MOZAMBIQUE ’S 2004 GENERAL ELECTIONS

Inge Ruigrok
Freelance journalist based in Portugal and a PhD candidate at the Free University Amsterdam, the  Netherlands.

Published in African Security Review Vol 14 No 1, 2005

The 2004 general elections

FRELIMO was the clear victor of both the presidential and the parliamentary elections. Armando Guebuza, who won 63.7 % of the vote, received a mandate to become the new president of Mozambique, while Afonso Dhlakama received only 31.7%. In the legislative body, FRELIMO has taken 160 seats (62.03% of votes), and RENAMO 90 seats (28.73% of votes). Compared to 1999, RENAMO lost 27 seats to FRELIMO.

As the results show (see results on next page), Mozambique remains a de facto twoparty -parliamentary -system. None of the 18 smaller parties received the share of the vote (5%) that was required to gain a seat in parliament. Even Raúl Domingos, who drew support from donors who hoped he and his Party for Peace, Democracy and Development (PDD) might create the long-hoped-for ‘third force’ in Mozambican politics, was unable to challenge the two giants. Even though the PDD showed itself to be active and organised during the election campaign, it received only 2% of votes. This means that Domingos lost the seat that he had held as an Independent in -parliament. Furthermore, as a presidential candidate Domingos won only 2.73% of votes, far behind Guebuza -and Dhlakama.

Presidential election results

Parliamentary election results

Whereas RENAMO had a majority in five provinces in 1994 and six provinces in 1999, Mozambique’s largest opposition party achieved a majority in only two provinces in the 2004 election. The number of blank votes was substantially reduced this year and the number of invalid votes was lower. FRELIMO won the two seats for Mozambicans in Europe and in Africa.

The Carter Center3 expressed concern about a number of irregularities that had been observed during the provincial tabulation. These included summary sheet results (editais) that lacked credibility, problems with the tabulation software, mismatched numbers of polling stations and editais, and mistrust among political party representatives in the provincial Technical Secretariat for Electoral Administration (STAE) structures. In a number of cases Carter Center observers found editais with unrealistically high voter turnout, including instances of polling stations in Niassa and Tete recording a 100% turnout and more than 90% support for FRELIMO. Given the low turnout nationwide, the Carter Center said, ballot boxes appeared to have been stuffed at polling stations in the Tete districts of Changara, Chifunde and Tsangano, as well as in the Niassa districts of Metarica and Marrupa, and in the Gaza district of Chicualacuala. This means that ballot box stuffing or falsification of the editais seemed to have been more common in the December 2004 elections than in 1999. A number of other observers, including the Electoral Observatory, a coalition of seven prominent Mozambican civil society groups, also cited these irregularities.

Nonetheless, it should be emphasised that all observers confirmed that the malpractice was not on a sufficient scale to account for the FRELIMO landslide victory. They all found that 85% or more of polling stations -functioned correctly.

RENAMO claimed that the attempt at fraud had started during the electoral registration earlier this year, with most of the population being excluded in the zones where the opposition has most influence. RENAMO also claimed that polling station presiding officers received orders to introduce ballot papers in favour of FRELIMO and Guebuza and at the same time to invalidate ballot papers for RENAMO and its candidate by making ink blots on the ballot papers.

At a press conference on 4 January 2005 the National Elections Commission (CNE) admitted two of RENAMO’s complaints: that 1 400 editais favourable to RENAMO had been stolen and thus were not included in the final results (5% of the total), and that there was ballot box stuffing in Tete province.4 In making its various corrections, the CNE confirmed that it had given an extra parliamentary seat to RENAMO in Zambézia, the province where the editais had disappeared. The CNE rejected all other allegations lodged by RENAMO, however.

Following the CNE’s announcement of results, Afonso Dhlakama announced that RENAMO did not recognise the results of the election. He said that no member of RENAMO who had been elected to parliament would take his or her seat, and that he would not take his place on the Council of State. RENAMO then took an appeal to the Constitutional Council, which is responsible for validating the results.

On 17 January 2005 the Constitutional Council announced that it had rejected RENAMO’s application on technical grounds. First, the council ruled that the application had missed the legally required deadline – this was denied by RENAMO. Second, the council noted that the application received from RENAMO was substantially different from the application originally made to the CNE, and that as a body of appeal the council had no jurisdiction over a new request. This second reason for refusal – one that seems wholly justifiable on legal grounds – was, ironically, the consequence of RENAMO’s attempt to make its demands less ambitious before -approaching the Constitutional Council. Whereas RENAMO had earlier rejected the results outright, the party’s application to the council demanded only ‘the correction of all irregularities’: namely the alleged instances of ballot stuffing, and the exclusion of certain polling stations from the final count.

On 20 January 2005 the Constitutional Council validated the election results, paving the way for the new parliament to convene and for Guebuza to assume the presidency. For his part, and in a somewhat anticipated fashion, Dhlakama reconsidered RENAMO’s threats to boycott parliament, and ‘for the sake of peace and stability’ as he put it, RENAMO took its seats in the assembly of the republic. On 31 January all 250 members of parliament who had been elected in the general elections were sworn in and President Joaquim Chissano handed over the presidency on 2 February to Armando Guebuza.

The run-up to the 2004 general elections

Electoral framework: overview

The legal framework for multiparty elections has not changed fundamentally since it was introduced in 1993, although there have been some modifications. The Mozambican parliament adopted a new electoral law in June 2004 which regulated the electoral process for the polls on 1 and 2 December.

The main body in charge of elections is still the CNE, established in 1993 and based in Maputo. Yet, whereas in 1994 the CNE was composed exclusively of members nominated by FRELIMO and RENAMO, since 1999 it has been composed of a chair nominated by civil society (Reverend Arão Litsure) and 18 members nominated by parties in proportion to representation in parliament (which in reality means ten members of FRELIMO and eight members of RENAMO).

Another change is visible in the decision-making process within the CNE: while consensus was legally compulsory in 1994, it is now only recommended. This means that the FRELIMO majority can push decisions through the CNE without the assent of other parties. This -attracted sharp criticism of the CNE around the time of the 2004 election. The Technical Secretariat for Electoral Administration (STAE), which is the implementing body, falls under the CNE.

By law, the CNE has to publish national election results within 15 days of the elections, with copies being submitted to the Constitutional Council for verification. There is no deadline for action by the Constitutional Council, except that two days after the validation and proclamation of the results it must publish its findings in the Boletim da República. Within 15 days of publication the new parliament opens, after which the new president is inaugurated within eight days.

The Constitutional Council oversees the entire electoral process, approving the presidential candidatures, validating the final results and deciding on any electoral complaints. The council, although provided for in the constitution of 1990, did not exist until the local elections of 2003, its role until then being fulfilled by the Supreme Court. Three members were elected by parliament: FRELIMO named two, former minister Teodato Hunguana and Lucia Ribeiro, director of the law faculty of the Universidade Eduardo Mondlane; RENAMO named the lawyer Orlando da Graça; President Joaquim Chissano appointed the chairman of the council, former minister Rui Baltazar. These four members selected their fifth colleague, João Ubisse Ngwenha, who was the director of the legislative services of the Mozambican parliament.

Since the local elections in 2003 there have been election commissions at provincial and district level with a five-year term which are smaller than the CNE but with the same structure. The commission at provincial level has seven members, one appointed by the government and six by the political parties represented in parliament. At district or municipal level the government nominates one member and the parties in parliament four members. In a move towards decentralisation in the December 2004 elections, party polling agents (delegados de candidaturas), who are nominated by parties and are assigned to a certain polling station, did not receive their credentials from the CNE, but from the district election commission.

The June 2004 law is similar to previous election laws, for instance by retaining the 5% threshold for political parties to enter parliament. However, a few significant changes have been introduced. After complaints from the opposition and observers about FRELIMO using government vehicles and facilities, a ban has been imposed on the use by any party of the goods or property of central or local government, or of state-owned or state-controlled companies. Polling staff, police and journalists may now vote at any polling station. In previous years they were effectively disenfranchised because they were allowed to vote only at the polling station where they were registered, although their duties on election day might have taken them elsewhere.

The new electoral law clarifies confusing rules about the limits on activities near polling stations. The law provides that within 300 metres of the polling station, no voter can say whom he or she will vote for or plans to vote for. Observers and journalists are not allowed to talk to voters within this area. Also no campaign poster or other material is permitted within 300 metres of the polling station, and no one is allowed to display political symbols, or wear a party T-shirt. Exit polls had already been banned in the 2003 local elections. No opinion polls may be published between the start of the campaign and the declaration of results. There is a ‘quiet period’ for the two days before voting, during which campaigning must cease.

In 2004 Mozambicans living abroad were allowed to vote for the first time. The electoral law allocates two of the 250 seats in parliament to those living abroad: one for Africa and one for the rest of the world. In July 2004 RENAMO unsuccessfully objected to the decision of the CNE to allow Mozambicans residing abroad to register and vote. The main argument was that the CNE had not received evidence that the registration process could be monitored. Diplomatic postings in many countries are political appointments, and registration would take place in embassies, nearly all of which are staffed by people linked to FRELIMO.

Election observation

On a practical level the Mozambican electoral process is complex. After the voting, the electoral law determines that there will be two counts: one at provincial and one at national level. After the polling stations have counted the votes, they have to submit their summary sheets (editais) and minutes (actas) to the district commission, which in turn sends them to the provincial commission. Invalid ballot papers and disputed editais are excluded from these results and must be checked by the CNE in Maputo. Meanwhile, the CNE does its own ‘provisional count’, which is identical to the provincial counts. The CNE then compares the provisional count to the provincial counts and tries to reconcile them. This is done at national level with press and observers being excluded. The reconsideration of editais to be added to the final result also takes place behind closed doors.

The latter two regulations on the exclusion of observers caused an international outcry before the 2004 elections, observers saying that the final counting of Mozambique’s votes would not meet international standards of -transparency. They claimed the procedure could shift the final results by 1% to 2%, which would be enough to make a difference in a very close result. Some donors believed that the decision to maintain secrecy had personally been taken by FRELIMO’s presidential candidate, Armando Guebuza, as he wanted room for possible manipulation in the event of a close election. There was a historical precedent for this: in 1999 some observers estimated that President Joaquim Chissano’s official vote was 0,5% higher than his real vote.5 In that year’s elections, 1 874 editais contained errors or erasures or seemed to have been tampered with, which was about 11% of the total. The CNE accepted about one third of these dubious editais and rejected the rest. RENAMO eventually walked out of the CNE because it did not accept the arbitration process.

In 2004 international observers, including 32 delegates from the EU, nine from the US-based Carter Center (headed by former US president Jimmy Carter), and one from Japan, protested against the secrecy surrounding the reconciliation process of doubtful editais. At first the FRELIMO majority on the CNE refused to make any further concessions. However, in mid-November 2004, the CNE gave the assurance that the press, observers and party delegates would be able to examine the excluded editais, allowing them to confirm that the results were so unclear that it would have been impossible to include them. Additionally, the CNE allowed international observers to attend the process of reclassification of invalid and disputed ballot papers (nulos). However, the final tabulation process was done in secret, which led to further criticism by observers after the elections in December 2004.

The number of voters

A second controversial issue concerned the number of voters. In the run-up to elections in December 2004, more than 1.2 million new voters registered, nearly double the 700 000 predicted by electoral authorities. This process included an update of the register books (cadernos), which included new registrations, second-issue voters’ cards (for those who had lost or damaged their original cards) and transfers (for people who had changed their residence to another municipality). However, the process had been marked by significant problems, and was accepted only reluctantly by election observers.

During the local elections in November 2003 it had already become clear that the register books were in a chaotic state, after which the Constitutional Council and observers called for a clean up. But little was done during the registration process in July 2004 to correct the problems. Since the voter registration for the general elections of October 1994 there had been three additional registrations: a totally new national registration in 1999 and two updates in 2003 and 2004. People were registered via hand-written entries in books, and these were then computerised. But the computerisation process had many errors, and it appeared that many of the registration books were not proofread for mistakes. As a consequence, many people were registered twice or not at all.6

Furthermore, the numbers of voters remained uncertain. Officially, there were 10.4 million voters on the registers of the STAE, but the National Electoral Institute (INE) calculated that the total voting age population as of August 2004 was 9.1 million. According to other estimates, however, at the time of the elections Mozambique had 8.1 million voters.

Confusion over different official figures caused chaos in the local elections in 2003 as the number of municipal assembly seats was based on the number of registered voters. The CNE published a list of seats on 20 August 2003 and on 21 October 2003, then changed the numbers of seats in the assemblies of eight municipalities when it announced the results on 4 December 2003. The Constitutional Council ruled that it was not allowed to change the number of seats after the elections, and that the list published on 20 August 2003 should stand, even if it was wrong. The final, correct set of results was announced by the Constitutional Council on 14 January 2004 and published in Boletim da República only 12 days later.

Polling stations and their locations

There were 12 804 polling stations in the December 2004 elections, compared to 8 334 in 1999. The number of polling stations reflects both the number of voters and population density. Thus, although the Zambézia electorate is slightly smaller than that of neighbouring Nampula, it has more polling stations because Zambézia’s rural population is more spread out. The division of polling stations in Mozambique was as follows:

Additionally there were 60 polling stations in nine foreign countries, where about 226 000 Mozambicans are living, according to the estimates of the Foreign Ministry. In South Africa there were 40 polling stations, including some at workplaces such as mines. There were also five polling stations in Zimbabwe, four in Tanzania, three in Swaziland, two in Malawi, two in Kenya, and one in Zambia. In Europe there were two polling stations: one in Portugal and one in Germany.

However, a final polling station list with registration book numbers and numbers of registered voters was never made available to political parties or observers. This list, reportedly described by the CNE as a ‘state secret’, is essential because it determines the number and location of polling stations and ought to coincide with the database used for the tabulation software. The reasons for the secrecy are not clear and the number of registered voters or polling stations on election day is still unknown.

In all parts of the country there were polling stations that did not open, or opened only after long delays, especially in rural areas. In some instances heavy rain was the cause, as in 43 polling stations in Cabo Delgado and 33 polling stations in Zambézia. In other instances register books were missing or were sent to the wrong polling station.7 There were reports throughout the country of people having trouble in finding their polling stations, claiming the stations were not at the place where the person had registered. However, the electoral law allowed people to vote if they had a valid voter’s card for a particular polling station, even if through error their name did not appear on the register. In addition, people who had lost their voters’ cards but had valid identity cards could still vote if their names could be found on the register.

Software problems

The tabulation software used in the general elections of 1999 caused many problems, and predictions of a further computer-related crisis in 2004 proved to be correct. As in 1999, STAE staff wrote the software. Because they began this task in September, the testing and auditing of the software could only be done at the last minute, a week before the elections and in a space of four hours. The Mozambican firm Soluções took on the task of auditing and reported to the CNE in November. Soluções was mainly concerned about the lack of security of the system, which appeared to allow the introduction of false editais. Soluções insisted that a much tighter password and security system should be introduced.8

After the CNE had decided to implement the recommendations, which included ending open access to the system, STAE computer staff were reluctant to cooperate and sometimes were even obstructive, and it was increasingly left to Soluções to implement the changes.9 STAE computer staff input the underlying databases of polling stations and register books. This was done at the last minute, in part because of the last minute clean-up of register books, and was not completed until two days before polling began.

STAE technicians distributed the software to the provinces only two days after polling closed, which meant that provincial counting started late. Problems soon began to appear. For example, register books bore identical numbers in different provinces, and this had to be corrected. But the biggest problem was that the databases proved to have been input hastily and without proper checks, and therefore contained a huge number of errors. Valid editais were rejected when the register numbers did not correspond to the incorrect register numbers in the database. Hundreds had to be input manually, which was in breach of the new security system, or sent to the CNE in Maputo. Since the system in Maputo had also crumbled, the results were announced only on 21 December, four days after the legal deadline.

Election campaigns

No donor funds were available to political parties, but the government allocated 45 billion meticais (about US$1.8 million), as called for in the electoral law. The amount was disbursed in the same way as in 1999: one third for presidential candidates, one third for parties currently holding seats in parliament based on the proportion of seats held, and one third for parties standing for parliament based on the number of candidates. Funds were disbursed in three parts: half was given on 16 October, with the rest delivered in two instalments of 25% each. Each additional instalment, however, was given only after a party justified its expenditure of the previous one.

The 43-day formal electoral campaign was generally peaceful and often festive in -atmosphere, but there was some tension and a few incidents of intimidation and violence early in the campaign, especially in Nampula, Gaza and Tete provinces. Also, the historical antagonism between FRELIMO and RENAMO manifested itself, particularly in the opposition stronghold of Sofala province. The province had accommodated RENAMO’s military headquarters in the district of Maríngué in the closing years of the war of destabilisation. Even though all government and RENAMO troops were supposed to have been demobilised and disarmed in 1994, under the terms of the 1992 General Peace Agreement, RENAMO has kept its Maríngué force operational. The group of perhaps between 50 and 150 men supposedly serve as bodyguards to protect their leader’s home, although Afonso Dhlakama now lives in a government house in Maputo.The government repeatedly offered to incorporate the men into the police force, but Dhlakama refused.

The problem of the illegal armed forces in Maríngué has dragged on for a decade. The government has been tolerant, since it could have sent in riot police to forcibly disarm the RENAMO men, and dismantle their base. So far it has declined to do so. However, political competition has intensified in the province since RENAMO, for the first time, claimed Beira in the local government elections in November 2003. In 2004 several cases of political violence were reported throughout the province, such as beatings of local officials and police officers, housebreaking and random shootings, allegedly perpetrated by the RENAMO troops in Maríngué.

There were also allegations that Dhlakama urged citizens in six of Sofala’s districts to contact him should they feel that local administrators or police were ‘committing abuses’, so that he could send his security men to seize them. However, it is important to be aware of possible bias, particularly towards the ruling party, in reports of the alleged incidents. RENAMO rejected these allegations as a strategy by FRELIMO to try to discredit RENAMO and its leader ahead of the 2004 polls. RENAMO’s justification for its members’ actions is that they have merely acted in response to FRELIMO provocation, and the party has accused the government of deliberately trying to discredit Dhlakama before the elections.

FRELIMO, in turn, was accused of buying voter’s cards in Sofala: an accusation denied by the ruling party. Dhlakama is adamant that those affected are supporters of RENAMO.

Since Sofala is a long-established opposition stronghold, RENAMO’s concern with maintaining its influence there is not new or surprising. However, RENAMO’s success in Beira in the local government elections has, perhaps, served to raise the political stakes for both parties in the province. The cause for concern is that the reported incidents of violence, whether they are accurate or not, serve to highlight the ongoing distrust between the two main political parties and the need for continuing efforts to establish a culture of political tolerance in Mozambique. The existence of RENAMO armed forces in Sofala could threaten not only law and order and local government authority, but also the possibility of cooperation between the two parties to ensure a stable political environment in the province.

Previous elections: overview and synthesis

1994: presidential and legislative elections

Mozambique’s first democratic elections took place on 27–29 October 1994. The preparations were undertaken in three phases: voter registration, voter education and the electoral campaign. Voter education, which covered electoral education (how to vote) and education for democracy (why vote), proved to be a difficult task. The high rate of illiteracy among Mozambicans and the lack of good communications networks throughout the country substantially hampered the effectiveness of voter education campaigns. Although methods using dance and theatre were often employed with good results, the rural areas were much less targeted than the urban areas.

A further issue was the funding of smaller opposition parties, which needed financial support if they were to be able to compete with FRELIMO and RENAMO. Ten smaller parties ran alongside the two giants, and ten presidential candidates contested the election in addition to Chissano and Dhlakama. Eventually the small opposition parties each received 100 million meticais (about US$15 000) from the government and an additional US$100 000 from the international donor community. Campaigning was not an easy task, however. In RENAMO-controlled areas, political activities by parties other than RENAMO itself were not allowed until the end of the demobilisation process a few months before the elections. Some opposition parties also accused the government of harassment in the areas it controlled.

During the election campaigns FRELIMO emphasised continuity and its experience in governance, while the opposition campaigned on the need for change. The opposition parties, including RENAMO, based their campaign on the mistakes made by FRELIMO during its 19 years in power and on the need for a more even distribution of power and resources to the central and northern regions of the country. They called for the establishment of a government of national unity after the elections: a concept that found interested ears within the international donor community and the Mozambican churches. However, FRELIMO refused to consider this option, arguing that the party that won the election should determine the composition of the new government.

The general elections were monitored by approximately 2 300 UN observers provided by individual UN member states, a range of UN offices, the UN Operation in Mozambique (ONUMOZ), the European Union, the diplomatic community in Maputo, NGOs working in Mozambique, the Organisation of African Unity (OAU) and the Association of European Parliamentarians for Southern Africa (AWEPA). The cost of the operation in 1994 was estimated as US$294.8 million.

A serious threat to the process occurred when RENAMO threatened to withdraw from the voting on the eve of the elections, alleging preparations for widespread fraud on the part of the government. The crisis was solved after tense negotiations between Dhlakama and donor representatives. The incident did not affect the course of events: the elections were characterised by an exceptionally high turnout. Around 87% of the 6.3 million registered voters cast their ballots, of which 91.4% were counted as valid.

Chissano won the presidential election, receiving 53.31% of the votes, against 33.73% in favour of RENAMO leader Afonso Dhlakama. Interestingly, Chissano proved more popular than his party, whereas RENAMO emerged stronger than its leader. FRELIMO failed to score an absolute popular majority in the legislative elections, winning only 44.33% of the vote, which came as a great shock to the party. RENAMO effectively became the leading opposition party, its candidates securing 37.78% of the parliamentary vote. Of the minor parties, only the UD (União Democrática) qualified for representation in the legislature.

FRELIMO and RENAMO showed that they are the only parties with national reach, although both possess regional strongholds: FRELIMO in the south and far north and RENAMO in the centre (Manica, Sofala, Tete, Zambézia and Nampula). RENAMO fared particularly well in Sofala, where it obtained more than 80% of the vote. The importance of this victory was enhanced by the fact that Sofala’s capital, Beira, is the second largest city in the country.

Different explanations for RENAMO’s electoral success, so soon after the war had ended, quickly emerged. According to some, RENAMO gained political support because the party was able to attract highly educated people who hoped that their RENAMO membership would lead to a speedy political career. Also, at the time of the elections RENAMO controlled about one third of the country. International donors delivered emergency aid and food in these areas, having sought the approval of RENAMO, and not of the government. Furthermore, having failed to secure a pre-election deal on ‘power-sharing’ or a government of national unity to guarantee that RENAMO had a place in the executive, international donors (in particular the US) and the Catholic Church urged people to vote for RENAMO. Only a strong opposition, they believed, would keep Mozambique on the road to peace. And it was peace, and not so much RENAMO or FRELIMO, that the population voted for.

After the elections FRELIMO adopted a winner-takes-all position. The argument was that the electoral results had given FRELIMO the right to govern alone and that one of the central characteristics of democracy was -precisely the existence of an opposition. RENAMO accepted FRELIMO’s arguments, perhaps out of the strategic conviction that during its five-year mandate FRELIMO would fail to meet expectations, while with the passage of time the memories of RENAMO’s cruelty during the war would fade away as the organisation built a new image.

1998: Local elections

Following the first democratic elections, Mozambique became a favourite among the international donor community. ONUMOZ was widely portrayed as an overwhelming success in that it contributed to a reasonably stable and quick transition from war to peace, and to the organisation of peaceful and fair elections, which marked the symbolic birth of Mozambique as a multiparty democracy. It soon became clear, however, that the world community had celebrated prematurely. In 1998, during the first local elections, the old enmities between FRELIMO and RENAMO resurfaced.

Power relations remained a sensitive issue in the country, and decentralisation was suggested as a viable solution. Devolution would not only bring citizens closer to state institutions that were proponents of the proposed decentralisation model; it would also enable power-sharing without having to create a government of national unity. Mozambique had always been a very centralised country in which all civil servants, even in the smallest locality, were appointed by the central government. In fact, the country had adopted a legal framework for local government reform in 1994 which proposed to turn all 128 districts into municipalities,10 constituting a radical break with the past. However, in 1997 FRELIMO’s parliamentary majority made it possible to repeal a law whose most relevant feature was the establishment of district municipalities linking urban and rural areas. This would have meant the dilution of an overly centralised political system, which practised a two-tiered administrative system: one for rural areas and one for urban areas.

The new law that was eventually promulgated promised to start by holding municipal elections in the existing 33 urban municipalities. Only after this was done would the elections be extended to rural areas.11 Hence, full -participation was -hindered by the new law, as the local government system left out most of the rural population, remaining subject to continued rule by officials appointed from Maputo.12

The first municipal elections were a sobering experience for the Mozambican polity as a whole. On the eve of the poll RENAMO and 15 smaller opposition parties decided to abstain from participation, arguing that the government was bent on fraud. Only 14,48% of the registered voters turned up, and the inevitable result was that FRELIMO candidates were elected as mayors in all 33 municipalities where elections were held. The ruling party achieved absolute control over the new municipal authorities.

1999: Presidential and legislative elections

Following the mistrust that surrounded the conduct of the 1998 local elections, greater efforts were made to ensure that the presidential and legislative elections that took place from 3 to 5 December 1999 were run in a transparent and accountable manner. The Mozambican parliament passed new electoral legislation to pave the way for the creation of a new CNE to oversee the 1999 elections.

Chissano was re-elected with 52.29% of the vote, against 47.71% for Dhlakama (who in 1994 had secured only 33.73% of the presidential vote). In the parliamentary poll FRELIMO consolidated its majority with 48.54% of the vote, against 38.8% for RENAMO.

RENAMO was seriously disappointed. The former rebel movement complained that FRELIMO had again corrupted the electoral process. On 22 December 1999 Dhlakama appealed against the election results. His case was based on 23 complaints focusing mainly on irregularities during the counting, and mistakes and numeric discrepancies in the officially published tabulation sheets. The principal problem concerned the re-classification of votes in the district commissions. Many votes that had been regarded as valid at local level were considered null when they came to be analysed at district level. The justification was that voters had not filled in the form properly, or had marked the vote in the wrong place. The opposition also claimed that it had not been able to conduct its election campaign in Tete because of FRELIMO aggression in the province.

The Supreme Court investigated the counting process and, on 4 January 2000, rejected all the complaints, thus validating the results. A seven-month boycott of parliament by RENAMO followed, while FRELIMO accused the former rebel movement of undermining the peace process. Dhlakama threatened to form separate governments in the six provinces where his party had won the majority vote, and indicated publicly that he intended to paralyse the country and to make it ungovernable if a power-sharing agreement could not be reached. In 2000 RENAMO moved from verbal attacks to nationwide demonstrations. Riots broke out in Montepuez in Cabo Delgado province after the police opened fire on demonstrating RENAMO supporters, an incident in which at least 45 people lost their lives, including seven policemen. Shortly afterwards the suspected agitators, some 100 RENAMO supporters were confined in a tiny police cell where they eventually died of asphyxiation – an incident that indicated the neglect of human rights and absence of the rule of law. Both parties showed a disregard for democratic process and behaved in a spirit of struggle.

2003: Local elections

Because RENAMO did not participate in 1998, the local elections of 19 November 2003 were the first inclusive ones. Turnout was still relatively low, although it varied regionally – from a low of 15% in the district of Nampula to highs of 46% in Mocimboa da Praia and 47% in Motize. Nationwide, only 27.56% of the more than 2 million registered voters cast their ballots in the 33 municipalities. The government party retained a firm hold on local politics in 28 districts, including Maputo.

RENAMO received a small majority of votes only in the province of Sofala: 49.98% against 45.64%. Of 33 municipalities, RENAMO mayoral candidates were elected in only five. In Beira, Daviz Simango of RENAMO was elected with 53.45% of the vote. Three opposition victories were registered in the port of Nacala, Angoche and Ilha de Moçambique, all in the province of Nampula. In the province of Sofala, João Agostinho of RENAMO won by a single vote in the small town of Marromeu, while FRELIMO controlled the municipal assembly.

Six small parties and citizens’ lists won seats in municipal assemblies, the same number as in 1998. The Independent Party of Mozambique (PIMO) won its first-ever victories, with -single seats in Nampula, Cuamba and Angoche, and 0.18% of the total votes. The Democratic Institute for Peace (IPADE), the grouping led by former RENAMO politician Raúl Domingos, won single seats in Beira and Dondo. Only two of the five local citizens’ lists that won seats in 1998 stood again in 2003, and they did badly. In Nacala the Organisation of Independent Candidates of Nacala Port (OCINA) was reduced from 11 seats in 1998 to just one in 2003. In Maputo, the Together for the City movement won five seats, compared to 15 in 1998.

The outcome of this local election showed that newly established political parties still had little say in Mozambican politics. If these results are taken as indicative of voting trends in the parliamentary elections, they suggest that only FRELIMO and RENAMO would be able to secure the 5% of the total vote necessary to gain seats in the legislature: a trend that was confirmed by the results of the 2004 general election.

In the days after the local elections, RENAMO reacted with outrage to its unexpectedly poor showing. ‘We do not recognise the results, and demand that the elections be annulled,’ RENAMO secretary-general Viana Magalhães declared on 4 December 2003. International observers had noted ‘irregularities’ during the election process and almost 20% of the ballots were considered invalid. The CNE had made mistakes in the initial announcement of seat distribution to winning parties, and the corrected results published on 11 December still contained errors. The Constitutional Council had to request clarifications from the CNE, which were forwarded on 18 December. Yet the international observers, including delegates from the Carter Center and the European Union, considered the elections generally to have been ‘free and fair’.

RENAMO nevertheless submitted protests to the CNE on irregularities noted during the electoral process, which allegedly favoured FRELIMO in almost half of the municipalities. The CNE published a deliberation addressing each complaint and the reasons that all had been rejected. In most cases the CNE stated that the irregularity had no influence on the final results, or that there was no material proof for the opposition’s claims. RENAMO was still not satisfied with the CNE’s arguments and submitted appeals to the Constitutional Council concentrating on non-compliance with Article 100 of the electoral law. FRELIMO also submitted appeals concerning the legality of certain RENAMO candidates. The Constitutional Council rejected all appeals and validated the results on 15 January 2004. Eventually RENAMO refrained from further action and returned to politics.

The next local elections in Mozambique will take place in 2008. The intention is to increase the number of municipalities, but it is not yet clear how and when this will happen. The Assembly of the Republic, however, recently approved a constitutional amendment that allows for elected provincial assemblies from 2008. These assemblies will relate to the provincial governments in much the same way as the Assembly of the Republic relates to the president and central government. Governors, however, will continue to be appointed by the president rather than elected, which means that perhaps half the provinces could find themselves with a governor from one party and an assembly controlled by another party.

Conclusion

From an analysis of the results from the four elections before the December 2004 presidential and legislative elections it is clear that FRELIMO obtained far better results in 2003 than in 1994 and in 1999, even in provinces previously dominated by RENAMO. In 2003 FRELIMO received 71.02% of the valid ballots cast, while in 1994 its presidential candidate, Joaquim Chissano, received a mere 53.66% and in 1999 no more than 52.3%.

In 1994 legislative candidates representing RENAMO won most of the seats in five provinces, a number that increased to six in 1999. In fact, in several of those provinces, RENAMO was able to secure more than two thirds of the seats. However, in the 2003 local elections only 26.55% of the electorate voted for RENAMO, which shows a sharp decline of support for RENAMO in local level politics.

Because general turnout was much lower than during the presidential and legislative elections, however, it would be unwise to draw too many conclusions from the 2003 local elections about long-term trends. The same is true of the first local elections in 1998, as RENAMO and ten smaller parties decided to abstain from participation. Moreover, the local elections excluded more than half of the Mozambican electorate, specifically the rural population, who may have different political preferences from those of urban voters.

What is worth noting about the 2003 vote is that, once again, it took place in an environment of profound distrust between the major contenders. The numerous legal, political and technical measures that were taken to guarantee transparency, freedom and fairness did not prevent allegations of fraud from being made throughout the process.

Taking a longer historical view, this lack of trust can be seen to have been present at every step of Mozambique’s democratic process since 1994. In that year RENAMO decided to withdraw from the peace process in the run-up to the elections and allegedly only money brought them back – even though, of course, the official explanation is that ‘diplomatic pressure’ convinced them. During the local elections of 1998, again on the eve of the elections, RENAMO decided to boycott, but this time no diplomatic pressure was brought to bear, and RENAMO was therefore not part of the local elections. A year later, in 1999, the rejection of all RENAMO’s complaints about the general election process led to a months-long boycott of parliament and violent clashes, bringing the political relations between RENAMO and FRELIMO to a low point. Furthermore, the allegations hindered fundamental progress in building confidence among the electorate that everyone was playing by the same rules, which is a basic indicator of the consolidation of democracy.

Only 3.3 million of 7.6 million potential voters (43%) cast their votes in the December 2004 elections, according to AWEPA.13 This low turnout contrasts with 1994, when 87% of the registered voters went to the ballot boxes. According to the Maputo-based newspaper Savana, the poor showing can be seen as a -‘yellow card’ for Mozambique’s political elite, a rejection of all political parties.14

Contrary to expectations, voters did not opt for alternatives. Opening the political sphere to non-FRELIMO and non-RENAMO political actors did not modify the main contenders for power in Mozambique: in effect the country has a two-party system clearly centred on the competition between FRELIMO and RENAMO, a dichotomy of political identities deeply rooted in the country’s past conflicts.

Of the two rivals, FRELIMO seems to have a more loyal electoral base than RENAMO, because RENAMO has been losing votes continuously and has become weaker with each successive election. Social marketing expert Juarez da Maia argues that RENAMO is the primary cause for the high abstention through its ‘strategy of defeat’. By constantly stressing its claim that it was denied power through fraud in 1994 and 1999, it created the image that there is no point in voting because FRELIMO will win no matter what. Accordingly, RENAMO has undermined its own image and has contributed to voters’ apathy.15

Moreover, it has been argued that Dhlakama failed to create solid party bases in the run-up to the December 2004 elections, while FRELIMO’s presidential candidate Armando Guebuza spent more than a year travelling around the country rebuilding the party foundations and ensuring its fidelity to him. The results suggest that the party successfully mobilised its core supporters. RENAMO, by contrast, has no effective mobilisation structure; Dhlakama runs the party in a personalised way and has dismissed those capable of creating a party machine because he feared they would challenge him. But elections in 2003 and 2004 show that people will vote only when a party urges them to do so.

Notes

1. From official CNE tables, AWEPA, Maputo.
   
   

2. Ibid.
   
   

3. Post-election Statement on Mozambique Elections, 21 December 2004, Carter Center, Maputo.
   
   

4. Mozambique Political Process Bulletin 34, January 2005, AWEPA, Maputo.
   
   

5. Mozambique Political Process Bulletin 24, August 2004, AWEPA, Maputo.
   
   

6. The problems caused by names being left off the register were to some extend mitigated by a law allowing people to vote solely on the basis of a valid voters’ card if their names did not appear on the list at the station where they ought to have been registered.
   
   

7. Mozambique Political Process Bulletin, 18 December 2004, AWEPA, Maputo.
   
   

8. Mediafax, STAE mente sobre dossier software, 24 November 2004.
   
   

9. Mozambique Political Process Bulletin 31, December 2004, AWEPA, Maputo.
   
   

10. In Mozambique, the municipalities have almost the same structure as central government, with the same ‘presidential system’. Every five years, a municipal president (mayor) and municipal assembly are elected. The municipal president then chooses local ministers, known as vereadores, who administer the various departments of the municipality. Apart from these municipalities, there are also so-called local state organs, consisting of 128 districts, 393 administrative posts and 1048 ‘localidades’ (local communities). In all cases the administrators, as well as provincial governors, are appointed centrally.
    
    

11. There were also reports of the poor performance of FRELIMO in the rural areas during the first national elections, which contributed to the -decision to exclude the rural areas from the process of municipal government.
    
    

12. E Braathen, Democratic Decentralization in Mozambique?, in M Newitt with P Chabal & N Macqueen (eds), Community and the State Lusophone in Africa, King’s College, London, 2003, pp 99-126.
    
    

13. If one takes the official voters’ roll figure of 9.1 million, then the turnout of 3.3 million represents a mere 36.4%. It is however universally recognised that voters’ roll numbers have been inflated by duplicate registrations and a failure to remove the names of deceased voters.
    
    

14. Savana, Absentismo venceu eleições?, 3 December 2004, p 2.
    
    

15. As cited in Mozambique Political Process Bulletin 31, op cit.