Since 1990, Yemen’s electoral commission—the Supreme Commission for Elections and Referendum (SCER)—has gained experience through the conduct of three parliamentary, two presidential and two local council elections. Following the 2003 parliamentary elections, international NGOs reported that—despite inevitable violations and irregularities—the SCER had shown major technical improvements in its conduct of the elections as compared with earlier polls. In 2006, the major challenge facing the SCER was to maintain these gains, despite the fact that the country’s September elections were to be the most complex it had ever mounted.

The 2006 elections were particularly complicated for three reasons. First, it was the first time that simultaneous elections were held for the presidency, the governorate councils and the councils of the administrative districts, which created significant logistical and voter education challenges for the SCER. Second, it was also the first time that the incumbent candidate of the ruling General People’s Congress (GPC), President Ali Abdullah Saleh, faced a credible opponent, Faisal bin Shamlan. Shamlan was endorsed by a coalition of opposition parties called the Joint Meeting Parties (JMP). Third, the elections were held in a political climate characterized by strong distrust of the SCER by the JMP. Although the SCER did make some sporadic efforts to promote a dialogue with the JMP, these attempts were not productive. Despite these challenges, the SCER succeeded in maintaining the technical progress it had made in the 2003 elections.

Preparing for the Elections

Yemen’s constitution requires the nine-member SCER to be independent and impartial. Operationally, the SCER conducts voter registration, elections and referendums through a series of  three-member committees. Although not legally required, Yemen’s practice is for members of all these committees to be party representatives, according to proportions agreed by the parties. In 2006, the SCER appointed a supervisory committee for each of Yemen’s 21 governorates, a main committee for each of the 333 administrative districts, and a total of 27,010 male and female sub-committees for 5,620 voting centers.

The SCER’s efforts to administer the fairest elections possible were hampered by two realities in place before they began preparations. First, in 2005 the government—apparently without consulting the SCER—ruled that the 2001 election law would not be amended before the 2006 elections. This decision was very unfortunate, because it meant that issues raised after the 2003 elections (in IFES reports and elsewhere) would not be addressed in time for the new polls. In addition, the 2001 election law is not well suited to simultaneous elections and does not specifically recognize the multi-member districts used for some local council elections.

Second, the boundaries of the local electoral constituencies were not revised according to the newest census data. According to statute, the SCER sets these boundaries so that the population in each constituency varies no more than 5 percent. A population census was held in December 2004, but the final results were not issued until January 2006—too late for the SCER to revise the boundaries of all 5,620 local constituencies before the voter registration update. It is certain that the constituencies within many administrative districts do not currently comply with the statutory population requirements.  

In its preparations for the elections, the SCER successfully distributed supplies, printed ballots, trained staff and registered voters, though it did encounter some unforeseen difficulties in voter registration. First, the SCER successfully managed the major logistical exercise of providing election supplies (e.g., ballot boxes, screens, forms, stationery, manuals, indelible ink, etc.) to the male and female sub-committees in each of the 5,620 voting centers. The UNDP was a major contributor of election supplies and also conducted voter education on behalf of the SCER.

Second, the task of printing the ballot papers was a major undertaking, and one tied to very tight timeframes. All ballots were printed centrally and distributed to the 27,010 sub-committees at the voting centers according to the number of voters registered with each sub-committee. The SCER contracted a private printer for the presidential ballot papers, but it printed the 333 different governorate council ballots and the 5,620 different district council ballots. Different colored security paper was used to distinguish each type of ballot paper. Despite the significant logistical challenges, the SCER was able to distribute the ballots as planned.

Third, the SCER and its staff needed to be well trained, which is a crucial ingredient in the efficient, effective and professional administration of elections. In cooperation with the UNDP Election Support Program, IFES developed, translated and implemented a SCER Professional Development Course that combined modules from IFES’ BEAT (Basic Election Administration Training) program and the BRIDGE (Building Resources in Democracy, Governance and Elections) program, which was jointly developed by UNEAD, International IDEA and the Australian Electoral Commission. International trainers conducted two workshops, each for 35 mid-level SCER staff, using Arabic training resources that can easily be adapted for use in other countries in the region. The effects of these courses were readily apparent in the much improved training program and materials the SCER implemented for the 2006 election committees.

Fourth, the SCER had to delay the pre-election voter registration update following the failure of the GPC and the JMP to agree on their proportions of relevant committee memberships. Eventually the SCER could wait no longer and appointed the committees from applicants for civil service employment. The total number of registered voters increased from 8.1 million in 2003 to 9.3 million in 2006. Following the update, 42% of registered voters were women, the same proportion as in 2003. Because there were large numbers of duplicate and under-age registrations, the SCER referred about 300,000 names to the public prosecutor, resulting in the removal of 250,000 names before the voter lists became final on July 26.

Candidates and the Campaign

While the voters had many candidates to choose from—five presidential candidates, 1,612 candidates for 431 governorate council seats, and 18,901 candidates for 6,896 administrative district council seats—a disappointingly small number of these candidates were women. Despite many pre-election statements by the GPC that at least 15% of its governorate council candidates and at least 20% of its district council candidates would be women, in fact it nominated very few female candidates. Islah (the largest party in the JMP) nominated no female candidates at all. Overall, 0.7 percent of candidates in 2006 were women, slightly more than at the previous local council elections in 2001 (though 35 women were elected in both years). This year, seven women won governorate council seats, and 28 women were elected to the district councils. There is certain to be renewed debate about legal changes to improve the participation of women as candidates before the 2009 parliamentary and local council elections.

During the campaign, the SCER is charged with ensuring candidates have equal opportunity to communicate their campaign messages through the official media. Unlike 2003, this year the SCER did ensure that opposition party presidential candidates received equitable coverage during the four-week campaign period. However, the fact that, by law, reporting of the president’s “daily activities” was not counted as part of his campaign meant that the incumbent received the most coverage by far. In addition, as in previous elections, government resources were used illegally to support the incumbent president and his party.  

The governorate council and administrative district council elections were largely ignored by the media during the two-week campaign period. Candidates’ campaigns seemed to consist largely of display of campaign posters.

Voting, Vote Counting and Results

Despite some fears of violence, election day passed without major incident. On the whole, it appears that the SCER’s sub-committees did a good job, although there were reports that some members of female sub-committees were not well trained. Voters appeared able to cope with the three different ballot papers, and voter turnout was 65 percent (for the presidential election), compared to 77 percent for the 2003 parliamentary elections. However, as expected, there were anecdotal reports of irregularities and violations, e.g., non-secret voting, campaigning on election day in or near voting centers, and vote buying.  

Vote counting started late and proceeded very slowly. First, all presidential ballots were counted and the results reported, then the governorate council ballots were tallied and then the district council ballots. The EU observer teams reported significant problems with counting in some areas where the counting was conducted in an extremely tense environment. There were also anecdotal reports of irregularities, such as the exclusion of party/candidate observers, the involvement of security forces in the counting process, and “creative” interpretations of the legal criteria for identifying a valid vote.

In the presidential race, incumbent President Saleh won 77 percent of the valid votes (compared to 96 percent in 1999). The JMP candidate, Faisal bin Shamlan, won 22 percent, and actually won a majority in 18 of the 333 administrative districts. The percentage of invalid votes was 3 percent (approximately the same as in 2003). In the governorate council election, the ruling GPC party won 85 percent of the seats, compared to 10 percent for JMP candidates and just over 5 percent for independent candidates (6 percent of votes were invalid). In the local council election, the GPC won 76 percent of seats, JMP candidates secured nearly 15 percent, and independent candidates got 8 percent (5 percent of votes were invalid). 

Conclusion

Overall, the SCER’s conduct of the 2006 presidential and local council elections consolidates the progress it made in 2003. However, because the 2006 elections were much more technically complex, they can be seen as a significant advance over the 2003 polls. Nevertheless, some problematic issues must be dealt with before the 2009 elections. First, the election law must be amended, the boundaries of parliamentary and local constituencies must be revised (using a transparent process), and action must be taken to ensure the integrity of the voter registers. Second, ways need to be found to improve women’s representation as political candidates. Third, the SCER needs to deal with technical issues such as the counting/reporting processes and improving disabled people’s access to the electoral process. Finally, the SCER needs to be more determined to enforce the election law, to be seen as more independent, and to be more open and transparent, particularly in its relations with political parties.

Paul Harris was IFES’ Country Director in Yemen from June 2004 to December 2006.