Ghana’s national election in 2016 was hailed as a success. It was the seventh poll since the birth of the Fourth Republic in 1992. But according to the results, nearly 5 million voters did not turn up to cast their ballots. What accounted for this situation? Voter apathy or there was a problem with the voters roll?
Elections are a test of a country’s democratic robustness and offer the citizenry an opportunity to participate in the selection of local and national leaders. In Africa, elections have in some cases been characterised by suppression of opposing parties through intimidation, puffed electoral rolls, opaque electoral processes, rigging and instances of vote buying by contending parties. This has resulted in post-election violence in many African countries – recent examples can be cited in Zambia, Zimbabwe, Kenya and Democratic Republic of Congo. Moreover, in some cases this resulted in situations of power-sharing governments – such as in Kenya and Zimbabwe. Another uncertainty that mars African elections is the possibility of incumbents failing to concede defeat and handover power should they lose the elections, or incumbents circumventing the constitution to stay in power beyond the constitutional term limits. The sum total of these multiple factors has culminated into cases of post-election violence and heightened confrontations in many African countries, which sometimes require international mediation from within and without Africa. Immediate reference can be made to the recent post-election impasse in The Gambia, in which incumbent Yahya Jammeh a few days after conceding defeat made a u-turn to say that the election was rigged and declared the results null and void. It should be noted that, there are no perfect electoral registers or elections mechanism anywhere in the world.
However, it is not all dark and gloomy for African democracy as some countries like Botswana, South Africa and Ghana present real hope for the future of competitive democracy in Africa. In the case of Ghana, very significant strides have been chalked in democratic practice, having conducted relatively free and fair elections, including the 2016 general elections and producing three turnovers. These notwithstanding there have been claims and counterclaims of irregularities and malpractices in the electioneering processes. Topmost of these allegations in the lead up to the 2016 election was a bloat of the voters register. Analysts have argued that the turnout of 68.62% was quite disappointing considering the euphoria preceding the elections and the fact that nearly 5 million voters failed to cast their ballot. In addition, this is largely being attributed to either the register being stuffed with names of unqualified persons or people just did not vote. This analysis looks further into both arguments and draws conclusions based on supportive evidence.
Electoral register and voter turnout in the Diaspora
The processes governing elections determine the validity and acceptability of the election outcome. One of the key variables in ensuring elections are fair and the outcomes reflect the will of the people is the electoral roll. Credibility of the register comes from institutionalised mechanisms that allow only qualified voters to have their names in the voters register. In almost every country around the world, the minimum qualification age for voting is 18 years. It is therefore quite straightforward to forecast the percentage of the population who should appear on the electoral roll, given a country’s population census and demographic profile.
According to The Economist Intelligence Unit (2015), the top 5 ranked democracies in Africa are Botswana, Cape Verde, South Africa, Ghana and Tunisia. In these democracies, the average number of people registered to vote barely exceeds 50% of national populations. In Ghana, the situation is different – it has the highest percentage of the population – 55% on the electoral role. For instance, in South Africa, 26,333,353 representing about 47% of its approximately 55.9 million population are registered voters. Whereas, the World Population Review estimates Botswana’s population to be 2,308,524 in 2016 (annual growth rate 1.73), out of this figure 824,073 (36%) appeared on the electoral roll for the last elections in 2014.
Election turnout in Ghana 1992 to 2016
Ghana has conducted seven successive elections since 1992 when it ushered in the Fourth Republic. Turnout at these elections has been encouraging with average of about 71% of all eligible and registered voters voting. Specifically, the 1992, 1996, 2000, 2004, 2008, and 2012 elections registered a total voter turnout of 53.75%, 78.20%, 60.40%, 85.10%, 72.90%, 79% and 68.62% respectively. The highest voter turnout in Ghana since 1992 was the 2004 general elections when 8,813,908 of 10,354,970 registered voters, representing 85%, cast their ballot. So, generally, Ghana has had considerably high voter turnout for all but 1992 and 2000 election.
It is therefore premature to think that the 2016 election recorded relatively low turn because Ghanaian voters have grown tangential to political activism, including voting in presidential elections. At best, the argument of voter apathy may account for just a fraction of a rather bigger issue. The following table gives a summary of voter turnout against registered voters from 1992 to 2016.
Voter turnout in the Fourth Republic (1992-2016)
Total votes cast
Voter turnout (%)
Apathy or a bloated register?
People’s decision to vote in elections may be influenced by several factors. In Ghana’s 2016 elections, there were concerns about the turnout which many considered as low, considering the massive number (4,930,890 million) that failed to cast their ballot on polling day. A number of competing explanations have been put forward to account for this. The first possible explanation is voter apathy. In other words people have grown discontent with the political system and have decided to withdraw their participation to register their fatigue and anxiety.
Voter apathy mainly occurs when voters are required to vote too often or the electorate feel disengaged, or both. The former is less likely in the Ghanaian situation since elections are called not too frequently on a national scale. The latter makes a lot of sense considering the enormity of socioeconomic challenges faced by the outgoing government and the subsequent public outcry for change. However, voter apathy may work best in a scenario where the electorate have no alternative candidate to vote for and where they will have only one candidate to endorse. In that case, people may simply decide not to vote if they feel the government will not work to serve their interest. In the case of Ghana, where multiple candidates contest the elections and where all have a fair chance of winning, dissatisfaction or disengagement of the previous administration is likely to serve as an incentive for people to come out in their numbers to vote out the perceived non-performing government. It therefore weakens the argument that disengagement from government may have angered the electorate from voting. We must look further in our attempt to trace source of this voter absenteeism.
The next possible explanation is electoral fraud. The issue of vote rigging or what we term as ‘paper vote’ – the illegal interference with the polling process to increase the vote share of a favoured candidate and reduce that of the rival candidates, or both – could account for the huge gap in the figures of 2012 and 2016. This is made possible due to a probable bloated electoral roll that made it seem like more people did not vote, though in reality, turnout was not low. To think that an election has received low turnout, we must first compare eligible and registered voters to the actual number that turned out to vote. In 2016, there were 15,712,499 people on Ghana’s electoral register. Out of this, 10,781,609 in 271 of 275  constituencies voted on Election Day December 7, 2016, representing 68.62% of registered voters.
It is important to draw attention to the fact that this voter population of 15,712,499 is largely based on the 2010 population and housing census which reported Ghana’s population as at September 16, 2010 as 24,658,823 (Ghana Statistical Service, 2012). This register was also used for the 2012 presidential and parliamentary elections. In 2012, the voter population was 14,158,890 out of which 11,246,982 voted, representing 79% voter turnout. The 2012 elections also recorded the second highest turnout in the history of elections under the Fourth Republic. Figure 1 gives a visual impression of registered voters versus voter turnout from 1992 to date. From this, while registered voters continue rising, voter turnout has generally been inconsistent with a peak in 2012, after which it has taken a nosedive.
Figure 1: registered voters versus voter turnout 1992-2016
So why is there a 10% gap in voter turnout between 2012 and 2016 elections. Multiple answers are possible. One school of thought is that some voters have grown indifferent to politicians in Ghana and decided to withdraw from voting in 2016. Nevertheless, as explained earlier, it is difficult to sustain this argument generally without concrete proof in an open democracy like ours, as dissatisfied voters are more likely to vote against the non-performing government than to refrain from voting. Additionally, there are some individuals and groups in Ghana who register merely to have access to the voter’s identification card for private uses such as opening a bank account among others. Typical example is the Jehovah’s Witnesses (JW). They have their names in the voters register yet their belief bars them from voting; so definitely they are part of the absentee voters. However, the JWs have been on the electoral roll since 1992 without voting, therefore, we cannot simply ascribe this turnout difference to them even though they might definitely be part of it, perhaps marginally.
Another school of thought explaining the 10% shortfall in 2016 turnout is a possible tightening of the Electoral Commission’s voting procedures and improved checks and balances. This might have prevented the possibilities of double and/or over-voting from occurring substantially which could have swelled the turnout figures. While this may be a problematic argument, the overall improvement in the 2016 voting and results tallying procedures over that of 2012 cannot be denied.
Critics have argued that Ghana’s voters register relative to its current population is on the high side, given the voter population as a percentage of total adult population. In 2010, Ghana had a population of 24,658,823 with average annual growth rate estimate of 2.3%, putting the estimated total population in 2016 at about 28,263,513. With this, 55.60% of the current estimated total population are registered voters in Ghana. This is relatively high given the indicative adult population (18+) of 13,632,299 in the 2010 census. Judging from the 2010 census, the adult population was 55.28%. At an annual average growth rate of 2.3%, the eligible voter population is estimated to be in the region of 15,625,103 as at 2016, which translates into 0.56% less than the current number of eligible voters on the electoral register. This means the current voters register is 100.56% of the estimated current adult population, or there are 0.56% more persons on the voters register assuming the voter register has all eligible adults on it. This suggests that all eligible voters in Ghana are registered on the electoral roll, a fact that is rare.
Two factors are likely to account for this: First, it should be noted that the possible blowing up of the register may be traced to its original compilation in 2012. Second, the 2010 housing and population census was fraught with numerous administrative and enumeration challenges capable of affecting the final figures.
In 2012, 14,158,890 were registered to vote, representing 54.87% of the 2012 estimated total population of 25,806,173, and 99.25% of the estimated adult population of 14, 265,652. Elsewhere in Africa, electoral rolls hardly exceed 50% of total population and 80% of the adult population. For instance, in South Africa, with a relatively higher adult population than Ghana, the voters register, as a percentage of total population, is 47%. In Botswana, 36% of its population are on the electoral register representing 77% of eligible adult or voter population.
Comparison of votes gained or lost by NDC and NPP, 2012 and 2016
Change in NDC votes 2012-2016
Change in NPP votes 2012-2016
Generally, the 2016 elections recorded low numbers across all the regions for the two major political parties (NPP and NDC) when compared with their recorded figures for the 2012 general elections. However, while the NDC had net losses in all the ten regions in the 2016 elections, the NPP recorded modest gains in their votes across all the regions but Greater Accra and Eastern regions. The loss in NDC votes in the regions especially Brong Ahafo, Western, Central, Upper East and Upper West regions can conveniently be traced to the likelihood of floating voters. This is because, in these regions, as the NDC loses votes the NPP gains in similar proportions. For example in the Brong Ahafo region, the NDC lost 68,455 votes in 2016 compared with 2012 figures, while the NPP gained 61,238 in 2016 from 2012. Similarly, in the Upper West Region, NDC lost 18,651 votes in 2016 while the NPP gained 19,385 additional votes in 2016 from the 2012 figures. Again, in the Ashanti region, as the NDC lost 109,248 votes in 2016, their NPP counterpart got additional 111,136 votes in 2016. Operating on the assumption that majority of the people who voted in 2012 are the same people who voted in 2016, we can largely attribute the loss in NDC votes in these regions to the gains in NPP votes in those regions.
However, there is an interesting twist to this given the voter numbers in the Western, Eastern, Northern, Volta and Greater Accra where the NDC lost heavily. Figures suggest the defeat of the NDC did not significantly reflect in the votes gained by the NPP. In the Eastern and Greater Accra regions, for instance, the NPP also lost votes in 2016. A number of competing explanations can be put forward in attempting to account for the missing voters in 2016, particularly the close to one million votes that the NDC lost between 2012 and 2016. The first plausible explanation is apathy in NDC strongholds, ghost names (dead, foreigners, double registration) or tightening of th electoral process that prevented rigging.
In democracies, voting during elections is one way to measure political participation and gives a signal of the public’s approval and support or otherwise specifically of the government of the day, but more generally of the system of governance. Voter apathy or eligible voters not coming out to vote on Election Day is a regressive revolt against the broader interest and aspirations of democracy. It is also signals that people are voting with their feet to register their discontent with government and the general direction of governance of their country. The lack of interest among eligible voters to come out and cast their votes especially in the strong-hold of the NDC and the swing regions was as a result of the perceived lack of development in the region or poor living conditions. A party in government has the enormous responsibility of ensuring that development is equitably distributed across the country. However, if a region or group feels they don’t have their fair share of the development projects, they tend to vote against the ruling party.
The effect of this is twofold: First, people will not turn out to vote at all, and second, the ruling party (in this case the NDC) is more likely to have lost more votes as a result of voter apathy than the NPP. From the above data, it can be argued that the NDC lost the 2016 elections because of apathy in its strongholds, especially the Volta Region where the party lost 121,436 votes in 2016. What makes this important to probe further is the fact that the loss of votes by the NDC in the Volta region did not translate into gains for the NPP. Thus, it is tempting to conclude that potential NDC voters just decided not to vote because they were dissatisfied with the performance of the NDC government. At the same time, while they failed to vote massively for the NDC, they did not also vote for the NPP.
For instance in the Ketu South constituency, the NDC got 81,880 votes in 2012 but that figure was slashed to 50,648 in 2016 representing 38% lost between the two elections. Yet, the over 30,000 people in the Ketu South who did not vote for the NDC did not vote for the NPP either. The NPP got 5,165 and 5,885 votes in 2012 and 2016 respectively. Similar thing occurred in Ho Central constituency where the NDC got 62,363 votes in 2012 but that figure was reduced to only 24,363 in 2016.
This relates to two of the proposed explanations to the missing vote – either those voters genuinely exist but failed to vote in 2016 as a sign of dissatisfaction with the government’s performance, or they are ghost names or foreigners who could not find their way into the ballot box in 2016 due to the strengthening of the electoral processes. Whatever the true cause, such significant reduction in votes in a four-year period should be a source of worry to the NDC party considering the high approval rating the party has enjoyed in Volta Region generally and specific constituencies like the Ketu South since 1992.
However, the NDC did not only lose significant votes in the Volta Region but also in other regions such as Greater Accra, Western and Ashanti. Even the NDC lost more votes in the Western and Greater Accra than the Volta Region. Though the NPP also lost about 20,000 votes in the Greater Accra region in 2016, the NDC lost over 213,000 votes in Greater Accra. Those voters did not necessarily vote for the NPP or other smaller parties. Again, in the Western Region the NDC lost over 126,000 votes in 2016 while the NPP got additional 57,000 votes. Whichever way you look at it, some votes in 2016 still remain unaccounted for from the 2012 turnout figures. An isolated reason which is very often not considered in such analysis is death – though there is little statistical data on the death rate in Ghana to buttress this point.
The 2016 general elections has by far been one of the fairest elections in Ghana since 1992. This can largely be attributed to the reforms instituted by the Electoral Commission before the elections. The results of the elections suggest about five million voters did not turn out to vote, a phenomenon which suggest some form of anomaly occurred during the compilation of the electoral register. However, the number of registered voters as a percentage of the adult population +18 is not very abnormal compared to other democracies on the continent. In other words, although there are challenges with the electoral register in Ghana, it is premature to narrow the challenge to a bloated register, especially as Ghana has a challenge with data generation or gathering.
However, from the foregoing analysis, it is suggestive the nearly five million voters who did not vote on Election Day can be accounted for by apathy and electoral fraud such as cases of multiple registration.
[The authors acknowledge Mr Tony Dogbe, Principal Consultant, PDA, and Mr Clement Sefa-Nyarko Research Manager, PDA, for providing useful information and comments on earlier drafts of this article.]
- Alexander Afram is Assistant Program Coordinator – Research and Kafui Tsekpo is Coordinator Advocacy and Communications at Participatory Development, a social development enterprise with over 15 years of organisational experience in research, facilitation, organisational development, policy analysis and advocacy.
GSS (2012). 2010 Population and Housing Census Final Results; available at http://www.statsghana.gov.gh/docfiles/2010phc/2010_POPULATION_AND_HOUSIN…
World Population Review (2016). Botswana population 2016. http://worldpopulationreview.com/countries/botswana-population/
The Economist Intelligence Unit (2016). Democracy Index 2015: Democracy in an age of anxiety. http://www.eiu.com/public/topical_report.aspx?campaignid=DemocracyIndex2015
The figures here as declared by the Electoral Commission is based on 271 constituencies out of 275. This is because at the time of the declaration on December 9, results from 4 constituencies were being contested and were excluded from the declaration and final analysis