A Case Study of the 1999 National Election in Indonesia
by Tim Meisburger
Election Advisor
The Asia Foundation
Introduction
The resignation of President Soeharto in May 1998 began Indonesia's remarkably rapid retreat from autocracy, and stunning rebirth as the world's third largest democracy. Although 'elections' had been held regularly during Soeharto's 32 years in power, they were closely managed by the state bureaucracy to ensure an appropriate result, their primary purpose being to provide Soeharto and GOLKAR (his party) with a thin veneer of legitimacy.
The history of managed elections, and the fact that the election administrators were still largely unchanged from previous elections, led many to suspect that the transitional elections scheduled for June 7, 1999 would not be free and fair. In the event, they were undoubtedly the freest and fairest in Indonesia's history, an outcome in large measure attributable to the role domestic observers played in ensuring the integrity of the election process.
Under Soeharto, Indonesia's election administration had been designed to ensure the government could control the outcome of each election. According to IFES's Indonesian Election Law Assessment:
Rather than providing transparency or 'checks and balances,' this system served to obfuscate lines of authority and insulate government control over the election process. In particular, Panwaslak the supposed supervisory and semi-judicial authority became a black hole for complaints and a legitimizing decoration for the government-run structure.409
Shortly after Soeharto's resignation, in an effort to stem public unrest, the interim government of President B.J. Habibie announced that elections would be held the following year, and that the existing election administration structure would be replaced with an independent election commission, and a more transparent election procedure.
The Election Law
The process of drafting the new law involved much backroom discussion and political horse-trading, as the various parties sought to gain a structural advantage. As would be expected of legislation resulting from political compromise, the law that eventually emerged from this process did not describe procedures in great detail, and lacked some internal coherence, but it was not all bad. According to IFES, 'the new General Election Law inspires neither great alarm nor great confidence'.410
Of course, the drafting of the law did not take place in a vacuum, and throughout the process student groups, NGOs and the international community were advocating for features that would improve the transparency and fairness of the election. This advocacy, and the desire on the part of the drafters that the law be viewed as credible both domestically and internationally, led to several important improvements in the new law.
To increase the transparency of the process, officially accredited foreign and domestics monitors were allowed to observe each stage of the election. Although the new law did not describe the polling and counting processes in great detail (specifically assigning that responsibility to the election commission), it did mandate that voters be marked with indelible ink, to deter multiple voting.
Unfortunately, the long drafting process meant that the law was not passed until January 1999, only five months before the election. This left very little time for setting up the new commission, drafting detailed rules and procedures, and hiring and training new staff.
The Election Administration Structure
The new election commission, called the Komisi Pemilihan Umum (KPU) in Indonesian, was finally established in March 1999, 86 days before the election. In the structure set out in the new law the KPU was only responsible for setting policy, while actual implementation was to be the responsibility of a subordinate body called the Panitia Pemilihan Indonesia (PPI or Indonesia Election Committee).
When it became clear that the KPU and PPI would have difficulty preparing for the election in the time available, a third body was added to the administration structure, called the Secretariat. Since the Secretariat was made up from the same local officials who had administered past elections, many felt this was an effort to subvert the independence of the commission.
It was intended that the Secretariat would work under the direction of the PPI to implement the new process, under the overall control of the KPU, but this complex system soon fell apart. The polling officials received little or no training in the new procedures, and the communication system between the PPI and the Secretariat broke down (or was never established).
With tremendous political pressure to have the elections on time, the election officials in the field pressed ahead, but without clear guidance from above, they tended to fall back on the systems and procedure they knew. The result of this lack of training and effective control was that the 1999 transitional elections were administered in much the same, deeply flawed, fashion as previous elections had been.
The KPU had a hugely difficult task. In a few short months it was expected to create a new organization and hire 2.7 million pollworkers, who would administer more than 300,000 polling stations for 121 million voters, scattered across and archipelago of more than 3000 inhabited islands. But civil society faced a challenge similar in scope, with far fewer resources. Student groups and NGOs had been in the forefront of the effort to topple the Soeharto regime, and as the election approached, they were eager to see the democratic transition continue. They recognized the weakness in the legislation, and saw that much of the election administration was controlled by a single political faction, and worried that their remarkable accomplishment might be stolen at the last moment. Many decided that close monitoring of the process was the best way to help ensure a free and fair election.
The Election Monitoring Organizations
Prior to the 1999 election, there was only one NGO election monitoring organization (EMO) in Indonesia, the Komite Independen Pemantau Pemilu (Independent Committee for Election Monitoring, or KIPP). KIPP was set up prior to the 1997 general election, but in that election the process was so 'managed' that few were interested in observing it, and KIPP was only able to field 9000 unaccredited (therefore unofficial) observers on election day. In the words of KIPP's founder, Mulyana Kusumah 'The government pretended to have an election, and we pretended to monitor it.'
In contrast to the '97 election, interest in the first post-Soeharto election was very high, and in the year prior to the election several other national monitoring groups were created, as well as numerous regional ones. KIPP, formed as a coalition of NGOs, expanded its volunteer base by recruiting both individuals and new coalition partners. The student groups, who had been instrumental in the struggle to bring down Soeharto, formed a university based observation network called UNFREL.
Another group tapping the universities for volunteers was the Rector's Forum, organized by university rectors. The last two national groups, JAMPPI and the JPPR, were organized around Islamic religious organizations. In total, the national groups, KIPP, Rector's Forum, UNFREL, JAMPPI, and the JPPR, fielded over 600,000 observers for the 1999 election, and average of two per station.
Only one of the national groups (the JPPR, or People's Voter Education Network) was formed for a purpose other than election monitoring. Although each of these groups were formed spontaneously, they all received significant technical and financial support from the international community. The Asia Foundation, in particular, made early and frequent efforts to encourage coordination among the EMOs. The Foundation also encouraged the development of a standard monitoring methodology that would allow the aggregation of results from different groups, to provide a more comprehensive picture of the entire process.
Direct financial support for domestic election monitoring was provided by USAID, The Asia Foundation, The American Center for International Labor Solidarity (which recruited, trained, and managed a relatively small contingent of long-term domestic observers). The European Union, and seven countries (led by Japan and Australia)also provided substantial support for monitoring through a UNDP-managed Trust Fund.
There were significant problems in accrediting these observers at the local level, primarily because election administrators were overwhelmed by their own task. As the election loomed, a political accommodation was reached whereby organizations rather than individual would be accredited, and organization identification would be sufficient to enter the polling station.
Registration Process
The voter registration process was wildly unorganized. Registration officials were inadequately trained in the new procedures. In many instances in the outer islands officials failed even to receive written instructions on the procedure. Essential materials and funds to pay staff arrived late, or never. In areas where it was impossible to conduct a proper registration, officials fell back on the voting lists from the 1997 election.
Lack of official accreditation, lack of financial resources, and a lack of adequate preparation among the EMOs meant that in most instances, quantitative and organized monitoring of the registration process did not take place. Several organizations were, however, able to spot check the process, and their reports tended to confirm the general impression of chaos during the registration.
The Polling and Counting Processes
The lack of time for preparation, and the use of the existing election administration infrastructure, meant that the polling process used in 1999 was essentially unchanged from that used in previous elections. The only major changes in the process were the addition of indelible ink (supplied by the international community) to mark the voter's fingers to prevent multiple voting, and provision in the legislation for the presence of non-partisan observers.
The process itself (which had been used in previous elections)was simple and transparent. The polling stations were set up in the open, under an awning made of canvas or other locally available material. Because there were no walls, local people as well as observers could watch the entire process.
In the morning the ballot box was shown to be empty and then sealed. As people arrived their names were checked on the voter's roll, they received a ballot, marked the ballot behind a screen, then dropped it in the box. At the end of the day, in full view of the observers, party agents, and crowd surrounding the station, the ballot box was opened. As the ballots were counted, each was held up so the crowd could see the mark, and the choice was recorded on a large blackboard. At the end of the count, marks are tallied up, and the final result for that station posted on a board.
The Role of Domestic Observers in Ensuring the Integrity of the Process
The over 600,000 organized domestic observers who monitored and reported on the election undoubtedly contributed to deterring fraud and other forms of electoral abuse throughout the process, but it is likely that their most important contribution was made in ensuring the integrity of the counting process. Because party agents, domestic observers, and ordinary citizens could all observe the polling and counting processes at the polling station, there was little opportunity for cheating there. Except for cases of multiple voting (controlled in 1999 with indelible ink), this had been true in previous elections as well. In fact, there had been no need for wholesale rigging at the polling station, because the desired result could easily be constructed when the polling station results were aggregated at sub-district and district levels, out of sight of prying eyes.
Since the non-transparent procedure for aggregating results remained unchanged in 1999, observers believed that the most serious attempts to subvert the election would most likely take place during the aggregation phase. To forestall cheating during this vulnerable period, several EMOs independently decided to conduct a parallel vote tabulation (PVT) by separately aggregating the results from polling stations covered by their observers.
Two PVT strategies were used. The first, adopted by Rectors Forum, was to conduct a statistical quick count, sampling the results from about a thousand randomly selected polling stations, and from that data projecting the final results. Technical assistance for this PVT was provided by the National Democratic Institute for International Affairs (NDI).
The second strategy, adopted by all of the other groups, was to conduct what could be termed a comprehensive PVT, where the observation group attempts to collect and aggregate the results from most or all of the polling stations in a particular constituency. The feeling was that although a quick count could be accurate, the procedure was non-transparent and complex, so it would be less effective in convincing people of the integrity and credibility of the count. The process of collecting the information for the PVT (and other monitoring data)and sharing it among the various EMOs was facilitated by a jointly-developed procedural manual and report form developed and published with technical financial assistance from the Asia Foundation.
These strategies were spectacularly successful. By recording and separately aggregating the results from polling stations, the EMOs were able to determine with a high degree of accuracy the expected outcome in any constituency. Since the EMOs were much more credible in the eyes of the population than the election administrators, there was no question that doctored results could be accepted as accurate. Although the aggregation of the data took a remarkably long time (even for the quick count), its mere existence, and the knowledge that it would eventually be aggregated and made public, seems to have been effective in deterring malfeasance.
For seven weeks after the election the national election commission delayed announcing the official outcome, claiming that they were still waiting for results from some districts. Some suggested that a desperate search was on to find a way to modify the results in favor of entrenched interests, but if so, the attempt was unsuccessful, and the official results that eventually were announced matched very closely those provided by the EMOs.
Political Party Monitors and Citizen Observers
The election law allowed the participation at every polling station of one monitor from each competing party. Because of the literal transparency of the polling stations, ordinary citizens could also observe the process, but for several reasons, neither groups contribution to ensuring the integrity of the process matched that of the non-partisan domestic observers.
The political parties were able to observe effectively in areas where their party was strong, and by carefully aggregating these results, they generally had a pretty good idea of the number of seats they had won well before the official results were out, but since their was little or no sharing of information between the parties, they had much less an idea of what had occurred outside their strongholds. The agents, because they were overtly partisan, were also less credible that non-partisan observers when making complaints or allegations against other parties.
Most of the foreign observers of the Indonesian election were struck by the excitement and enthusiasm of ordinary citizens as they gathered to watch the polling from beginning to end. There presence in such large numbers graphically demonstrated public support for a fair election process, and their impact in ensuring the overall integrity of the process should not be discounted. Nevertheless, by their presence, they did not obviate the need for organized observation.
Because the procedure was simple and transparent, ordinary, engaged citizens could ensure that the process at their polling station was honest and fair. Unfortunately, without an organized system for collecting and compiling information, they had no way to determine if the final result delivered by an aggregation process that took place far from prying eyes would accurately reflect the results generated at their polling station.
Conclusions and Lessons Learned
From a traditional point of view, there were simply not enough safeguards in the election process to ensure a fair election, yet the Indonesian National Election of 1999 is viewed by many as one of the freest and fairest transitional elections of the entire decade. If the integrity of the election was not assured by the legislative framework, judicial oversight, or neutral election administration, how was it assured? In my opinion, the credibility and integrity of the election rests on three pillars:
- The Polling and Counting Procedures
IIndonesia was blessed, perhaps accidentally, with a simple and totally transparent polling procedure that allowed the scrutiny of civil society (observers, party agents, and ordinary citizens) to make up for the lack of a neutral administration, and the absence of special security features in the process and materials. This inexpensive and transparent procedure, with minor modification, could serve as a model for sustainable election systems in other transitional democracies.
- The Indelible Ink
AA poor registration system meant that the only check on multiple voting was the marking of fingers with indelible ink. The addition of this safeguard was promoted by both the EMOs and the international community, and the ink itself was supplied by an international donor. Provision of ink may be one of the most cost-effective commodity contributions to improving the integrity and credibility of transitional elections.
- The Domestic Observers
Although there was tremendous momentum for change in Indonesian society in the months leading up to the 1999 election, without the widespread involvement of EMOs in monitoring the election process it is certainly possible, and probably likely, that entrenched interests would have attempted wide scale fraud. By closely monitoring the election process and the count, the EMOs were able to deter fraud, and increase confidence in and acceptance of the results.
In conclusion, the Indonesian experience suggests that in a transitional election characterized by weak administration and a lack of procedural safeguards, committed citizens organized by election monitoring NGOs can be an important or critical element in ensuring the overall integrity of the election process. By observing and reporting on every step of the process, EMOs were able to exert considerable pressure on both political parties and election administrators, increasing responsiveness and accountability.
