This paper explores the ideal that election observation and electoral reforms have a bearing on election related violence in most SADC states. It also explores the idea that countries that resisted electoral reforms and interfered with election observation continue to experience electoral difficulties and sometimes widespread violence. The paper relied on data of various election observation reports, published and unpublished literature. The next section looks at the theoretical perspectives of election observation and electoral violence. Building on this, the authors then examined the general political impact of election observation in SADC in the section that follows. The subsequent section of the article looks at how different institutions have a bearing on electoral violence and show how they can be reformed to deepen democratisation in the SADC region.

Election observation and electoral violence

Violence in elections is often used to intimidate voters and manipulate or rig elections. Electoral violence is a very broad concept and includes activities not limited to physical assault, disruption of political rallies, issuing of threats and vote rigging. Albert (2007:133) broadly defines electoral violence as ‘all forms of organized acts or threats– physical, psychological, and structural aimed at intimidating, harming, blackmailing a political stakeholder before, during and after an election with a view to determining, delaying, or otherwise influencing an electoral process’. Oftentimes, observers from membership organizations such as United Nations (UN) and SADC are often reluctant to be harsh in their criticism of a member state regarding flaws  in  an  election  and hence often condone electoral flaws (Carothers 1997: 25). But the point here is that observers of all sorts suggest important reforms that help to move the country forward.

For Khadiagala (2010: 16), electoral violence is the outcome of events and circumstances that emanate from broader political conflicts, particularly in societies that are beset by ethnic, communal, and sectarian fissures. Electoral violence also results from structural deficiencies, particularly the institutional rules that govern the conduct of elections. As Khadiagala observes, ‘electoral violence is a consequence of imperfect electoral rules, imperfections that allow some parties to manipulate elections through electoral fraud, vote buying and rigging (2010:pg17). Our observation is that election observation helps to bring the necessary reforms that reduce the imperfections and reduce chances of manipulation that trigger protests.

The implication therefore is that electoral violence can be alleviated by putting in place electoral rules and reforms that would curtail chances of manipulation and electoral fraud. To this end, election observation has a fundamental role to ensure that electoral rules are implemented, that elections are conducted in a peaceful atmosphere, and that reforms are carried out. Our argument is that election observation supplies recommendations that could provide directions towards reforms, thus contribute to reductions in the levels of electoral violence. Election observation has risen to prominence as a democratic assistance process in the 1980s.

Carothers (1997: 2) argues that election observation is the best-established, most visible, and often best-funded type of democracy-related assistance. The importance of observation of elections in ensuring credible elections and exposing irregularities in an electoral cycle cannot be overemphasized. According to Kelley (2009), election monitoring helps to uphold electoral norms and to report on the quality of elections. In this way, observers partly ensure that laws and regulations governing elections are adhered to by all parties and candidates. The presence of observers and the international media helps to curtail election rigging, promote better understanding between the contending parties and may bring about electoral reforms, a combination that helps to quell electoral violence. The involvement of international observers may also convince skeptical opposition politicians that competing in the elections, is preferable to engaging in civil disobedience or violence (Carothers, 1997:4). Central to this argument is the idea that the presence of international observers may avert state sponsored violence on opposition candidates and supporters. Awareness of the fact that the world is watching may influence not only the potential victims of intimidation but also potential intimidators (Mair, 1997). In his assessment of the 2000 presidential elections in Zimbabwe, Laakso (2002:21) stated that the EU concluded that, 'the presence of a substantial number of international observers throughout the country was important in reducing tensions and calming conflict at a local level’. More importantly, the argument is that reportage by election observers identify areas for reforms and helps to move the country forward in the democratization process.

It is acknowledged that election monitoring can fail to curb or reduce violence in protracted societies. As Carothers (1997) argues, “foreign observers cannot force profoundly polarized political factions to cooperate with one another. In deeply divided societies, averting electoral violence may prove a difficult undertaking where a long time ruling party has high chances of losing power. Carothers states that they cannot counter the deeply antidemocratic instincts of strongman intent on holding on to power. The case of the 2008 Zimbabwean presidential election runoff is illustrative on this regard. A highly volatile election environment may often lead to restraint on the part of observers’ reports which ultimately distorts their findings. In a tension filled environment where adverse reports can lead to violence, observers often “soft-pedal their findings” (Carothers, 1997: 25). According to Kelley (2009: 765), election observation can help curtail electoral violence, it can also work to promote endorsement of less credible elections by concealing the truth. Pre-electoral violence and unstable environment may discourage observers from delegitimizing elections for fear of post-electoral violence. Thus, incumbent-dominated pre-election violence may dissuade monitors from denouncing the elections (Kelley, 2009: 9). The result is that not only would elections be fraudulent, but they may also cast some doubts and aspersions on the overall election observation process.

THE POLITICAL IMPACT OF ELECTORAL OBSERVATION IN SADC

Since the end of the Cold War in 1989 and the collapse of apartheid in 1994, the SADC region has held numerous elections and all of them were observed by diverse election observer groups. Groups that commonly observe elections in the southern Africa region include national nongovernmental organizations (NGOs), the SADC Election Observer Mission (SADC SEOM), SADC Parliamentary Forum (SADC PF), Electoral Commissioners Forum (ECF) and Electoral Institute for the Sustainability of Democracy in Africa (EISA). International bodies such as the European Union, the European Parliament and the AU also observe elections. All these election observation groups share the commonality of assessing the credibility of elections in terms of the electoral laws governing elections, demarcation of constituency boundaries, management of registration, acquisition of electoral material, and policing and security of the poll. However, SADC places more emphasis on collecting data for dispute resolution in the post-election period. To meet that objective, SADC allegedly produces two reports: one for the public which is general and friendly to the hosting country; and a  second one which is detailed, secretive, and constitutes inter-government communication meant for targeted political intervention.

Election observation has become a prominent feature of African politics and an important mechanism, to induce sustained democratisation. The African Union (AU) and the SADC have since resolved to observe all national elections within their respective mandates. Election observation gathered material critical for pronouncing on the legitimacy of governments, on political intervention options and on political reforms. Early observations linked elections with widespread electoral violence and therefore, with contested state legitimacy as there were observable human rights abuses, electoral malpractices, low level conflict, and sometimes civil war. Election observation has also helped to classify elections into credible and non-credible elections, and to legitimise or delegitimise certain players, and to guide mediation that helped to end civil wars, resolve political tensions, and aid political reforms.

Election observation plays the important role of helping to distinguish between credible elections from non-credible ones. MISA’s Outside the Ballot Box (2004) constitutes one of the first efforts to analyse, document and publish information on election observation and its political impact. In its introduction, outside the Ballot Box (2004: 4) observed that ‘elections represent an important dimension in the efforts towards democratic consolidation in any country, not least in African countries. Many African elections continue to fail human rights and democratic tests, and have often served as the casus belli for low intensity conflict or outright war’. This important early observation, linked African elections with human rights abuses, electoral malpractices of all sorts and low level conflict and war. Such an observation partly justified the need for extensive election observation. However, it must be admitted that low level conflicts and wars have largely ended in the SADC region (including in Angola, South Africa, Mozambique, Namibia and the DRC) and that credible elections are held in a number of countries. However, human rights abuses and electoral malpractices still occur in a few other countries.

Election observation has the other role of legitimising or delegitimising electoral contenders. For instance, De Brito (2009: 35) notes about Angola that, ‘the September 1992 elections were declared free and fair, but this was not accepted by Savimbi. UNITA withdrew its former soldiers from the united army (FAA) and clashes began to occur between Unita‘s forces and those of the government’. By declaring the 1992 Angolan election free and fair, election observation helped to delegitimise Unita and to rally the international community behind the Angolan MPLA government that was regarded as the legitimate government. Even the United States changed its posture and accepted the elected MPLA government as legitimate, effectively abandoning Unita to its own devices. Later that year, the UN Security Council imposed an arms and  fuel embargo on Unita. Thus, election observation in this case helped to provide legitimacy indicators and foreclosed further violence between the contestants.

Election observation showed that Angola, South Africa, DRC, Namibia, Tanzania and Zambia have joined Botswana in the characterisation of peaceful elections and some of these states are able to work closely with civil society. This improvement was very dramatic in South Africa and shows that even high incidence of violence can be brought down quickly through observation and political reform. Booysen and Masterson (2009: 416) write that, “according to the IEC statistics, South African elections have become increasingly peaceful. Election 2004 was the least conflictual of the three elections from 1994 to 2004”. In its preliminary post-2009 election statements, the IEC suggested that the trend of increasingly peaceful elections was continued into election 2009. In the national and provincial elections of 1994, the IEC received 3,594 official complaints, and more than 1,000 people were killed in election-related violence. In the 1999 national and provincial elections, the number of official complaints was 1,114 and the number of people killed was less than 100. In 2004, the IEC received just 253 complaints and claimed no election-related deaths”. Similarly, elections campaigns are peaceful in Zambia where the police were the only ones accused of favouring the ruling party, but have since adopted a neutral position. However, Ilona Tip et al. (2009: 603) noted that “most stakeholders agreed that for the 2006 and 2008 elections, the police managed the process fairly and most proposed meetings and rallies were allowed to take place”. This trend continued during the 2011 tripartite election. Election observation also showed that the DRC has also managed to minimise violence in its electoral politics, and to postpone elections in areas experiencing violence and to work closely with civil society (Kadima and Tshiyoyo, 2009).

Malawi and Zanzibar also made progress in arresting electoral violence. For instance, in the latter, the ruling CCM and CUF negotiated the Muakafa 11 Accord that was signed in 2000 leading to significant reforms of electoral laws, of judiciary, management of the electoral process and installation of neutral mechanisms.

Masterson (2009: 512) noted that when the Tanzania and Zanzibar elections were de-linked in 2005 due to the death of a vice presidential candidate that necessitated postponement of elections on the mainland, an inordinately large number of international observers descended on Zanzibar where they observed insignificant scale of violence and orderly processes. As a way of addressing electoral violence, South Africa’s Electoral Act of 1998 Schedule 2 (Section 99) requires participating political parties to sign to a binding code, which requires all party candidates to uphold everyone’s rights by, publicly condemning any action that undermines the conduct of free and fair elections; and by accepting the results of an election or challenge the result in a South African court (quoted in Booysen and Masterson, 2009: 412-413).

Namibia and Lesotho established similar electoral codes of conduct (Blaauw and Letsholo 2009: 370). Given the significant reduction in electoral violence in countries that adopted codes of conduct, a common code of election campaigning for SADC could minimise electoral violence.

In contrast, some countries experienced regression in many fronts, and various electoral observation groups cast doubts on legitimacy of their elections. This was most evident in Zimbabwe in the presidential run-off elections in 2008, in Lesotho after the 2007 election and in Seychelles after the 2011 election. ECF (2008) questioned the legitimacy of the presidential run-off election in Zimbabwe on the basis of undue delay in releasing the 29 March 2008 presidential elections results and consequent tension and uncertainties, widespread violence against the opposition and civil society, media coverage biased in favour of ZANU PF, and arrests of opposition politicians and elections officials. 

In the eyes of the national, regional and international election observers, the security situation in Zimbabwe had collapsed and the environment was not conducive to the holding of credible elections. Even SEOM, the official SADC observer group raised concerns on the Zimbabwean presidential run-off in a press brief dated 25 June 2008 by Joao Marcos Barrica, head of SEOM appealing to supporters of political parties and candidates to restrain from all forms of violence. SADC also urged the law enforcement agencies to ensure that there is law and order in the country.

This acknowledgement of widespread violence prompted SADC to propose the establishment of a coalition government in Zimbabwe in order to avert open conflict and continued suffering of people. During the 2008 elections, Zimbabwe could have benefited from a code of conduct for political parties similar to the ones of Namibia, Tanzania and South Africa. The observers of the 2013 Zimbabwe elections reported them to be free and peaceful, restraining from pronouncing on fairness due to the problems associated with the registration and the high number of assisted voters.

Similarly, Lesotho in 2007 experienced challenges that generated political instability and delegitimized the electoral process (Musanhu, 2009: 151). The central question was whether the electoral commission should have recognized ‘illegal political alliances’ which were aimed at subverting the electoral law itself? The Lesotho electoral commission recognized the alliances and rewarded them accordingly, plunging the country into a post-election crisis. However, the challenges facing Lesotho are bigger, including the fact that the 2005 population census was not followed by a constituency delimitation exercise. According to Musanhu (2009: 167) the country went into the 2007 snap election without new constituency boundaries even after a population census had been conducted in 2005. In addition, the voter registration period was too short, and the voters’ roll was updated inadequately. Furthermore, these challenges got

spur from a tradition where Lesotho had a culture of rejecting election outcomes (Musanhu, 2009: 183). In this sense, Lesotho might benefit from a cultural re-orientation and political reform under the auspices of SADC so that election results are accepted as a norm.

Similarly, the situation in Seychelles in 2011 qualified that country for de-legitimation. For instance, local stakeholders raised concerns over disqualification of an independent candidate, media, civil service and security apparatus partiality, lack of copies of voters’ roll and violence.

Interestingly, SADC (2011: 3) in Seychelles observed that “although some of the concerns raised were pertinent, they are nonetheless not of such magnitude as to affect the credibility of the overall electoral process”. This conclusion from SEOM is not surprising coming from a team of short term observers, who are in the field for brief periods of time and also representing a regional organisation that downplays (in public) negative factors observed in a sister country. 

Observers’ reports from Seychelles indicated the need for a code of conduct for political parties similar to the ones in South Africa and Tanzania. In all the above instances where observation qualified the electoral process, a case is made for continued election observation. SADC could make it categorically clear that widespread violence creates an atmosphere that is not conducive to the holding of credible, legitimate and peaceful elections, and that the elections should either be postponed or not accepted as credible in cases of widespread violence.  

Election observation and election management bodies (EMBs)

Election observation has pointed out that the general competence and independence of the election management body are major contributory factors to peaceful elections. Booysen and Masterson (2009: 427) observed that the acceptance of the outcome of elections in democratic South Africa has been unequivocal to a considerable extent due to the credible electoral commission operations. Of course, acceptance of outcomes is relatively guaranteed in conditions where the ruling ANC has retained both widespread popular legitimacy and commanding majorities. Opposition parties in the first four democratic elections have not come anywhere close to posing electoral threats to the ANC.

Thus, the competence and impartiality of the South Africa IEC and the dominance of the ANC makes the acceptance of results easier. However, South Africa also has superior election registration mechanisms which are highly welcomed by all stake holders. Booysen and Masterson (2009: 406) observed that, the 2004 and 2009 voter registration processes remained largely similar to that of 1999, although voters who had previously registered were not required  to  re-register.  In  addition,  during  the registration periods returning voters were able to verify their details on the voters’ roll without having to complete the registration process again. It was possible for voters to call a toll-free telephone number or visit the IEC’s website to check whether and where they registered.

Therefore, the electoral reforms leading to competence of the IEC contributed significantly to the high credibility of the South Africa electoral process that prompted the EU election observer mission to conclude that there was no need for it to observe that country’s elections anymore. In contrast, Musuva (2009: 223) observed this about Malawi Electoral Commission (MEC). While the first Electoral Commission that presided over the 1994 elections was lauded for its performance, subsequent commissions have put the institution in disrepute. The handling of the 1999 and 2004 elections, particularly in the pre-election preparation and post-election phases of the electoral process, impacted negatively on the efficiency and capacity of the Malawi Electoral Commission (MEC) in discharging its responsibilities. The MEC that oversaw the 2009 elections displayed more professionalism and transparency in the management of the electoral process and won the confidence of the majority of the electoral stakeholders. The in-consistencies in the capacities of the EBMs should be addressed as they pose danger that could incite political violence. Intimacy between the governments and the electoral bodies has emerged as an issue of scholarly concern in several countries. Such high intimacy negatively affected the independence of the electoral management body. Musuva (2009: 235) on Malawi added that, the period following the announcement of the 1999 and 2004 election results was marked by violent protests, loss of life and a host of litigation. Widespread allegations of vote-buying during the campaign period in the run-up to the 2004 elections were not investigated by the MEC. Campaign violence and voter intimidation in the 2004 and 2009 elections were less prevalent than the 1999 and 1994 elections. However, the MEC was temporarily closed for four months (December 2010-April 2011) at the instruction of Head of State. Thus, Malawi appears to be failing to consistently improve the competence of its electoral management body.

Election scholars and observers agree that lack of independence of the EMB seemed to be a common feature among countries that experienced electoral difficulties. Election scholars such as Booysen and Toulou (2009: 643) who rely extensively on election observers reports note that Zimbabwe’s electoral challenges include the closeness between the Zimbabwe Electoral Commission (ZEC) and the ruling ZANU-PF party.

The changes in election management mentioned above left some issues unaddressed regarding ZEC’s autonomy as election organizer, particularly its close relationship with ZANU-PF and the government. Thus, the ZEC’s role in the electoral process appears to be more supervisory than policy-making and effective management. Evidence of this institution’s shortcomings were inter alia its failure to proclaim the results of the presidential  vote  immediately after the 29 March 2008 Harmonised Elections, to challenge the demand for a recount of votes expressed by ZANU-PF, and to postpone the June 2008 presidential run-off owing to the climate of violence, fear and intimidation resulting in the decision by Morgan Tsvangirai to pull out of the electoral process.

Such intimate relations between the state and Electoral Management Bodies (EMBs) worked against impartiality and objectivity. Thus, the intimacy between the ZEC and the ZANU-PF government was noted as a crucial stumbling block towards holding credible elections in Zimbabwe. The same closeness was observed in a number of SADC member states. Common SADC standards could guide operations and the independence of the Electoral Management Bodies. A country that takes these observations seriously and reforms its electoral laws is likely to suffer less electoral violence in the future. It is those countries that ignore recommendations from international election observers that continue to experience heightened electoral tensions and violence.

Election observers noted that some EMBs were poorly funded and were completely unprepared to conduct free and fair elections. Scholars have observed that the capacity of an EMB is an important element in the holding of credible elections. Such observations indicate what is wrong within the election administration machinery, and provide signals to the state concerned and to donors who support elections that more funding is required by the EMBs. Thus, election observation identifies the crucial gaps that need to be plugged to prevent future electoral violence.

Election observation and state security agencies

Election observation has noted that biased security agencies are now less common in some SADC member states. They noted that Swaziland and Zimbabwe were the exceptions in SADC, where the security agencies arrested opposition candidates, restricted the movement of civil society groups and sometimes condoned violence against opponents of the ruling elite. Election observers note that the intimacy between the ruling political elite, and the security agencies in these two countries, has sometimes compromised the security of the election. There are two points here. The first is that other countries in SADC have moved on, by making their security forces less of a threat to opposition parties and candidates. The second point is that only two countries still lagged behind in this regard. By celebrating achievements and exposing gaps, election observers and election scholars help to pinpoint where problems are and it becomes easier to plug them and prevent electoral violence in the future.

Constituency delimitation

Election observation has pointed  out  that  constituency delimitation and unequal representation are still serious problems facing some countries in the SADC region, exposing them to the dangers of electoral violence. Such countries include Malawi, Tanzania, Zimbabwe and Lesotho. For instance, poor funding in Malawi affected the demarcation of constituencies. Musuva (2009:246) noted that for the 1999 elections, the MEC did not invoke the principle of constituencies containing an approximately equal number of voters who are eligible to register, subject only to population, communication and geography. It failed to draw constituencies that crossed regional or district areas and merely subdivided large constituencies rather than re-demarcating all constituencies with many districts. As a consequence, constituency populations were very uneven. For example, in Nkhata Bay East there were 3,407 voters and in Machinga North there were 59,494 voters. Another obstacle in the demarcation process in Malawi was that no census had been conducted since 1987. This lack of accurate population data greatly hampered the commission in carrying out its mandate.

The phenomenon of unequal constituencies is common in several SADC states as shown below. In addition, the failure to make the demarcation of constituencies a consultative process, posed the danger of the process becoming cultural insensitive, and had the potential to incite electoral violence. In Tanzania for instance, constituency delimitation posed serious challenges. Masterson (2010: 523) noted that one of the biggest concerns as regards the delimitation of constituencies in Tanzania relates to the large discrepancies between the size of constituencies in Zanzibar and those on the mainland. Zanzibar’s population of little more than 1 million persons entitles it to 50 representatives in the Union National Assembly (a ratio of 1 seat : 20 000 citizens), while the nearly 37 million citizens in Tanzania in 2005 were split into 232 constituencies, giving the mainland a representative ratio of 1 seat : 160,000 citizens. This weighting gives Zanzibar citizens a disproportionate amount of influence in Union matters in comparison with citizens on the mainland, highlighting one of the unusual accommodations of Zanzibar’s status within the Union.

Thus, unequal representation and unequal delimitation of constituencies create tensions that could easily incite protests and lead to electoral violence. As long as the union was structured in manners that favour Zanzibar, sentiments of unequal representation expose Tanzania to electoral violence.

Similarly, Zimbabwe has not yet resolved issues surrounding constituency delimitation that compromised peace during the Harmonized Elections of 2008. Booysen and Toulou (2009: 645) noted concerns by Zimbabwean civil society groups that suggested the possibility of gerrymandering. “In December 2004, the Delimitation Commission recommended a reduction of the number of constituencies by one each in Harare, Bulawayo and Matebeleland South, and an increase by one each in Mashonaland East and Manicaland. These concerns were justified by the fact that the decrease in the number of constituencies affected mostly urban areas which had been known for their support to the opposition. Harare Metropolitan was cited as an example where a decrease of constituencies occurred in spite of an increase in its voting population. On the other hand, the areas which enjoyed an increase in the number of constituencies coincided with the rural areas, which traditionally supported the ruling ZANU-PF. Further, domestic elections observers noted that the Delimitation Commission completed its report in December 2004, well before the final voters’ roll had been compiled. Thus, Zimbabwe is still exposed to electoral violence.

In addition, Booysen and Toulou (2009: 645) noted that Zimbabwean opposition parties and civil society organisations made representations to the Pan African Parliament Election Observer Mission regarding the fact that the delimitation process had not been transparent enough and contrary to the legal requirement of it being a consultative process. They quoted Amendment Act No.18 which states that, the President shall cause the report to be laid before the Parliament within the next seven days after he has received it (section 13.d). “Neither of the commission’s preliminary and final reports was presented in parliament after being sent to the president. Only one copy of the reports was made available and few parliamentarians, let alone the general public, were able to scrutinize it.

Thus, constituency delimitation in Zimbabwe was allegedly secretive and violated constitutional provisions which called for consultation. This added to the mistrust that creates room for electoral violence. In contrast, Angola, the DRC, Mozambique and South Africa seem to have found a working formula to make unequal representation an irrelevant electoral factor. Kadima and Tshiyoyo (2009: 108) summarized the DRC formula for allocating seats per electoral constituency within each province and the National Assembly on the basis of actual numbers of registered voters in each constituency and province. There is also a fixed electoral quotient for the National Assembly which is equivalent to the total number of registered voters in the country, divided by the total number of seats to be filled in the National Assembly. Thus, there exists a better formula out there that could help countries that face serious challenges associated with unequal representation and constituency delimitation.

In Angola, while constituency delimitation is not really a challenge, unequal representation is.  De Brito (2009:40) observed that regardless of the population or number of voters in a province, all 18 provinces are allocated five seats each. For instance, Luanda, with the highest number of registered voters (1,971,963), has the same number of dedicated provincial representatives in parliament as Bengo, with only 108,758 registered voters. These seats are allocated provincially according to the population each party or coalition received in that province (in an application of the D’Hondt system).

Thus, while the principle of equal representation is not being met, the use of proportional representation eliminated the need for constituency delimitation. In contrast, Mozambique, Namibia and South Africa eliminated problems associated with constituency delimitation through the proportional representation system, making demarcation of constituencies redundant. In addition, the responsibility of delimiting the entire geography of South Africa into voting districts is the responsibility of the Delimitation Directorate (DD), using a satellite mapping system and census information available to do so. The DD, which is part of the IEC, reviews the districts prior to each election. Since 2001, GPS technology has been deployed to accurately record the positions of voting stations. The 14,650 voting districts of 1999 increased to 16,966 in 2004, and 19,726 in 2009.

South Africa may be required to share its technological advances with the other SADC states that are willing to do so, to help map and locate voting districts. In addition, proportional representation appears to have elements that help to depoliticise issues surrounding constituency demarcation and eases tension within the political society. Thus, the PR system partly eliminates problems associated with constituency delimitation and unequal representation.

Election observation and the media

Exposing state journalists to electoral violence were widespread complaints about unbalanced reporting of state-owned media, cutting across many SADC states- from Angola to Botswana, Lesotho, Madagascar, Malawi, Mozambique, Namibia, Seychelles, Tanzania, Zambia and Zimbabwe. Several authors in Outside the Ballot Box (2004) observed that governments control the public media that are explicitly biased in favour of the ruling party in Botswana, Mozambique, Lesotho, Namibia, Malawi, Tanzania and Zimbabwe. The same concerns were raised in the 2009, 2010 and 2011 elections across the region. This implies that freeing the media (particularly state media) remains a serious challenge that limits the full realization of democracy in the SADC region. In Zambia, Ilona Tip et al. (2009) noted a general improvement in the environment in which the media conducts its duties in the run-up to elections. However, incidents of harassment of journalists were reported during the 2008 presidential elections in Zambia.

Ironically, SADC states were attempting to strengthen their controls on the private media, an emerging situation observed in Botswana with the passing of the Media Act of 2009 that has largely been abandoned. Similarly, Musanhu (2009: 162) observed about Lesotho that reports were received from the Media Institute of Southern Africa (MISA) that a number of journalists had received death threats from members of the ruling party over coverage of issues which tainted the image of the ruling party. These threats or perceptions are reported to have led to a certain

amount of self-censorship on the part of the media. These SADC states (Botswana, Malawi, Mozambique, Namibia, Seychelles, Zambia and Zimbabwe) are being challenged to loosen their grip over the state media and to lessen threats against journalists. In contrast, only South Africa and the DRC in SADC seem to have overcome complaints about state media and therefore reduced threats against state journalists. According to Booysen and Masterson (2009: 414), all public media communications in South Africa are regulated by the Independent Communications Authority of South Africa (ICASA). ICASA is responsible for publishing a code of conduct for public broadcasts during an election period, and is also responsible for arbitrating and resolving disputes that occur regarding infringements of regulations governing the public media and electoral matters. ICASA is managed by a board of seven councillors, who are appointed by the president, in consultation with the National Assembly. According to ICASA regulations, paid political campaign advertising and announcements during an election period must be submitted to ICASA 96 h prior to airing or publication, and must conform to standards that do not contravene the constitution, Section 9 of the Electoral Code or Broadcasting Act.

Setting up an independent functional regulatory authority over media reporting on elections seems to be the way to go for the rest of the SADC countries. This is expected to instil a spirit of professionalism and lessen threats against journalists who cover elections.   

This paper sought to establish links between election observation, electoral violence and political reforms. This paper has shown that as election observation became widespread and generated extensive reports, electoral reforms followed and election related violence declined in most SADC states. Through recommendations from election observer missions, several countries within SADC (including Namibia, South Africa, and Tanzania) have negotiated and signed binding codes meant to reform the electoral space and this contributed to the prevention of electoral violence. As a result of implementing the codes of reforms in the presence of subsequent election observers, these countries witnessed a dramatic decline in electoral violence. In contrast, a few countries that resisted electoral reforms and interfered with election observation (such as, Lesotho, Seychelles, and Zimbabwe) continue to experience electoral difficulties and sometimes widespread violence. In this regard, SADC could provide leadership through appropriate and timely political intervention that is critical for these countries to move forward in consolidating democracy. SADC is not yet ready to enforce common electoral standards.

The paper has also shown that election observation helped to categorise elections, to legitimise some electoral players and delegitimise  others,  to  point  out  areas needing reforms and to expose hotspots that exposed countries to electoral violence. The authors have also shown that while most countries in the region have positively reformed in many areas, liberalising the media has been the common problem area facing all SADC states. The questions is on the media, that almost all states (except South Africa) in the region have regressed. Thus, SADC states, particularly those whose ruling parties face defeat at the polls, are still exposed to the dangers of electoral violence and therefore need continued election observation.

The authors have not declared any conflict of interests.