The significance in both the international and domestic public eye of the elections, following as they did 32 years of elections with a high degree of coercion, and low degrees of freedom and transparency make a case study of the count especially relevant. Mechanisms were chosen to highlight the transparency and correctness of these elections that could serve as a model for other transitional elections.
On the other hand, the notorious problems of the consolidation of the results, will be brought up as a cautionary reminder of the importance of planning and training as carefully for the count and results process, as for the polling process.
The 1999 Indonesian transitional elections do not reflect procedurally the conduct of earlier elections, nor do they reflect necessarily the conduct of subsequent elections.
The main differences with previous elections were the increase in political participation (48 parties compared with 3 the 32 previous years), and the transparency built into the procedures. Accredited national and international observers were permitted, in unlimited numbers, to watch both the voting and the count. In addition, party agents (maximum one per party), were not only free to observe, but also to challenge elements of the polling or count.
The 7 June 1999 elections were for 3 levels - district, provincial, and regional legislatures. For the national elections, a total of 462 seats were at stake (with 38 reserved for the military and police, making a total of 500). These 462 seats were apportioned to the provinces on the basis of population, but with some favouring of the more sparsely populated outer islands over the densely populated Java.
Polling was conducted from 8 am to 2 pm in approximately 300,000 polling stations across the country - no mean logistical feat in a country of 17,000 islands, some mountainous, some densely forested.
Following the closing of the polls, votes were counted at the polling station, by the polling station officials. As 80% of Indonesia ́s polling stations were outdoors, there was ample opportunity for not only the observers and party agents, but also the general public to watch (which they did! - the count across the country was characterised by a carnival like enthusiasm on the part of the onlookers).
The Voting Station Count
For the sake of this case study, a random sample polling station count is used to exemplify certain stages of the counting process, in this a village polling station in Sorkom Sub-District, North Sumatra.
Before opening the ballot boxes, the polling station officials prefilled the Official Reconciliation Forms to the extent possible (see example model C1 form). These forms were in the form of rather unwieldy, large books, a necessity considering that there were not only 48 parties times three elections, but also that all the polling station officials, as well as all the party agents present were to sign, endorsing the correctness of the documents. The C1 book/forms were to be filled out in 3 copies for the station, as well, each party agent had a copy.
The first entry on the form Ballots Received should have been filled out already at the opening of the polls. The ballot papers for the polling station were sealed in inside the ballot boxes upon receipt, and were to be opened, counted, and recorded as part of polling station opening procedures, in full view of all present, including the first voters the day.
The ballots received total should equal the amount of registered voters at the polling station, plus a 3% contingency (for spoiled ballots, or for persons added to the register with valid registration cards, or for persons registered elsewhere with a special authorisation form). The first ballot box, for national elections, was then opened, and ballots counted. The Indonesian 1999 ballots was small-poster size, and folded in a special way to ensure secrecy of the ballot. An official ballot was one signed by the polling station chairman, the vice-chairman, and a third polling station member as well as affixed with an official hologram sticker.
The three levels of ballots had three different colours for differentiation; the ballot boxes were in corresponding colours. When the first ballot box was opened, a number of wrong colour ballots had crept in - a common problem for the count when several elections are held at the same time. With the approval of all the party agents present, the misplaced ballots were placed in the correct boxes. The expectation would be of course that any missing ballots for this count would be found in the boxes to be opened subsequently.
For this reason, the official reconciliation form results at each polling station were not to be filled in until the end of the count of all three boxes. Back to the opening of the ballot box - the first task upon opening was to sort out the most obvious invalid ballots, that is the unofficial lines, that were missing the three signatures or the official hologram, fake ballots, photocopies ballots, ballots larger or smaller than the official ballots, or ballots in a different colour, which were to be removed and not counted, should any exist.
The official ballots were then counted, the total noted, hopefully matching the amount received, minus unused and minus spoiled, and minus unofficial/invalid. If the numbers did not reconcile, chances are that this was because ballots for this election were still in the other boxes, and yet to be counted. One by one, in front of all party agents, observers, and the general public, the ballots were unfolded, held up in full view, and the party called out by the chairman, and tallied on a large visible boards.
The party agents, and observers if they so wished, had forms to do a parallel tally at the same time. Interestingly for the Indonesian elections, ballots were punched with a hole rather than marked. Because of this, the back of the ballots was shown to all present, so that the hole would be clearly visible. As the ballot papers were thin, translucent, and the party symbols bright and easily visible from the back of the paper, this was fine except had there been worse lighting, (mostly the count was done in the daytime, between 2-6 pm, but in some places delayed materials resulted in delayed counts, meaning the count was conducted in poor lighting) this could have been a problem.
When voters had not properly punched the cards, but rather just made a dent in the paper, these were considered valid if so approved by all the representatives present. Ballots marked with a pen, punched more than once, unclearly punched (i.e. on the line between two parties), or punched were considered invalid, were set aside along with the unofficial ballots already removed.
The chairman made special effort with the invalid ballots to make sure that the invalidity was approved by all. This strengthened the sense of transparency, good will, and participation. Valid votes were both tallied during the count, and the ballots divided into piles, subsequently counted and checked against the large tally board. The parallel count by the party agents and observers present served as a triple check for accuracy.
Following the tally, and the determination of votes per party, the ballots were rubber banded together on a party basis, and placed into an official envelope that was then sealed. Invalid ballots, were placed into another envelope, recorded and sealed. These two envelopes, plus the envelope with the spoiled ballots were placed in a plastic bag. This procedure was repeated for the provincial and district level ballots.
After the ballots were tallied for each election, the polling station chair asked the vice chair to count the number of votes for each party, for each election, on each tally sheet. The chair announced the total number of votes for each party so that party agents and observers could hear. The vice chair and another polling station member then record the number of votes for each party on the official reconciliation form (3 copies). Another polling station member recorded the numbers at the same time on a large, visible form on the board.
The three copies were signed by all polling station officials and all party agents present. One form was to be kept sealed, with the other election material, another to be posted publicly at the polling station, and the third to be given to the village level election committee for the consolidation of the results.
The sealed envelopes were locked inside one of the ballot boxes, other materials placed in a second, and all the materials accompanied by the polling officials, party agents, observers, and general public, to the location designated by the village level election committee for safekeeping overnight.
The Consolidation of the Results
The village election committees, known as the PPS, were each responsible for 2 to 9 polling stations, depending on the population of the area. The Indonesian elections were organised based on a hierarchy of 6 (!) levels: (1) the national election committee and commission (KPU/PPI), (2) Provincial (PPD1), (3) Regency/Municipality (PPD2), (4) District (PPK), (5) Sub-district/Village (PPS) and (6) polling station levels.
Official reconciliation forms were consolidated at each level - that is, the PPS (Village) consolidated its polling station results, which were sent to the Sub-district (PPK) level. The District level consolidated and sent official forms to the next level up (Municipality), but also sent the results directly to an electronic parallel counting centre in Jakarta, run largely by internationals (Australia). The official manual results, following consolidation at the provincial and national levels was scheduled to take approximately 2 weeks, followed by a legalization process and official announcement. The computerised count, though unofficial, provided considerably quicker results. Certain observer groups also conducted parallel tallies.
Under the current electoral law, candidates are assigned seats on a constituency basis (the constituency, but within a proportional representation system). List of candidates for each party/constituency are posted in advance.
The use of reconciliation of numbers is a very important feature to consider on the election forms, and yet one area that many election officials seemed to have difficulty with during the election. Using the sample polling station as an example, the results were spread over 11 parties, with 2 parties neck in neck, resulting in a marginal victory for Golkar, the party that only a few years before had earned over 90% of the vote. This gives an indication of the changes that a new electoral law, free campaigning, and clear and transparent procedures can bring to a small village.
