
20%
18%
19%
17%
16%
25%
11%
22%
22%
31%
26%
34%
25%
27%
15%
30%
39%
21%
18%
38%
34%
31%
37%
32%
26%
21%
34%
38%
53%
49%
17%
18%
24%
15%
21%
20%
17%
19%
18%
24%
25%
17%
34%
24%
24%
17%
22%
14%
25%
27%
39%
25%
17%
35%
39%
19%
23%
28%
23%
29%
41%
46%
35%
36%
22%
25%
24%
25%
36%
40%
24%
38%
21%
22%
26%
29%
27%
27%
22%
22%
21%
28%
20%
20%
23%
22%
28%
19%
10%
20%
22%
21%
16%
23%
22%
20%
19%
20%
20%
18%
10%
11%
51%
39%
31%
18%
35%
28%
37%
37%
24%
27%
26%
30%
22%
27%
25%
19%
18%
28%
21%
19%
18%
18%
21%
12%
13%
17%
14%
16%
13%
15%
8%
8%
11%
7%
10%
9%
1%
7%
2%
10%
10%
2%
6%
3%
13%
4%
5%
2%
11%
5%
8%
5%
14%
4%
6%
6%
8%
13%
12%
8%
5%
13%
4%
5%
4%
5%
4%
1%
1%
6%
5%
0% 20% 40% 60% 80% 100%
Gabon
Sudan
Nigeria
São Tomé and Príncipe
Algeria
Liberia
Togo
Ghana
Sierra Leone
Benin
Cameroon
Côte d'Ivoire
Cape Verde
Kenya
Zimbabwe
Zambia
Mozambique
Guinea
Average (35 countries)
Uganda
Madagascar
Tunisia
Lesotho
Egypt
Swaziland
Malawi
Senegal
South Africa
Burkina Faso
Botswana
Tanzania
Mauritius
Mali
Namibia
Niger
Burundi
A lot Somewhat Just a little Not at all Don't know
Copyright © Afrobarometer 2016 5
Across 18 countries2 tracked since 2005/2006, trust shows a modest negative trend: Strong
positive sentiments (trust “a lot”) have declined slightly, replaced by modest positive
sentiments (trust “somewhat”), while a decrease in “don’t know” responses is matched by
small increases in “not at all” and “just a little” responses (Figure 2).
Figure 2: Change in public trust in national electoral commissions | 18 countries
| 2005-2015
Respondents were asked: How m uch do you trust each of the following, or haven’t you heard enough
about them to say: National Electoral Com m ission?
While average changes are small, there is significant over-time variation within countries in
public trust. Between 2011/2013 and 2014/2015, trust in electoral comm issions increased
sharply in Mali (by 28 percentage points), Egypt (22 points), and Namibia (17 points) while
dropping precipitously in Ghana (by 22 percentage points), Sierra Leone (19 points), Benin
(14 points), and South Africa (11 points) (Figure 3).
For some countries, increases in trust between 2011/2013 and 2014/2015 may reflect reforms
and technological advances. Namibia, for example, introduced electronic voting in its 2014
general elections. In Mali and Madagascar, electoral commissions may have garnered trust
from the way they managed elections that brought an end to political instability.
Interestingly, among those with significant recent decreases in trust are countries that have
high levels of electoral competition and are generally viewed as relatively democratic,
including Ghana, Benin, South Africa, Cape Verde, and Zambia. This may reflect citizens’
rising expectations of what constitutes a high-quality election; increasing oversight of
election processes, especially by opposition parties and civil society; and the view that even
minor irregularities in elections may influence the outcome in very tight contests.
2 Countries tracked since 2005/2006 are Benin, Botswana, Cape Verde, Ghana, Kenya, Lesotho, Madagascar,
Malawi, Mali, Mozambique, Namibia, Nigeria, Senegal, South Africa, Tanzania, Uganda, Zambia, Zimbabwe.
30% 26%
28%
26%
23% 24%
25%
27%
20% 20%
23% 22%
17%
21%
17%
20%
11% 9%
7% 6%
0%
10%
20%
30%
40%
2005/2006 2008/2009 2011/2013 2014/2015
A lot Somewhat Just a little Not at all Don't know
Copyright © Afrobarometer 2016 6
Figure 3: Recent changes in public trust in national electoral commissions | 32 countries* | 2011-2015
Figure shows increase or decrease, in percentage points, from 2011/2013 to 2014/2015 in the proportion
of respondents saying they trust their national electoral com mission “som ewhat” or “a lot.” (*Question not asked in Morocco, not asked in Tunisia in 2011/2013; Gabon and São Tom é and Príncipe were not
surveyed in 2011/2013.).
Examining longer-term trends among 18 countries tracked over the past decade, significant
increases in trust were recorded in Namibia (18 percentage points), Zimbabwe (17 points),
Mali (15 points), and Zambia (15 points, although in Zambia trust started declining again in
2011/2013) (Figure 4). But several countries registered alarming decreases in trust over the
past decade, led by Ghana (a drop of 38 percentage points), Mozambique (24 points), and
Tanzania (19 points) (Figure 5).
Moreover, trends over the past decade may obscure significant short -term ups and downs
within countries, such as dips in Senegal (2008) and Mali (2013) followed by increases. In
Kenya, an overall 7-percentage-point decline in trust over the past decade includes a
dramatic 27-point drop in the 2008 survey, reflecting the country’s violent presidential
election, followed by a full recovery in the 2011 survey. From 2011 to 2014, public trust in
Kenya’s electoral commission declined again (by 7 percentage points) – perhaps an
example of the fact that in some cases, efforts to improve election practices may fail to
-22
-19
-14
-11
-10
-10
-8
-7
-6
-6
-4
-3
-1
-1
0
1
2
3
3
3
4
5
8
9
9
10
10
11
12
17
22
28
-30 -20 -10 0 10 20 30 40
Ghana
Sierra Leone
Benin
South Africa
Cape Verde
Liberia
Zambia
Kenya
Malawi
Mozambique
Mauritius
Nigeria
Guinea
Burkina Faso
Botswana
Sudan
Algeria
Côte d'Ivoire
Togo
Cameroon
Lesotho
Tanzania
Senegal
Burundi
Swaziland
Niger
Zimbabwe
Uganda
Madagascar
Namibia
Egypt
Mali
Copyright © Afrobarometer 2016 7
meet expectations and may even erode public confidence as existing weaknesses are
revealed. Kenya implemented biometric voter registration and identification before its 2013
general elections, but procurement, management, and technical problems led many
Kenyans to question the integrity or competence of the electoral commission. A more recent
example comes from Ghana, where the electoral commission was directed by the judiciary
to conduct a full audit of the voter list. The revelation of many inaccuracies in the voter list
and questions about whether the electoral commission will have time to correct these
inaccuracies by December 2016 may have aggravated the decade-long downhill slide of
public trust in the commission.
Figure 4: Changes in public trust in national electoral commissions | countries with
largest increases | 2005-2015
Respondents were asked: How m uch do you trust each of the following, or haven’t you heard enough
about them to say: National Electoral Com m ission?
(% who say “som ewhat or “a lot”)
Figure 5: Changes in public trust in national electoral commissions | countries with
largest decreases | 2005-2015
Respondents were asked: How m uch do you trust each of the following, or haven’t you heard enough
about them to say: National Electoral Com m ission?
(% who say “som ewhat or “a lot”)
56% 63%
57%
74%
29% 31% 36%
46%
53% 53%
40%
68%
33% 41%
56% 48%
20% 28%
33% 31%
49%
34%
50%
58%
0%
20%
40%
60%
80%
100%
2005/2006 2008/2009 2011/2013 2014/2015
Namibia Zimbabwe Mali
Zambia Nigeria Senegal
75%
66%
59%
37%
72%
66% 54%
48%
86%
77% 62% 67%
67%
51%
56% 62%
41% 43%
53% 53%
25%
53% 46%
0%
20%
40%
60%
80%
100%
2005/2006 2008/2009 2011/2013 2014/2015
Ghana Mozambique Tanzania
Lesotho Uganda Kenya
Copyright © Afrobarometer 2016 8
Freeness and fairness of elections
Citizens’ assessments of the quality of their elections surpass their trust in their electoral
commission: Two-thirds (65%) say their most recent election was either “completely free and
fair” (41%) or “free and fair, but with minor problems” (24%) (Figure 6).
Figure 6: Freeness and fairness of elections | 36 countries | 2014/2015
Respondents were asked: On the whole, how would you rate the freeness and fairness of the last
national election, held in [20xx]?
This overall assessment has been fairly stable, both across 34 countries surveyed in 2011/2013
and across 18 countries tracked over the past decade (Figure 7),
Figure 7: Freeness and fairness of elections | 18 countries | 2005-2015
Respondents were asked: On the whole, how would you rate the freeness and fairness of the last national election, held in [20xx]?
(Note: This question was not asked in Afrobarom eter Round 4 (2008/2009).)
41%
24%
14%
11%
10%
Completely free and fair
Free and fair, but with
minor problems
Free and fair, with major
problems
Not free and fair
Don’t know
41%
47%
44%
24% 23% 24%
12% 11% 14%
10% 9%
9% 10%
0%
10%
20%
30%
40%
50%
2005/2006 2011/2013 2014/2015
Completely free and fair Free and fair, but with minor problems
Free and fair, with major problems Not free and fair
Don’t know
Copyright © Afrobarometer 2016 9
However, countries vary greatly in their citizens’ perceptions of overall election quality.
Citizens in Mauritius (91%), Senegal (87%), and Madagascar (84%) believe in overwhelming
numbers that their elections are completely or mostly free and fair (Figure 8). Malagasy views
on their 2013 general election are intriguing given persistent concerns about the
management of recent elections. Burundians’ views of their 2010 elections are also
overwhelmingly positive (81%); these views were expressed well before the 2015 presidential
election rejected by the international community.
In contrast, only minorities see their elections as free and fair in Sudan (29%), Morocco (34%),
Gabon (37% for its 2011 election), Algeria (43%), Nigeria (46%), and Ghana (46%).
Figure 8: Freeness and fairness of elections | 36 countries | 2014/2015
Respondents were asked: On the whole, how would you rate the freeness and fairness of the last
national election, held in [20xx]?
(% who say “free and fair, but with m inor problem s” or “com pletely free and fair”)
29%
34%
37%
43%
46%
46%
52%
54%
55%
56%
56%
58%
60%
61%
62%
64%
65%
65%
67%
69%
69%
70%
70%
72%
75%
75%
76%
77%
78%
80%
81%
83%
83%
83%
84%
87%
91%
0% 20% 40% 60% 80% 100%
Sudan
Morocco
Gabon
Algeria
Nigeria
Ghana
Mozambique
Liberia
Côte d'Ivoire
Cape Verde
Zimbabwe
Cameroon
Egypt
Togo
Kenya
Burkina Faso
Average
Benin
Guinea
Uganda
Sierra Leone
Malawi
Tunisia
South Africa
Lesotho
Tanzania
Swaziland
Botswana
Zambia
Namibia
Burundi
Niger
São Tomé and Príncipe
Mali
Madagascar
Senegal
Mauritius
Copyright © Afrobarometer 2016 10
Within countries, men and women offer similar assessments of the freeness and fairness of
elections. But rural residents are more likely to see elections as free and fair than urban
dwellers, 69% vs. 60%, which may reflect the amount and quality of information that reaches
rural residents and the fact that many ruling parties invest heavily in rural areas. Gaps in rural
vs. urban perceptions of whether elections were free and fair are largest in Burundi (85% rural
vs. 55% urban), Zimbabwe (63% vs. 43%), Burkina Faso (68% vs. 48%), and Togo (69% v. 50%)
(Figure 9).
Figure 9: Rural-urban differences in perceptions of elections as free and fair | 36 countries | 2014/2015
Respondents were asked: On the whole, how would you rate the freeness and fairness of the last
national election, held in [20xx]?
(% who say “free and fair, but with m inor problem s” or “com pletely free and fair”)
67%
50%
36%
72%
93%
77%
54%
57%
68%
87%
78%
75%
28%
54%
82%
67%
76%
71%
80%
41%
77%
66%
50%
70%
57%
60%
35%
52%
60%
41%
71%
72%
58%
50%
48%
43%
55%
54%
43%
30%
67%
89%
74%
51%
54%
66%
87%
78%
75%
30%
56%
85%
71%
79%
75%
84%
46%
82%
71%
57%
77%
64%
69%
44%
63%
71%
53%
85%
86%
72%
69%
68%
63%
85%
0% 20% 40% 60% 80% 100%
Egypt
Nigeria
Morocco
Tunisia
Mauritius
Lesotho
Mozambique
Cape Verde
Guinea
Senegal
Zambia
Tanzania
Sudan
Côte d'Ivoire
São Tomé and Príncipe
Malawi
Botswana
South Africa
Mali
Algeria
Namibia
Sierra Leone
Liberia
Swaziland
Kenya
Average
Gabon
Cameroon
Uganda
Ghana
Niger
Madagascar
Benin
Togo
Burkina Faso
Zimbabwe
Burundi
Rural Urban
Copyright © Afrobarometer 2016 11
While average assessments of election quality may be fairly stable over time, individual
countries have experienced significant changes. Since the previous survey round in
2011/2013, the proportion of Kenyans who describe their last national election as completely
or mostly free and fair has increased by 31 percentage points, and similar improvement are
seen in Madagascar (30 points) and Mali (26 points). At the other extreme are Ghana (a 34-
percentage–point drop), Nigeria (25 points), and Liberia (22 points) (Figure 10).
Figure 10: Freeness and fairness of elections over time | 34 countries | 2011-2015
Figure shows increase or decrease, in percentage points, from 2011/2013 to 2014/2015 in the proportion
of respondents who say their m ost recent national elections were “com pletely free and fair” or “free and fair, but with m inor problem s.”
Looking over the past decade, the proportion of citizens who see their previous election as
completely or mostly free and fair has increased by 49 percentage points in Zambia, by 27
points in Malawi, by 20 points in Zimbabwe, and by 19 points in Mali (Figure 11). This does not
necessarily mean that these countries have attained very positive views on election quality,
only that there has been a large positive change. For instance, in Zimbabwe, while positive
-34
-25
-22
-20
-12
-9
-8
-7
-7
-7
-7
-7
-6
-5
-5
-4
-3
0
1
1
2
4
5
5
5
7
8
12
14
15
18
19
26
30
31
-40 -30 -20 -10 0 10 20 30 40
Ghana
Nigeria
Liberia
Cape Verde
Mozambique
Benin
Tanzania
Burkina Faso
Zambia
Lesotho
Sudan
Morocco
Botswana
South Africa
Algeria
Senegal
Burundi
Average
Mauritius
Malawi
Sierra Leone
Tunisia
Namibia
Uganda
Guinea
Niger
Swaziland
Cameroon
Togo
Egypt
Zimbabwe
Côte d'Ivoire
Mali
Madagascar
Kenya
Copyright © Afrobarometer 2016 12
assessments have risen by 20 percentage points since 2005 (after the severely flawed 2002
elections), still only 56% of citizens say in 2014/2015 that the 2013 election was free and fair
(Figure 12a).
Notably, several countries that election support experts see as making progress on election
management experienced sharp declines in public opinion about the freeness and fairness
of elections, including Ghana (a drop of 31 percentage points), Kenya (17 points), and Benin
(9 points) (Figure 12b). As with public trust in electoral commissions, these decreases may
reflect increasingly vibrant political competition and higher citizen expectations of what a
free and fair election should look like.
Again, changes over the past decade can obscure large short -term swings in citizens’
perceptions of election quality that are highly dependent on the context of a particular
election. Some of the biggest swings have occurred in Kenya, Madagascar, and Nigeria
(Figure 12c). In Kenya, respondents in the 2011 survey rated the violent 2007 general election
process very poorly, in contrast to much more positive views of the 2013 general election.
However, ratings of the 2013 election are still more unfavourable than t hose given in the 2005
survey (concerning the 2002 general election).
For Nigeria, the 2011 election marked a high-water mark in election quality approval, after
severely flawed elections in 2003 and 2007. However, in the tense run-up to the 2015 election,
citizens may have begun to recast even the image of the 2011 election in light of what they
saw as potential problems in the upcoming contest.
Figure 11: Change in perceived freeness and fairness of elections | 18 countries
| 2005-2015
Figure shows the increase or decrease, in percentage points, from 2005/2006 to 2014/2015 in the
proportion of respondents who say their m ost recent elections were “free and fair, but with m inor problem s” or “com pletely free and fair”)
-31
-25
-17
-9
-7
-4
-3
-3
-2
2
3
4
7
9
16
19
20
27
49
-50 0 50 100
Ghana
Mozambique
Kenya
Benin
Botswana
Lesotho
Tanzania
South Africa
Cape Verde
Namibia
Average (18 countries)
Uganda
Madagascar
Senegal
Nigeria
Mali
Zimbabwe
Malawi
Zambia
Copyright © Afrobarometer 2016 13
29%
85%
78%
43%
69% 70%
36% 38%
56%
64%
58%
83%
0%
20%
40%
60%
80%
100%
2005/2006 2011/2013 2014/2015
Zambia Malawi
Zimbabwe Mali
77% 80%
46% 64% 52%
79%
31%
62%
74% 74% 65%
0%
20%
40%
60%
80%
100%
2005/2006 2011/2013 2014/2015
Ghana Mozambique
Kenya Benin
79%
31%
62% 77%
53%
84%
30%
71%
46%
0%
20%
40%
60%
80%
100%
2005/2006 2011/2013 2014/2015
Kenya Madagascar
Nigeria
Figure 12: Change in perceived freeness and fairness of elections | largest increases,
decreases, and swings | 18 countries| 2005-2015
12a: Largest increases 12b: Largest decreases
12c: Dramatic changes
Respondents were asked: On the whole, how would you rate the freeness and fairness of the last
national election, held in [20xx]?
(% who say “free and fair, but with m inor problem s”
or “com pletely free and fair”)
Citizens vs. “experts” as judges of election quality
A comparison of Afrobarometer survey findings regarding election quality with ratings by
Freedom House and the Electoral Integrity Project’s Perceptions of Election Integrity (PEI)
survey suggests that citizens generally reach similar conclusions about election quality as
experts with access to election observation reports, feedback from political leaders, and
other specialized information.
In addition to its broad ratings of civil and political freedoms, Freedom House provides a
subscore for election quality (Freedom House, 2009-2016).3 These Freedom House scores and
3 The Freedom House election scores used here correspond to the elections that the Afrobarometer
questionnaire asked about. For instance, Kenyans were asked about the 2013 election in the Afrobarometer survey, so the 2013 Freedom House election score is used.
Copyright © Afrobarometer 2016 14
citizens’ perceptions of election quality are moderately correlated (at 0.30, p < 0.076) (Figure
13). If we remove from the analysis four countries where citizens rate election quality much
higher than the experts (Madagascar, Mali, Niger, and Swaz iland, shown in the upper-left
corner of the figure), the correlation is much stronger (at 0.59, p < 0.000). In two countries
(Ghana and Cape Verde, in the lower-right corner of the figure), citizens give significantly
more critical assessments than experts.
Figure 13: Election quality: Citizen vs. Freedom House assessments | 36 countries
| 2014/2015
Correlation coefficient: 0.30, two-tailed Pearson’s correlation at p < 0.076
Source: Afrobarom eter and Freedom House
Another source of expert opinion on election quality is the Perceptions of Election Integrity
(PEI) Index (Norris, Martínez i Coma, Nai, & Groemping, 2016), which has surveyed country
experts about the quality of elections since 2012 across 49 subject areas (e.g. voter
registration and roll, boundary demarcation, media freedom). PEI calculates an overall score
for each election using the expert survey results. The imputed overall score (included in the
PEI data set) used in this analysis corrects for any missing data from the expert survey. For the
18 countries for which both PEI and Afrobarometer data are available, citizen and expert
opinions on election quality are correlated (at 0.34, p < 0.171).
While PEI Index scores were calculated only for elections from 2012 to 2015, with additional
coding in PEI format4 we can calculate a PEI Index score for all elections covered in the
Afrobarometer Round 6 questionnaire. Adding data in the PEI format for 17 countries not
covered in the PEI data yields a similar correlation (at 0.36, p < 0.031) (Figure 14).
4 As part of his dissertation project, Peter Penar has coded elections not included in the PEI Index since 2000 using the same PEI survey. These data were merged with the existing PEI data to produce Figure 14.
Algeria
Benin
Botswana
Burkina Faso
Burundi
CameroonCape VerdeCote d'Ivoire
Egypt
Gabon
Ghana
Guinea
Kenya
Lesotho
Liberia
Madagascar
Malawi
Mauritius
Morocco
Mozambique
Namibia
Nigeria
Sao Tome & Principe
Senegal
Sierra Leone
South Africa
Sudan
Swaziland Tanzania
Togo
TunisiaUganda
Zambia
Zimbabwe
Mali Niger
.2.4
.6.8
1
Ele
ctio
n q
ualit
y (
% fre
e a
nd
fair
or
with m
inor
pro
ble
ms)
0 1 2 3 4 5 6 7 8 9 10 11 12
Freedom House election subscore
Copyright © Afrobarometer 2016 15
Thus, comparisons with both Freedom House and PEI assessments suggest that citizens are
largely drawing the same conclusions about election quality as experts with extensive access
to additional information about the quality of elections.
Figure 14: Election quality: Citizens vs. Perceptions of Election Integrity Index (imputed) and additional data | 35 countries* | 2014/2015
Correlation coefficient: 0.36, two-tailed Pearson’s correlation at p < 0.031
Sources: Afrobarometer, PEI 4.0, Penar original PEI form at coding. (* No data are available for Morocco.)
Components of election quality
A free and fair election presupposes a level playing field for candidates and a safe
environment for voters to express their preferences. Elections are full cycles that include
components before Election Day (voter registration, maintenance of the voter roll), on
Election Day (voting), and after Election Day (vote count, results announcement).
Afrobarometer’s 2014/2015 survey asked citizens about their personal safety during
campaigns and at the polls, the opposition’s ability to campaign freely, whether media
coverage of all candidates was fair or biased, the fairness of the vote count, and bribery of
voters.
Fear during campaigns and at the polls
Ensuring voter safety during election campaigns and at the polls is an important part of
protecting the freedom of citizens to express their political preferences. Government security
forces can protect voters from ad hoc political violence, but they can also be used to
repress political dissent and intimidate voters under the guise of ensuring public safety.
Across 36 countries, a majority (55%) of respondents say they do not fear becoming a victim
of intimidation or violence during campaigns, whereas 30% say they fear this possibility “a
little bit” or “somewhat” and 13% say they fear violence “a lot” (Figure 15). Fear is most
Algeria
Benin
Botswana
Burkina Faso
Burundi
CameroonCape VerdeCote d'Ivoire
Egypt
Gabon
Ghana
Guinea
Kenya
Lesotho
Liberia
Madagascar
Malawi
Mali
Mauritius
Mozambique
NamibiaNiger
Nigeria
Sao Tome & Principe
Senegal
Sierra Leone
South Africa
Sudan
Swaziland Tanzania
Togo
TunisiaUganda
Zambia
Zimbabwe
.2.4
.6.8
1
Ele
ctio
n q
ualit
y (
% fre
e a
nd
fair
or
with m
inor
pro
ble
ms)
35 40 45 50 55 60 65 70 75 80
PEI overall score (with additional data)
Copyright © Afrobarometer 2016 16
common in countries with a history of election-related violence and civil conflict, such as
Côte d'Ivoire (where 33% of citizens express “a lot” of fear), Uganda (32%), Zimbabwe (25%),
Kenya (24%), and Nigeria (24%). At the other extreme, almost nine in 10 citizens in Niger,
Cape Verde, and Madagascar say they are “not at all” afraid during election campaigns.
Figure 15: Fear during election campaigns | 36 countries | 2014/2015
Respondents were asked: During election cam paigns in this country, how m uch do you personally fear
becom ing a victim of political intim idation or violence?
Similarly, about half (48%) of respondents say voters are “never” threatened with violence at
the polls on Election Day (Figure 16). But substantial minorities say that violence threatens
voters “sometimes” (25%), often (12%), or “always” (8%). At the aggregate level, there are
only small differences between urban and rural dwellers and between men and women in
their assessments of how often voters are threatened.
18%
28%
30%
32%
33%
35%
36%
36%
39%
42%
44%
46%
47%
48%
50%
52%
55%
55%
57%
59%
61%
62%
62%
63%
64%
65%
65%
65%
67%
71%
71%
72%
75%
77%
86%
87%
89%
57%
38%
52%
54%
42%
41%
31%
47%
36%
41%
32%
29%
31%
46%
28%
34%
24%
30%
30%
33%
26%
31%
17%
24%
30%
24%
31%
17%
25%
22%
21%
21%
18%
21%
14%
9%
8%
24%
33%
12%
11%
24%
23%
32%
10%
25%
11%
22%
17%
22%
5%
23%
7%
20%
13%
9%
4%
12%
5%
19%
12%
6%
11%
3%
16%
8%
5%
7%
5%
5%
1%
1%
2%
2%
0% 20% 40% 60% 80% 100%
Nigeria
Côte d'Ivoire
Egypt
Liberia
Kenya
Gabon
Uganda
Mozambique
Zimbabwe
Sudan
Zambia
Sierra Leone
Togo
Tanzania
Guinea
Algeria
Malawi
Average
Cameroon
Morocco
Swaziland
South Africa
Lesotho
Ghana
Benin
Mali
Namibia
Tunisia
Senegal
São Tomé and Príncipe
Botswana
Burundi
Burkina Faso
Mauritius
Madagascar
Cape Verde
Niger
Not at all A little bit/Somewhat A lot
Copyright © Afrobarometer 2016 17
Citizens of Botswana, Guinea, and Madagascar are least likely to express concern about
violence at the polls: More than 70% say it “never” occurs. Guineans, who have experienced
election-related violence, appear to differentiate between violence associated with
campaigns (which only 50% say they fear “not at all”) and the threat of violence on Election
Day (which 72% say “never” occurs). This indicates that citizens’ concerns about violence
can vary depending on the stage in the election cycle.
Taken together, survey findings suggest that fear is a very real part of the election process in
a number of African countries, especially Nigeria, Côte d'Ivoire, Egypt, Gabon, Kenya, and
Zimbabwe. This may call into question the legitimacy of the election process, since violence
and intimidation can be used as manipulative tools to suppress campaign activities, voter
turnout, and civil-society activities designed to ensure a free and fair election.
Figure 16: Threats of violence at the polls | 36 countries | 2014/2015
Respondents were asked: In your opinion, how often do the following things occur in this country’s
elections: Voters are threatened with violence at the polls?
14%
21%
21%
26%
27%
29%
29%
30%
31%
32%
39%
39%
42%
42%
44%
44%
45%
48%
49%
49%
50%
52%
53%
53%
59%
60%
61%
62%
62%
63%
67%
67%
69%
70%
71%
72%
73%
46%
37%
36%
35%
41%
32%
29%
30%
34%
40%
28%
24%
21%
24%
21%
32%
31%
25%
22%
23%
28%
21%
21%
26%
24%
17%
20%
27%
10%
13%
12%
16%
14%
8%
22%
15%
8%
34%
28%
42%
33%
27%
21%
25%
29%
32%
24%
27%
20%
20%
17%
21%
18%
20%
20%
17%
16%
13%
24%
14%
17%
14%
15%
6%
10%
9%
24%
18%
13%
11%
21%
6%
10%
13%
0% 20% 40% 60% 80% 100%
Nigeria
Egypt
Côte d'Ivoire
Zimbabwe
Kenya
Morocco
Mozambique
Ghana
Gabon
Tanzania
Uganda
Algeria
Sierra Leone
Cape Verde
Cameroon
Zambia
Liberia
Average
Sudan
Tunisia
South Africa
Togo
São Tomé and Príncipe
Senegal
Malawi
Swaziland
Mauritius
Benin
Lesotho
Mali
Burkina Faso
Burundi
Namibia
Niger
Madagascar
Guinea
Botswana
Never Sometimes Often/always
Copyright © Afrobarometer 2016 18
Opposition’s ability to contest
While multiparty elections cannot happen without opposition candidates, countries vary
greatly in how much freedom the opposition is given to operate. Many election-related
grievances involve treatment of opposition parties that is perceived as unfair, sometimes
leading to protests and election boycotts.
Survey findings suggest that many African countries do not fully honour the spirit of multiparty
competition. On average, fully 43% of respondents say that opposition candidates are
“sometimes,” “often,” or “always” prevented from running for office. Majorities in 12 countries
say that opposition candidates are at least sometimes prevented from contesting elections.
This view is most common in Côte d'Ivoire (64%), Zimbabwe (61%), Burundi (61%), Egypt (61%),
Nigeria (60%), and Gabon (56%) (Figure 17). Perceptions of opposition exclusion are lowest in
Botswana (22%), Niger (22%), and Guinea (23%).
Figure 17: Oposition candidates prevented from running | 36 countries | 2014/2015
Respondents were asked: In your opinion, how often do the following things occur in this country’s
elections: Opposition candidates are prevented from running for office? (% who say “som etimes,” “often,” or “always”)
22% 22% 23%
27% 27% 28% 30%
33% 33%
35% 37% 37% 38% 38%
43% 43% 44% 45% 45% 46% 46% 46% 47% 48% 48% 50% 50% 51% 51% 52% 53%
56% 60% 61% 61% 61%
64%
0% 20% 40% 60% 80% 100%
BotswanaNiger
GuineaMali
Cape VerdeLesotho
MauritiusBenin
SenegalBurkina Faso
MoroccoTunisia
Sierra LeoneNamibiaAverage
ZambiaGhana
South AfricaMalawi
CameroonSão Tomé and Principe
UgandaKenyaSudan
TogoMozambique
LiberiaAlgeria
TanzaniaSwaziland
MadagascarGabonNigeria
EgyptBurundi
ZimbabweCôte d'Ivoire
Copyright © Afrobarometer 2016 19
Media coverage during elections
Although election observation reports consistently mention bias in media coverage,
particularly in countries with a dominant state-controlled media outlet, efforts by civil society
and international organisations to develop media regulatory frameworks and engage in
media monitoring during elections can contribute to more balanced coverage. In addition,
some countries have seen the expansion of private media outlets, which can provide a
wider range of coverage.
Nonetheless, more than four in 10 Africans (43%) say the media “never” or only “sometimes”
provides fair coverage of all candidates (Figure 18). Gabonese are the most critical of media
bias (77%), followed by citizens of Madagascar (62%), Malawi (59%), Nigeria (59%),
Zimbabwe (57%), and Sudan (57%). Most citizens in Senegal, Niger, Burundi (in late 2014), and
Botswana say the media does a good job of providing fair coverage of all candidates.
Figure 18: Perceptions of unfair media coverage | 36 countries | 2014/3015
Respondents were asked: In your opinion, how often do the following things occur in this country’s
elections: The m edia provides fair coverage of all candidates? (% who say “never” or “som etimes”)
23%
25%
29%
29%
30%
30%
31%
33%
34%
35%
36%
36%
38%
38%
39%
40%
40%
40%
41%
42%
43%
43%
47%
47%
48%
48%
50%
50%
51%
55%
55%
57%
57%
59%
59%
62%
77%
0% 20% 40% 60% 80% 100%
Senegal
Niger
Burundi
Botswana
Swaziland
Lesotho
Sierra Leone
Mauritius
Kenya
Benin
Cape Verde
Uganda
Guinea
Morocco
South Africa
Mali
Ghana
Tanzania
Algeria
São Tomé and Principe
Tunisia
Average
Burkina Faso
Togo
Cameroon
Mozambique
Zambia
Namibia
Côte d'Ivoire
Egypt
Liberia
Sudan
Zimbabwe
Nigeria
Malawi
Madagascar
Gabon
Copyright © Afrobarometer 2016 20
Bribery of voters
The exchange of cash or “gifts” for votes is generally illegal and therefore difficult to observe
directly. Although available evidence questions the effectiveness of vote-buying (Bratton,
2008), survey findings suggest that it persists as a common campaign strategy.
Almost seven in 10 Africans (69%) say that voters are bribed at least “sometimes”; 43% say this
happens “often” or “always” (Figure 19). Bribery of voters is seen as pervasive in Mali (78%
often/always), Gabon (71%), and Senegal (68%). In contrast, fewer than one in five citizens of
Lesotho (15%), Namibia (16%), and Burundi (18%) believe that voters are routinely bribed.
Urbanites are more likely than rural residents to say that voters are often or always bribed
(48% vs. 40%), while men and women are about equally likely to hold this view.
Figure 19: How often voters are bribed during elections | 36 countries | 2014/2015
Respondents were asked: In your opinion, how often do the following things occur in this country’s
elections: Voters are bribed?
15%
16%
18%
20%
21%
25%
27%
29%
31%
33%
33%
34%
34%
34%
35%
37%
39%
41%
43%
45%
46%
49%
49%
52%
54%
54%
54%
56%
56%
57%
58%
58%
60%
64%
68%
71%
78%
15%
23%
21%
20%
44%
24%
28%
24%
30%
34%
34%
16%
30%
45%
34%
44%
26%
32%
26%
39%
23%
26%
25%
20%
20%
30%
20%
16%
32%
33%
19%
20%
25%
26%
17%
20%
11%
48%
50%
54%
52%
35%
28%
35%
30%
33%
17%
20%
35%
22%
10%
25%
9%
29%
21%
22%
13%
15%
12%
18%
15%
19%
14%
7%
24%
9%
6%
10%
16%
13%
4%
9%
7%
10%
22%
10%
7%
8%
23%
11%
16%
6%
16%
12%
14%
15%
10%
6%
10%
6%
7%
9%
3%
14%
12%
8%
13%
7%
3%
18%
4%
4%
4%
13%
6%
3%
6%
7%
2%
0% 20% 40% 60% 80% 100%
Lesotho
Namibia
Burundi
Guinea
Madagascar
Mozambique
South Africa
Sierra Leone
Malawi
Algeria
Zimbabwe
Botswana
Sudan
Egypt
Zambia
Mauritius
Niger
Tanzania
Average
Liberia
São Tomé and Príncipe
Ghana
Uganda
Cameroon
Togo
Côte d'Ivoire
Cape Verde
Burkina Faso
Kenya
Nigeria
Tunisia
Swaziland
Benin
Morocco
Senegal
Gabon
Mali
Often/always Sometimes Never Don’t know
Copyright © Afrobarometer 2016 21
Fairness of the vote count
The vote count is arguably one of the most important parts of the election cycle. Counting
votes requires high levels of logistical competence as well as transparency to overcome
public skepticism about the intentions of the government and electoral commission to
conduct the count fairly. Low levels of public confidence in the vote count put a country at
risk of post-election protest and violence in response to perceived voter fraud and
manipulation.
Across 36 countries, just one-third (34%) of citizens think that votes are “always” counted fairly
(Figure 20). One in five (19%) say the vote count is “often” fair, while a plurality say the vote
count is “never” (13%) or only “sometimes” (25%) fair.
Figure 20: Fairness of the vote count | 36 countries | 2014/2015
Respondents were asked: In your opinion, how often do the following things occur in this country’s
elections: Votes are counted fairly?
6%
15%
15%
16%
16%
19%
20%
22%
22%
24%
26%
26%
27%
28%
30%
30%
31%
32%
33%
34%
34%
36%
36%
37%
37%
38%
38%
40%
41%
45%
47%
56%
57%
59%
63%
67%
69%
17%
27%
12%
19%
17%
25%
25%
15%
17%
12%
24%
17%
29%
14%
13%
27%
13%
9%
13%
19%
31%
33%
9%
14%
27%
21%
26%
12%
31%
22%
19%
9%
16%
20%
17%
22%
14%
0% 20% 40% 60% 80% 100%
Nigeria
Morocco
Gabon
Algeria
Sudan
Cameroon
Egypt
Liberia
Zimbabwe
Malawi
Tunisia
Kenya
Tanzania
Ghana
Cape Verde
Côte d'Ivoire
Togo
Mozambique
Zambia
Average
Benin
Burkina Faso
Sierra Leone
Uganda
Madagascar
Lesotho
Burundi
Guinea
Mali
Swaziland
South Africa
São Tomé and Príncipe
Botswana
Namibia
Senegal
Mauritius
Niger
Always Often
Copyright © Afrobarometer 2016 22
Large majorities have faith in the vote count in Niger (where 69% say it is “always” fair),
Mauritius (67%), and Senegal (63%), but in 30 out of 36 surveyed countries, fewer than half of
all citizens fully trust the vote count. Citizens in Nigeria (6%), Morocco (15%), Gabon (15%),
Algeria (16%), and Sudan (16%) are the most skeptical. (See Appendix Table A.2 for a
detailed breakdown.) Unsurprisingly, countries w ith a history of election-related protests and
violence express quite low levels of confidence in the vote count, including Kenya (26%),
Zimbabwe (22%), and Nigeria (6%). Rural residents are more likely to believe that votes are
counted fairly than urban dwellers (57% v. 49%).
Do elections deliver?
Beyond issues of election management and conduct, Afrobarometer asked three questions
to explore whether people perceive that elections function effectively as a way to select
representative and accountable leaders: 1) whether voters are offered a genuine choice at
the polls, 2) how well their elections ensure that the people’s views are represented, and 3)
how well their elections enable voters to remove underperforming leaders from office.
With regard to choice, six in 10 Africans (61%) affirm that elections “often” or “always” offer
voters a genuine choice (Figure 21). More than eight in 10 citizens in Niger (89%), Senegal
(86%), Mali (84%), and Madagascar (81%) say they are offered real choices. But less than four
in 10 share this view in Nigeria (32%), Sudan (34%), and Egypt (39%). While the Sudan and
Egypt results are not surprising, given government’s role in vetting candidates, the small
proportion of Nigerians who feel they have a real choice in their country’s competitive
elections may reflect a general skepticism or dismay about the electoral process.
Even if most Africans believe they are offered real choices at the ballot box, many are
dissatisfied with elections as a vehicle for ensuring that leaders are representative and
accountable. Half (50%) of respondents say that their elections function “not very well” or
“not at all well” to ensure that those elected reflect the views of the people (Figure 22).
Similarly, 51% believe that their elections do not work well in enabling voters to remove
leaders who don’t do what the people want .
Copyright © Afrobarometer 2016 23
Figure 21: Voters given genuine choice at the polls | 36 countries | 2014/2015
Respondents were asked: In your opinion, how often do the following things occur in this country’s
elections: Voters are offered a genuine choice in elections? (% who say “always” or “often”)
32%
34%
39%
40%
43%
45%
46%
48%
50%
52%
52%
54%
54%
56%
60%
60%
60%
61%
62%
63%
64%
65%
65%
67%
68%
69%
71%
73%
74%
74%
74%
75%
78%
81%
84%
86%
89%
0% 20% 40% 60% 80% 100%
Nigeria
Sudan
Egypt
Mozambique
Algeria
Gabon
Zimbabwe
Tanzania
Liberia
São Tomé and Principe
Morocco
Kenya
Cape Verde
Cameroon
Namibia
Côte d'Ivoire
Sierra Leone
Average
Zambia
Ghana
Guinea
Togo
Uganda
Tunisia
South Africa
Swaziland
Benin
Burundi
Lesotho
Malawi
Mauritius
Burkina Faso
Botswana
Madagascar
Mali
Senegal
Niger
Copyright © Afrobarometer 2016 24
Figure 22: Performance of elections in Africa | 36 countries | 2014/2015
Respondents were asked: Think about how elections work in practice in this country. How well do
elections: 1. Ensure that m embers of Parliam ent/National Assembly representatives reflect the views of
voters?
2. Enable voters to rem ove from office leaders who do not do what the people want?
Namibians, Batswana, and Tunisians are most confident that elections can ensure that
constituents’ views are represented, while Gabonese, Malagasy, and Moroccans are most
skeptical (Figure 23; see Appendix Table A.3 for details).
Figure 23: Representation function of elections | best- and worst-performing
countries | 2014/2015
Respondents were asked: Think about how elections work in practice in this country. How well do elections ensure that m em bers of Parliam ent/National Assem bly representatives reflect the views of
voters?
50% 51%
42% 40%
8% 9%
0%
10%
20%
30%
40%
50%
60%
Elections ensure voters' views are
reflected
Elections enable voters to remove
underperforming leaders from office
Not at all well/not very well Well/very well Don’t know
63% 62% 58% 57% 56% 56% 51% 51% 51%
76%
67% 67% 64% 62% 62% 62% 60% 60%
0%
20%
40%
60%
80%
100%
Na
mib
ia
Bo
tsw
an
a
Tun
isia
Ma
uritiu
s
Nig
er
Ma
li
Gh
an
a
Zam
bia
Gu
ine
a
Ga
bo
n
Ma
da
ga
sca
r
Mo
roc
co
Alg
eria
Nig
eria
Sw
azi
lan
d
Lib
eria
Su
da
n
Cô
te d
'Ivo
ire
Best performers
(% well or very well)
Worst perfomers
(% not at all well/not very well)
Copyright © Afrobarometer 2016 25
When it comes to holding elected officials accountable, Senegalese, Cape Verdeans,
Batswana, and Ghanaians are most optimistic that elections enable voters to remove non-
performing leaders (Figure 24; see Appendix Table A.4 for details). In contrast, Gabonese,
Swazis, Sudanese, Moroccans, Nigerians, and Algerians are most skeptical that elections help
ensure accountability. Four of these countries face significant barriers to the election
accountability mechanism: Swaziland and Morocco are kingdoms with tight controls over
who is able to compete in elections and limited power for those elected, and Sudan and
Algeria have been de facto one-party states.
Figure 24: Accountability function of elections | best- and worst-performing countries
| 2014/2015
Respondents were asked: Think about how elections work in practice in this country: How well do elections enable voters to rem ove from office leaders who do not do what the people want?
Across 18 countries tracked since 2005/2006, the proportion of citizens who say elections
ensure that people’s views are represented has remained stable; about 6% of respondents
have moved out of the “don’t know” column to say that elections do not perform this
representation function well (Figure 25). A similar trend holds for the question on
accountability.
Looking at over-time changes at the country level (Figure 26), Zambia, which has
experienced several transfers of power between different political parties, registers the
greatest improvement among 18 countries tracked over the past decade – a 22-
percentage-point increase in the proportion of citizens who say that elections function well
to ensure representation. Zimbabwe, Senegal, Malawi, and Mali are among countries that
have seen more modest improvements.
Several countries with higher or recently increased levels of political competition exhibit
decreases in the proportion of citizens who say that elections ensure representation: Ghana
(a decrease of 19 percentage points), Mozambique (14 points), Benin (7 points), and South
Africa (6 points). In some cases, this may reflect a public questioning of liberation-party
hegemony (e.g. Mozambique, South Africa) and voters’ growing expectations of what true
representation should look like.
66% 64% 61% 61% 59% 54%
50% 50%
79%
69% 69% 69% 66% 65% 63% 63% 61%
0%
20%
40%
60%
80%
100%
Se
ne
ga
l
Ca
pe
Ve
rde
Bo
tsw
an
a
Gh
an
a
Zam
bia
Tun
isia
Gu
ine
a
Ke
nya
Ga
bo
n
Sw
azi
lan
d
Su
da
n
Mo
roc
co
Nig
eria
Alg
eria
Be
nin
Ma
da
ga
sca
r
Cô
te d
'Ivo
ire
Best performers
(% well/very well)
Worst performers
(% not at all well/not very well)
Copyright © Afrobarometer 2016 26
Figure 25: Changes in performance of elections | 18 countries | 2005-2015
Respondents were asked: Think about how elections work in practice in this country. How well do
elections: 1. Ensure that m embers of Parliam ent/National Assembly representatives reflect the views of
voters?
2. Enable voters to rem ove from office leaders who do not do what the people want?
(Note: This question was not included in the 2011/2013 survey round.)
Figure 26: Changes in proportions who say elections perform ‘well’ or ‘very well’ in ensuring representation | 18 countries | 2005-2015
Respondents were asked: Think about how elections work in practice in this country. How well do
elections ensure that m em bers of Parliam ent/National Assem bly representatives reflect the views of
voters? (Figure shows the difference, in percentage points, between surveys in 2005/2006 and 2014/2015 in the
proportion of respondents who say that elections perform “well” or “very well” in ensuring that elected
officials represent the views of citizens. Positive num bers indicate increasingly positive assessments.)
41% 40% 47% 40% 41% 47%
46% 47% 46%
47% 46% 46%
13% 12% 7% 13% 13% 7%
0%
20%
40%
60%
80%
100%
2005/2006 2008/2009 2014/2015 2005/2006 2008/2009 2014/2015
Elections ensure voters' views are
reflected
Elections enable voters to remove
underperforming leaders from office
Not at all well/not very well Well/very well Don’t know
-19
-14
-7
-6
-6
-6
-3
-3
-3
-1
0
3
5
5
7
8
8
8
22
-25 -20 -15 -10 -5 0 5 10 15 20 25
Ghana
Mozambique
Benin
Madagascar
South Africa
Uganda
Botswana
Namibia
Tanzania
Lesotho
Average, 18 countries
Nigeria
Cape Verde
Kenya
Mali
Malawi
Senegal
Zimbabwe
Zambia
Copyright © Afrobarometer 2016 27
Do your own analysis of Afrobarometer data – on any question, for any country and survey
round. It’s easy and free at
www.afrobarometer.org/online-data-analysis.
With respect to the accountability function of elections, citizens in Ghana, Madagascar,
Mozambique, and Benin express increasing pessimism, whereas Zambia and Cape Verde
experienced the greatest gains in public confidence (Figure 27).
Figure 27: Changes in proportions who say elections perform ‘well’ or ‘very well’ in enabling voters to remove non-performing leaders | 18 countries | 2005-2015
Respondents were asked: Think about how elections work in practice in this country. How well do
elections enable voters to rem ove from office leaders who do not do what the people want? (Figure shows the difference, in percentage points, between surveys in 2005/2006 and 2014/2015 in the
proportion of respondents who say that elections perform “well” or “very well” in enabling voters to
rem ove underperform ing leaders. Positive num bers reflect increasingly positive assessments.)
Election quality and trust in electoral commission
As we have seen, two important concerns about African elections are the fairness of the
vote count and the freedom that opposition parties have to contest elections. A vote count
that lacks transparency is an easy place for leaders to manipulate election outcomes, and
interference with the opposition is likely to
slow even the strongest opposition parties
from gaining momentum and expanding
their reach.
As might be expected, Afrobarometer
results indicate that public perceptions of
how fairly votes are counted and the
opposition is treated affect citizens’ overall
assessments of election quality as well as their trust in the electoral commission. Citizens who
believe that votes are counted fairly are more likely to believe that elections are “completely
free and fair” or “free and fair, but with minor problems” (Figure 28). The same is true for the
relationship between opposition suppression and election quality (Figure 29).
-18
-16
-16
-14
-11
-7
-3
-3
-2
-2
-1
1
3
4
6
11
11
15
30
-30 -20 -10 0 10 20 30 40
Ghana
Madagascar
Mozambique
Benin
Namibia
Malawi
Botswana
South Africa
Mali
Uganda
Average, 18 countries
Tanzania
Nigeria
Lesotho
Zimbabwe
Kenya
Senegal
Cape Verde
Zambia
Copyright © Afrobarometer 2016 28
Figure 28: Votes counted fairly and overall election quality | 36 countries
| 2014/2015
Respondents were asked:
1. On the whole, how would you rate the freeness and fairness of the last national election, held in [20xx]?
2. In your opinion, how often do the following things occur in this country’s elections: Votes are
counted fairly?
Figure 29: Opposition party suppression and overall election quality | 36 countries
| 2014/2015
Respondents were asked:
1. On the whole, how would you rate the freeness and fairness of the last national election, held
in [20xx]? 2. In your opinion, how often do the following things occur in this country’s elections: Opposition
candidates are prevented from running for office?
83%
69%
57%
41%
11%
23%
35%
51%
0%
20%
40%
60%
80%
100%
Votes always
counted fairly
Votes often
counted fairly
Votes sometimes
counted fairly
Votes never
counted fairly
Election completely free and fair or with minor problems
Election not free and fair or major problems
74%
60% 57% 60%
19%
33% 34% 34%
0%
20%
40%
60%
80%
Opposition never
prevented from
running
Opposition
sometimes
prevented from
running
Opposition often
prevented from
running
Opposition always
prevented from
running
Election completely free and fair or with minor problems
Election not free and fair or major problems
Copyright © Afrobarometer 2016 29
Since electoral commissions are responsible for managing elections, one would expect that
citizens’ assessments of election quality influence their trust in the commission. Survey findings
confirm that citizens who view elections as free and fair are more likely to have higher levels
of trust in their electoral commissions (Figure 30). Similarly, respondents who hold favourable
views of the fairness of the vote count and the treatment of the opposition are also more
likely to express trust in the electoral commission.
Figure 30: Overall election quality and trust in electoral commissions | 35 countries*
| 2014/2015
Respondents were asked:
1. On the whole, how would you rate the freeness and fairness of the last national election, held in [20xx]?
2. How m uch do you trust each of the following, or haven’t you heard enough about them to
say: National Electoral Com m ission? (* Question was not asked in Morocco.)
Conclusion
The past decade has seen significant investments by civil society, international organisations,
and some national electoral commissions in facilitating dialogue and reforms in election
management in Africa. Despite these efforts, citizens continue to hold mixed views of the
quality of elections and of how well they ultimately perform their functions of ensuring that
voters’ views are represented and officeholders are accountable. By and large, citizens’
views on the state of election quality align with assessments by country experts.
Fears of voter intimidation, perceptions of unfair treatment of opposition parties, and
concerns about the manipulation of election processes and results persist in many countries,
and countries with upcoming elections (2016-2018) have some of the lowest levels of trust in
their national electoral commissions. Such concerns have the potential to influence whether
the people’s will is expressed on Election Day, since fear or a perception of unfairness can
cause voices to remain silent.
These findings suggest that additional reforms and greater transparency are needed to
ensure free and fair elections. Increasing public trust in electoral commissions will depend on
better management of components of the election process, from voter rolls to vote counts.
Election assistance should be sustained, rather than focused only on early pre-election and
Election Day activities, and should be accompanied by public sensitization efforts to ensure
that steps forward are matched by gains in public perceptions and trust.
83%
75%
63%
43%
11% 17%
29%
47%
0%
20%
40%
60%
80%
100%
Trust EC a lot Trust EC somewhat Trust EC just a little Trust EC not at all
Election completely free and fair or with minor problems
Election not free and fair or major problems
Copyright © Afrobarometer 2016 30
References
Bentley, T., Han, K., & Penar, P. H. (2015). African democracy update: Democratic satisfaction remains elusive for many. Afrobarometer Dispatch No. 45. Available at www.afrobarometer.org.
Bratton, M. (2008). Vote buying and violence in Nigerian election campaigns. Electoral Studies, 27(4), 621-632.
Freedom House. (2009-2016). Freedom in the world. Available at www.freedomhouse.org.
Norris, P., Martínez i Coma, F., Nai, A., & Groemping. M. (2016). The expert survey of perceptions of electoral integrity, PEI_4.0. Available at www.electoralintegrityproject.com and http://thedata.harvard.edu/dvn/dv/PEI.
Appendix
Table A.1: Surveyed countries, election years, and survey fieldwork dates
Country Last national
election preceding Round 6 survey
Months when Round 6 fieldwork was
conducted Previous survey rounds
Algeria 2014 May-June 2015 2013
Benin 2011 May-June 2014 2005, 2008, 2011
Botswana 2009 June-July 2014 1999, 2003, 2005, 2008,
2012
Burkina Faso 2012 April-May 2015 2008, 2012
Burundi 2010 September-October
2014 2012
Cameroon 2013 January-February
2015 2013
Cape Verde 2011 November-December
2014 2002, 2005, 2008, 2011
Côte d'Ivoire 2011 August-September
2014 2013
Egypt 2014 June-July 2015 2013
Gabon 2011 September-October
2015 N/A
Ghana 2012 May-June 2014 1999, 2002, 2005, 2008,
2012
Guinea 2013 March-April 2015 2013
Kenya 2013 November-December
2014 2003, 2005, 2008, 2011
Lesotho 2012 May 2014 2000, 2003, 2005, 2008,
2012
Liberia 2011 May 2015 2008, 2012
Madagascar 2013 December 2015-
January 2015 2005, 2008, 2013
Malawi 2009 March-April 2014 1999, 2003, 2005, 2008,
2012
Mali 2013 December 2014 2001, 2002, 2005, 2008,
2013
Mauritius 2010 June-July 2014 2012
Morocco 2011 November 2015 2013
Mozambique 2014 June-August 2015 2002, 2005, 2008, 2012
Namibia 2009 August-September
2014 1999, 2003, 2006, 2008,
2012
Niger 2011 April 2015 2013
Nigeria 2011 December 2014-
January 2015 2000, 2003, 2005, 2008,
2013
São Tomé and Príncipe
2014 July-August 2015 N/A
Senegal 2012 November-December
2014 2002, 2005, 2008, 2013
Copyright © Afrobarometer 2016 1
Country Last national
election preceding Round 6 survey
Months when Round 6 fieldwork was
conducted Previous survey rounds
Sierra Leone 2012 May-June 2015 2012
South Africa 2014 August-September
2015 2000, 2002, 2006, 2008,
2011
Sudan 2015 June 2015 2013
Swaziland 2013 April 2015 2013
Tanzania 2010 August-November
2014 2001, 2003, 2005, 2008,
2012
Togo 2013 October 2014 2012
Tunisia 2014 April-May 2015 2013
Uganda 2011 May 2015 2000, 2002, 2005, 2008,
2012
Zambia 2011 October 2014 1999, 2003, 2005, 2009,
2013
Zimbabwe 2013 November 2014 1999, 2004, 2005, 2009,
2012
Copyright © Afrobarometer 2016 2
Table A.2: Perceived fairness of the vote count | 36 countries |2014/2015
Always Often Sometimes Never Don’t know
Niger 69% 14% 9% 3% 6%
Mauritius 67% 22% 6% 1% 4%
Senegal 63% 17% 9% 3% 8%
Namibia 59% 20% 13% 4% 4%
Botswana 57% 16% 13% 7% 7%
São Tomé and Príncipe 56% 9% 16% 5% 13%
South Africa 47% 19% 21% 7% 7%
Swaziland 45% 22% 20% 5% 9%
Mali 41% 31% 17% 9% 3%
Guinea 40% 12% 21% 20% 7%
Burundi 38% 26% 23% 7% 6%
Lesotho 38% 21% 22% 7% 13%
Madagascar 37% 27% 22% 13% 1%
Uganda 37% 14% 26% 15% 8%
Sierra Leone 36% 9% 21% 14% 21%
Burkina Faso 36% 33% 14% 11% 7%
Benin 34% 31% 25% 7% 2%
Average 34% 19% 25% 13% 8%
Zambia 33% 13% 39% 9% 6%
Mozambique 32% 9% 32% 15% 11%
Togo 31% 13% 18% 27% 11%
Côte d'Ivoire 30% 27% 23% 17% 3%
Cape Verde 30% 13% 28% 11% 18%
Ghana 28% 14% 33% 15% 9%
Tanzania 27% 29% 33% 6% 5%
Kenya 26% 17% 35% 18% 4%
Tunisia 26% 24% 25% 12% 13%
Malawi 24% 12% 34% 23% 8%
Zimbabwe 22% 17% 33% 20% 7%
Liberia 22% 15% 49% 12% 2%
Egypt 20% 25% 35% 8% 12%
Cameroon 19% 25% 21% 21% 14%
Sudan 16% 17% 27% 25% 14%
Algeria 16% 19% 31% 23% 12%
Gabon 15% 12% 31% 41% 2%
Morocco 15% 27% 33% 10% 16%
Nigeria 6% 17% 52% 23% 3%
Respondents were asked: In your opinion, how often do the following things occur in this country’s
elections: Votes are counted fairly?
Copyright © Afrobarometer 2016 3
Table A.3: Performance of elections in ensuring elected leaders reflect voters’ views | 36 countries | 2014/2015
Not at all well/not very well Well/very well Don’t know
Gabon 76% 23% 2%
Algeria 64% 24% 13%
Morocco 67% 25% 8%
Cameroon 57% 29% 13%
Côte d'Ivoire 60% 31% 9%
Togo 59% 32% 10%
Swaziland 62% 32% 6%
Sierra Leone 51% 32% 17%
Liberia 62% 33% 5%
Madagascar 67% 33% 0%
Sudan 60% 33% 7%
Nigeria 62% 34% 4%
São Tomé and Príncipe 45% 36% 19%
Egypt 49% 38% 13%
Tanzania 57% 38% 5%
Malawi 58% 38% 3%
Zimbabwe 56% 39% 5%
Burkina Faso 54% 40% 6%
Lesotho 41% 41% 18%
Mozambique 42% 42% 17%
Average 50% 42% 8%
Benin 53% 43% 4%
South Africa 50% 44% 6%
Burundi 51% 44% 5%
Cape Verde 39% 45% 15%
Kenya 47% 47% 6%
Senegal 46% 48% 6%
Uganda 44% 49% 7%
Guinea 42% 51% 7%
Zambia 38% 51% 11%
Ghana 41% 51% 7%
Niger 34% 56% 10%
Mali 43% 56% 1%
Mauritius 34% 57% 8%
Tunisia 34% 58% 8%
Botswana 31% 62% 7%
Namibia 34% 63% 3%
Respondents were asked: Think about how elections work in practice in this country. How well do
elections ensure that m em bers of Parliam ent/Nation Assembly representatives reflect the views of voters?
Copyright © Afrobarometer 2016 4
Table A.4: Performance of elections in enabling voters to remove underperforming leaders from office | 36 countries | 2014/2015
Not at all well/not very well Well/very well Don’t know
Gabon 79% 19% 2%
Swaziland 69% 23% 8%
Sudan 69% 23% 8%
Morocco 69% 22% 9%
Nigeria 66% 29% 5%
Algeria 65% 20% 15%
Benin 63% 32% 4%
Madagascar 63% 37% 0%
Côte d'Ivoire 61% 30% 9%
Burkina Faso 60% 32% 9%
Togo 59% 32% 9%
Mali 57% 42% 0%
South Africa 57% 36% 7%
Tanzania 56% 39% 5%
Cameroon 56% 28% 15%
Zimbabwe 56% 38% 6%
Egypt 54% 30% 15%
Average 51% 40% 9%
São Tomé and Príncipe 50% 31% 18%
Mozambique 50% 32% 18%
Namibia 50% 47% 3%
Malawi 48% 49% 3%
Burundi 47% 48% 5%
Uganda 47% 45% 8%
Liberia 47% 47% 6%
Kenya 46% 50% 4%
Niger 46% 43% 11%
Mauritius 44% 44% 11%
Guinea 42% 50% 8%
Sierra Leone 38% 44% 18%
Tunisia 36% 54% 11%
Lesotho 35% 46% 18%
Ghana 33% 61% 6%
Botswana 31% 61% 8%
Zambia 31% 59% 10%
Senegal 28% 66% 6%
Cape Verde 24% 64% 12%
Respondents were asked: Think about how elections work in practice in this country. How well do
elections enable voters to rem ove from office leaders who do not do what the people want?
Copyright © Afrobarometer 2016 5
Other Round 6 global releases
Where to start? Aligning sustainable development goals with citizen priorities. (2015).
Afrobarometer Dispatch No. 67. Available at http://afrobarometer.org/sites/default/
files/publications/Dispatches/ ab_r6_dispatchno67_african_priorities_en.pdf.
Building on progress: Infrastructure development still a major challenge in Africa.
(2016). Afrobarometer Dispatch No. 69. Available at www.afrobarometer.org/
publications/ad69-building-progress-infrastructure-development-still-major-challenge-
africa.
Africa’s growth dividend? Lived poverty drops across much of the continent. (2016).
Afrobarometer Policy Paper No. 29. Available at http://www.afrobarometer.org/
publications/pp29-africas-growth-dividend-lived-poverty-drops-across-the-continent.
Good neighbours? Africans express high levels of tolerance for many, but not for all.
(2016). Afrobarometer Dispatch No. 74. Available at http://afrobarometer.org/
publications/tolerance-in-africa.
Off-grid or ‘off-on’: Lack of access, unreliable electricity supply still plague majority of Africans. Afrobarometer Dispatch No. 75. Available at
http://afrobarometer.org/publications/ad75-unreliable-electricity-supply-still-plague-
majority-of-africans.
Lack of safe water, sanitation spurs growing dissatisfaction with government performance. Afrobarometer Dispatch No. 76. Available at http://afrobarometer.org/
publications/ad76-lack-of-safe-water-and-sanitation-spurs-growing-dissatisfaction.
Despite gains, barriers keep health care high on Africa’s priority list. Afrobarometer
Policy Paper No. 31. Available at http://www.afrobarometer.org/publications/pp31-
despite-gains-barriers-keep-health-care-high-on-africas-priority-list.
Strong public support for ‘watchdog’ role backs African news media under attack. Afrobarometer Dispatch No. 85. Available at http://afrobarometer.org/publications/
ad85-media_in_africa_world_press_freedom_ day_2016.
Regional integration for Africa: Could stronger public support turn ‘rhetoric into reality’? Afrobarometer Dispatch No. 91. Available at http://afrobarometer.org/
publications/ad91-regional-integration-africa-could-stronger-public-support-turn-
rhetoric-reality.
Does less engaged mean less empowered? Political participation lags among African youth, especially women. Afrobarometer Policy Paper No. 34. Available at
http://www.afrobarometer.org/publications/youth-day-2016.
Do trustworthy institutions matter for development? Corruption, trust, and government performance in Africa. Afrobarometer Dispatch No. 112. Available at
http://www.afrobarometer.org/publications/ad112-do-trustworthy-institutions-matter-
development-corruption-trust-and-government.
Copyright © Afrobarometer 2016 6
Peter Penar is a research assistant for Afrobarometer and PhD student in the
Department of Political Science at Michigan State University. Email:
[email protected]. Twitter: @PPenar.
Rose Aiko is a researcher in Dar es Salaam, Tanzania. Email:
Thomas Bentley is a research assistant for Afrobarometer and PhD student in the
Department of Political Science at Michigan State University. Email:
Kangwook Han is a research assistant for Afrobarometer and PhD student in the
Department of Political Science at Michigan State University. Email:
Afrobarometer is produced collaboratively by social scientists from more than 30
African countries. Coordination is provided by the Center for Democratic
Development (CDD) in Ghana, the Institute for Justice and Reconciliation (IJR)
in South Africa, the Institute for Development Studies (IDS) at the University of
Nairobi in Kenya, and the Institute for Empirical Research in Political Economy
(IREEP) in Benin. Michigan State University (MSU) and the University of Cape Town
(UCT) provide technical support to the network.
Core support for Afrobarometer Rounds 5 and 6 has been provided by the UK’s
Department for International Development (DFID), the Mo Ibrahim Foundation,
the Swedish International Development Cooperation Agency (SIDA), the United
States Agency for International Development (USAID), and the World Bank.
Donations help the Afrobarometer Project give voice to African citizens. Please
consider making a contribution (at www.afrobarometer.org) or contact Aba
Kittoe ([email protected]) to discuss institutional funding.
For more information, please visit www.afrobarometer.org.
Cover photo: MONUSCO/Myriam Asm ani [CC BY-SA 2.0 (http://creativecommons.org/licenses/by-sa/2.0)], via Wikimedia Commons.
Infographic design by Soapbox, www.soapbox.com
Contact: [email protected]
Afrobarometer Policy Paper No. 35 | September 2016