Electoral Governance in Papua New Guinea: Chimbu Poll Diary, June 2002
Bill Standish - 28 June 2002
'Election 2002' is still under way in Papua New Guinea as this is written. Two major changes to electoral governance are being trialled in PNG. First, the Organic Law on Provincial and Local level Government 1995 requires that elections be held for Local Level Government council wards at the same time as elections for the National Parliament. The second major reform is the coming into operation of the Organic Law on the Integrity of Political Parties and Candidates (Integrity Law) 2000, which is designed to strengthen political parties and thus create stability in government. The Integrity Law's main effects will only be tested through the life of the parliament after the election process is over and parties have formed the next government, but this report will not explore that topic.
A further reform, the 2002 amendments to the Organic Law on National and Local Level Government Elections 1997, will introduce a system of Limited Preferential Voting (LPV) in any by-election held after the Return of the Writs on 15 July. How this will operate is still not clear, and there have been no public awareness programmes on the changes, although there was considerable publicity urging the case for Limited Preference Voting last year. While the fieldwork reported here carries implications for this change, especially the difficulty faced by officials in managing complex electoral processes, that will be discussed in a later report which covers the count.
In the central Highlands province of Chimbu there are about 260 candidates for the six Open seats in Parliament, including nine women, and 55 candidates for the Regional seat which covers the entire province, two of them women. I do not yet have a full listing of the thousands of candidates for the hundreds of council Wards in Chimbu. Aspirants for the councils include 28 women, some of whom had withdrawn from the parliamentary race. A few council wards are contested seriously, but usually these are seen as low key local races compared to the bitter feuds involved in parliamentary elections. Indeed, there are often hundreds of council ballot papers left over in the polling places after all those for parliament have been used up. Here I focus on the National Parliament contest.
One effect of combining the two elections has been to divide the electorate administratively into hundreds of council wards and thus to fragment villagers used to voting in larger units. Simbu society is already divided by personal, clan and tribal rivalries. Aside from the obvious administrative complexity, the decision to combine the polls has increased tension levels in a politically volatile region, and requires further analysis and review.
Free and fair elections?
This paper examines the elements of electoral governance designed to ensure free and fair elections. Democratic elections require real competition and choice between candidates and parties, free discussion of alternative policies, confidentiality of individual voters' choices and freedom from intimidation. In other words, democracy requires competition, the promotion of civil and political liberties and political participation (Diamond et al, 1988) Legally speaking, it is compulsory for adults over 18 years to enrol in PNG. Voting itself is not compulsory, although villagers are under strong pressure to vote if there is a local candidate.
In his 1997 Report the Electoral Commissioner stated that an accurate register of electors is 'absolutely essential' for free and fair elections, but added 'Compiling an accurate register of voters in Papua New Guinea … continues to be a seemingly impossible task …Attitudes have to change in the country before satisfactorily accurate roll of electors can be compiled.' Inflated rolls 'introduce a high probability of cheating at the polls', he reported (PNGEC 1997:3,5). Other mechanisms to prevent electoral fraud, especially double or multiple voting, include the dyeing of the left hand little fingernail with what is intended to be indelible ink. In the Chimbu Province there is plenty of choice for voters, and many of the 43 registered parties have endorsed candidates, yet most of the techniques designed to ensure a free and fair election were severely compromised during the polling.
Simbu people commenting on the polls themselves report the abuse of the election process by their fellow, with the encouragement of candidates who seem willing to do 'whatever it takes' in order to win. This follows the findings of PNG commentators (such as Gelu, 2000), as well as the Electoral Commissioner himself, who has commented on 'the level of electoral violence, the role of money and bribery in the electoral process, the deliberate and determined effort of some would-be candidates to tamper with the enrolment process, and so on' (PNGEC 1997: 20.). Electoral staff, nominally under the control of the Electoral Commission, have been unable to prevent electoral malpractice, if not actually colluding in the abuse of process. Nevertheless, Simbu observers show distress at 'foul play' in the electoral system, and some have already expressed their rejection of the outcome of the final count.
National wide administrative problems
PNG media reports indicate that there are serious administrative problems across the entire country in this election, which the Prime Minister has described as 'chaos'. The main problems have been with the Common Roll, which is intended to prevent double-voting but has also omitted many eligible people who participated in the updating exercise. The Electoral Commission blames late and inadequate funding for the delays in updating the roll, the Prime Minister blames the Electoral Commission The Electoral Commissioner has acknowledged that the roll is imperfect, with many false or duplicated 'ghost' names, even though some 1,566,832 names were eliminated from the working draft (Post-Courier, 6 June 2002). However media focus has been on the urban and coastal areas where many people have been unable to find their names and been denied the vote. On 17 June at the end of scheduled voting in the capital the polling had to be extended for a day in order to deal with crowds, especially in Moresby Northeast. Polling has been quiet in most coastal areas, despite a few allegations of tampering with ballot boxes in Port Moresby.
By contrast, the local press and TV have reported some violence and intimidation of voters and officials in Eastern Highlands. In the Western Highlands several ballot papers and boxes have been hijacked, three officials kidnapped and another seriously wounded. The Regional member, Fr Robert Lak, has spoken of police inactivity in polling areas where there has been widespread intimidation, a point reiterated on EMTV by a Mount Hagen woman voter. I stress that while I have seen people pressured in their voting, I have seen no physical violence except in the Kundiawa police station. As in previous elections, multiple voting has been rife (only discussed in The National 28 June 2002, for the Eastern Highlands). Polling in Enga and the Southern Highlands had yet to start at the time of writing, but senior police officers say they are expecting trouble.
From news reports it would appear that in some coastal and island regions of PNG officials have been able to maintain the controls built into the electoral system. This report does not claim to present the pattern across the entire country, but the Chimbu mode of voting seems typical of the poll so far in the Highlands region, which supplies 39 of the 109 Members of Parliament (MPs). That in itself justifies a detailed examination of methods by which many of them are being elected.
In the months before the one day polling scheduled in Chimbu for 20 June, several candidates told me the Common Rolls were quite unreliable. In Kundiawa Open seat, for instance, several clan and sub-clan groups (numbering thousands or hundreds of people) were located on the printed roll in the wrong council Ward and at the wrong polling place. Some are in the wrong Open electorate: e.g., hundreds of Yonggamugl people are located in the Kamanegu tribe (in Waiye Census Division, now Waiye LLG, in the area north and west of the junction of the Wahgi and Chimbu Rivers), a mistake repeated since elections in 1987. This error was one complaint in an unsuccessful Court of Disputed Returns appeal in 1992, after district administration officials ('kiaps') simply moved the wrongly located pages from Waiye CD in Kundiawa Open into the roll for Yonggamugl CD in Sinasina-Yonggamugl Open.
I was told by two candidates (one a likely beneficiary) that there are only 3-400 eligible voters in Kombaku clan of Naregu ('Narku') tribe, but the rolls show 1,881, and wrongly placed them in the NAURU tribe. This particular inflation of the roll occurred after the initial update of the roll in January-February 2002, when only 400 enrolment claims were made. (The figure is rounded because forms were issued in 100s.) The 'corrected' rolls were printed and released in mid-June, listing 1,881 voters for Kombaku. After discussion with various candidates several changes were made. Electoral officials agreed to let the Kombaku sub-clan vote in Naregu rather than NAURU, and accordingly 1,881 ballot papers were delivered to Naregu. This was seen among those people as a victory for their leading candidate.
Many people complained as polling got under way that the alphabetical listing of individuals (whose names may be spelt in various ways), which started in 1997, was disastrous. For 60 years people have been censussed in family groups and voted the same way for over 40 years. It is a pity computers cannot understand family groupings, they say. Many names were missing from the rolls, but also there could be three or four combinations of most people's names inserted in the update, and - if they wished - they could vote accordingly. The roll was simply not trusted as a control measure, and if used to deny people the vote in rural areas could be overcome by sheer pressure from the exuberant crowd. That is what happened in rural Chimbu.
People told me various strategies can be used to inflate the rolls and a local candidate's possible core or 'base vote' of clan or tribal members. One is to say a large number of people have migrated in to an area (to gain access to better land and services). Simbu observers say the most efficient technique is to create non-existent or ghost 'rest houses'. These were the colonial era huts for visiting administration patrols and meeting areas for clans. Each new 'rest house' gives scope to add thousands of names. This process is believed to have led to the election with several thousand votes of one Simbu MP in both 1992 and 1997, although his clan has only about 900 people. It is also said to have been applied by a prominent Kundiawa candidate in 2002, where the actual voters come from a closely allied group in the Kerowagi electorate. If these allegations are true then electoral fraud has reached a truly massive scale. With the turnover of staff following the 1995 provincial 'reforms' district administration staff clearly do not know their areas well, but such large scale manipulation required the collusion of at least some of those involved in the updating process.
Roll update poorly managed
Across the country the rolls update was only conducted in early 2002, due to late release of funds. It was insufficiently resourced and in some provinces poorly managed. Even before the printed rolls became available, candidates said they knew exactly how many voters names were registered in their claimed base vote areas. The update, supervised by the local Electoral Commission (EC) Election Manager, was run by people appointed by district administration staff, some poorly trained, with casual workers such as school leavers and former public servants often working in their own clan areas. Deputy Electoral Commissioner Andrew Trawen knew that nationally the updated roll had flaws; he told me that very few people were reported from Highlands provinces as having died, whereas the death rate in the coastal areas was normal. As in the national census, growth rates in Highlands provinces were extraordinary - about 0.35 per cent annually. 'Unbelievable', smiled Mr Trawen in May. It was argued that during 'Census 2000' that people had thought they were doing the update for the common roll, although these are separate processes.
Like the national census, the Common Roll update was not a head count or home visit census. The recorders simply collected names, and when they got tired let school children or leavers do the job. It is alleged (and certainly possible) that these additional workers were funded/bribed or inspired by local political aspirants to inflate the rolls. Young people have gleefully told me they have done that for previous elections, using very imaginative ghost names. Last April one man told me in front of 30 fellow clans people that his team had lifted the enrolled number of his own tribe from 9,000 eligible voters to 11,000. Along with those under 18, this approximately doubles the actual population. The claim forms were not vetted within the province, and the entering of the data and attempts to remove duplication were done by teams working double shifts in Port Moresby.
Two candidates, past and present MPs, told me the race would be decided by the number of names on the printed roll, not the number of people. The printed rolls arrived only a week before the election, scheduled for 20 June, and could not be checked. 'This roll is a disaster' was a frequent comment in Simbu political circles, even by some who would benefit from its weakness.
How to make it work in future?
Thoughtful Simbu are wondering how the roll can be made to work. It is clearly impossible to have a perfect roll if the citizens seek to manipulate it for their own ends. Interestingly, in South Africa in 1994 there was no time to prepare a roll, and voting was treated as a solemn right and ritual, at least on that occasion. Some local observers and even candidates in Chimbu have started discussing whether there is another way of achieving a fair election without using a roll. One suggestion gaining currency is that voting should be based upon the showing of receipts for payment of (reintroduced) head taxes. Unfortunately, the old council taxes were also subject to corrupt exemptions, could easily be avoided, and by 1975 the colonial administration lacked the capacity to collect them. Sadly, receipts for taxes are likely to be lost, and the tax teams issuing them and polling teams checking them could also be subject to corrupting pressure from voters or candidates. Proponents argue that the incentive to pay tax would be the opportunity to vote. They say that the benefits in avoiding the present abuses would outweigh the removal of the universal right to choose one's legislators. An old proposal to issue photographic identity cards is also gaining some currency. However that would require complex technology and is expensive, and could also be subject to corrupt manipulation by those with the motivation.
According to prominent candidates, in 2002 the main change, heightened from 1997, is the 'blocking up' of candidate's votes, which for Open seats are usually located in candidate's natal clan, in-laws and matrikin. This blocking process has two main dimensions.
First, candidates seek to ensure the unity of their clan or tribe, emphasising traditions of unity in peace and war. Since the first council votes in 1959 Simbu has had 'haus lain' (village or clan) consensus based voting, in effect a clan-based pre-selection or pre-poll. Often in the past - when there were slightly fewer candidates - Simbu people have presented over 90 percent voting solidarity in ballot boxes. That means that least 90 per cent of all those in each polling place, men and women (and some children!), vote for the same candidate. Such block voting is an assertion of a clan's strength, but it restricts the freedom of choice of all, especially the married women - who all were born into different clans. This helps explain why very strong women candidates have received very low votes in Simbu in the past. The women's solidarity with their menfolk has also been ascribed in some instances to pressure (including actual or threatened violence) from their husbands or fellow clansmen.
In addition, some entrepreneurial types have traded blocks of votes (sometimes with lists of names) to other candidates, and especially where there was no Regional candidate from their immediate area. Open candidates have also traded with Regional candidates, swapping Regional votes for support in Open contests. These tactics have been reported to me in 2002.
Both forms of block voting are only possible with the collaboration of clan members, but sometimes their participation is reinforced by pressure, a combination of threats of violence and positive rewards, or the obligation created by hospitality in the year running up to the election itself. Usually at each election the number of candidates has risen, driven by personal ambition and local rivalries, which fragments the potential for large voting bocks and can increase community tension as candidates compete for the same clan or tribal constituency, known as their 'base vote'.
A second aspect of block voting in 2002 is that candidates seek to prevent others from poaching their base vote. In the past it could be difficult for candidates to campaign in another's base area, and by 1997 it was unusual. In 2002 it was almost impossible and certainly dangerous, especially for Open candidates. Block campaigning preventing the entry of anyone who might sway voters with sweet talk, money, gifts and food, and this in turn could lead to division within clans. People were already hyped up in April, ready to repel all visitors. There was fighting between rival candidates' teams in Kundiawa in the nomination week. Two candidates were shot at early in the campaign, when outside their home base, and several vehicles had their windscreens smashed as a deterrent to campaign visits. This practice is clearly not consistent with those of a liberal democracy. The Bomaikane clan of Kamanegu tribe, based at Kurumugl in the Singganigle valley of Waiye LLG, is one such case of a clan with two candidates where fighting erupted soon after polling (see below)
Escalation of intimidation
The current situation is a progressive escalation of intimidation which started in 1977. In Chimbu the 1992 (national) poll and the 1993 (provincial) poll, especially, involved the frequent display and firing of firearms to impress and intimidate a candidates' own base group and his rivals. Officials reported that in 1993 guns and axes were used to force them to allow voters through the polls many times, the voters sometimes using stocks of stolen ballot papers (Standish 1996). A Simbu polling official told me he had had the same experience in 1997. After every Chimbu election since the provincial assembly vote in 1980 there has been fighting and house burning as payback for undelivered votes or alleged disloyalty. In 1997 there was gun warfare within NAURU tribe (5 killed), and in Dom (with 30 killed and about 1000 refugees driven off their land). Simbu people, especially women, are currently predicting similar recriminations by losers in 2002, and this fear clearly restricts the freedom of choice (see below).
By contrast, some Regional candidates were able to campaign relatively freely around their much larger electorate, because Regional seats - which will disappear in 2007 - are now seen as relatively insignificant, a rabis vot. Some Regional candidates said they had campaigned far and wide, although one prominent Regional candidate told me he had only campaigned in his base areas, across parts of two districts, and that he could afford to ignore the rest of the province.
As noted, the relatively smooth process of clan voting is endangered when large groups have several candidates standing, often on the ambition of the individual, but sometimes out of jealousy to block another rising star. But it is also alleged that these are externally funded vote splitters. Although clan groups share traditions of common descent, and many clans are clustered in tribes said to have descended from brothers. In several cases large groups are really loose confederations of clans ambitiously 'tribalised' by early colonial officials. As such, they are easily divided and internal conflict is likely. In the two major clans of Dingga group in Sinasina, based at Emai, tensions have been especially high this election. There are three Dingga men standing, two of them powerful businessman, and two Dingga women's activists are also standing. Several people within this group have told me that there has been heavy intimidation of voters by well armed henchmen and that the Dingga group had to vote at four separate locations this time in an attempt to prevent internal conflict.
In general, given the prolonged economic slump and low expectations, campaigning was not as expensive or showy this year as in previous elections, but there were some big spenders. One candidate in NAURU told me that he had paid K2-300,000 in compensation for deaths and property damage arising from the 1997 post election fight within the group. One Sinasina candidate admits to spending K500,000 on his campaign, after a lifetime of preparation. He said he had not paid for the 800 pigs slaughtered as part of his campaign feasting, but clearly he has incurred considerable debt from these events. By contrast, two women candidates costed their relatively modest campaigns at around K30,000, this in an area where average cash incomes would be less than K500 per annum.
Parties in decline
Political parties do not seem to have been significant in the campaign, although the People's Democratic Movement (PDM, formed by Paias Wingti) is so unpopular that people say they will vote against any PDM candidate. Most candidates were Independents. There were major rallies organised by Peoples Labour Party (led Peter Yama the wealthy former member from Usino-Bundi in Madang Province). There were fleeting visits by Sir Michael Somare (National Alliance) and Sir Mekere Morauta (the PM and parliamentary leader of Paias Wingti's PDM), and Lady Carol Kidu (of the Melanesian Alliance). Basically, as in previous elections, candidates sought some form of party endorsement hoping to gain a refund if they won, but usually they had no claim without any long-standing commitment to the party. The parties themselves lacked the staff to assess the quality of the candidates on offer, and some good candidates shopped around after being rejected, or simply ignored, by the party of their first choice. A former PDM organiser for the Highlands could not even name his party's nominees in some seats, and was himself running in Kundiawa Open as an Independent.
Very few candidates mentioned their parties' names, and even fewer their policies, in discussion with me. The election was about competition between traditional groups and alliances, and individual patrons and their enthusiastic close followers, usually known as 'supporters'. Parties seem to be weaker now than in 1992, a decline which one former MP described as having occurred in 1997.
The independence and transparency of electoral governance largely depends on the quality of the personnel enlisted. I do not know much about the processes behind the choice by District Administration officials of the Provincial Returning Officer (PRO) and Assistant Returning Officers (AROs) and the composition of each polling team - especially the Presiding Officer - for each polling place. Some of these officials have worked in the province for a decade, and most are Simbu people with strong links into the wider community. Electoral staffing was common knowledge; lists naming over 100 teams and 600 officials were posted on a notice board outside the Provincial Office. Although many are public servants or teachers, others were previously unemployed casuals. All wore white Electoral Commission T-shirts.
Just before the voting started, several candidates wrote formal submissions to the Provincial RO and AROs, complaining that polling teams in their areas were closely allied to particular named candidates. In one such letter threats were made in writing that a team would be ejected from the intended polling place because of its alleged bias. Allegations of bribery are frequently rife in Simbu, and the normal high level of suspicion is heightened during elections, damaging the credibility of the entire democratic process. Some electoral staff made no attempt to keep their distance from candidates. When the PRO on 21 June reminded his teams of the legal requirement for neutrality and the political need to maintain the integrity of the poll he was cheered by the nervous men - and a few women - waiting to go out to meet the waiting voters.
Distribution of ballot papers
The printed papers (with some mistakes such as duplicated photos, or the wrong party endorsement for candidates) arrived only on 14 June. After reprints some papers reached Chimbu on 18 June or later. Many more papers arrived than there were names on the rolls. Some of the Local Level Government (council) ballot papers were photocopied commercially in Kundiawa, the rest printed in Goroka by a commercial printer. The ballot papers (for the National Parliament) are meant to carry a water mark, and all carry serial numbers and butts (not unlike huge raffle tickets).
On 19 June the polling teams started pre-counting or checking their papers in the area behind the Kundiawa police station. Throughout this process there was little security at the compound where ballot papers are kept in shipping containers, as at each election. Neither police nor electoral officials seemed to see security of this area as their responsibility. A police janitor minded the padlock on the gate much of the time; a dwarf, he was taunted by the watching crowd. Almost anyone, schoolboys and even candidates, but not women, wandered freely into the area where papers and boxes were being distributed. At times frustrated police threw stones at the crowd to drive them away, but security remained slack until boxes filled with votes started returning on 21 June and were redeposited in the containers.
Late on Wednesday 19 June officials started the actual distribution of papers to teams so that the Open, Regional and LLG council ballot papers matched the numbers of voters on the rolls for each polling place. This was a complex and slow process, delayed by unseasonal heavy rain from late afternoon throughout the night. Only one room was available but no other shelter. On 21 June two tarpaulins were erected before a second all night shift sorting papers. It may not have occurred, but there was ample opportunity for malfeasance throughout this procedure.
Village people had been up all night in men's (campaign) houses during the last week of the campaign and were tense with anticipation for polling on Thursday 20 June. Then at 7.15 p.m. that day the PRO Steven Yakali went on provincial radio and announced that polling would start on 21 June. He apologised for the delay caused by late arrival of electoral materials and additional police from other areas. At 5p.m. on 20 June the first vehicles overladen with ballot papers, ballot boxes and staff left for Salt-Nomane, an eight hour drive on dangerously deteriorated roads. For protection against the weather, the staff had only a few large sheets of plastic and without proper camping gear.
The allocation of voting papers started late on the intended polling day in the case of Sinasina Open. The process continued through that night. That evening one polling team allocated to the Emai village area spent the night at the motel run by the wealthiest candidate for the Dingga tribe. After 6 p.m. I saw one poll clerk arrive at the motel with a sheet of paper which appeared to contain the numbers of the allocated ballot boxes for Emai, and the relevant numbers of the ballot papers. He gave these to the candidate, saying 'Here are the numbers', and the motel manager immediately opened up the office computer and set to work on these figures. I do not know precisely what was happening. This polling team, including its Presiding Officer, was also based at the same motel the next night, and were in close contact with the candidate through the next several days.
On the Saturday morning 22 June this same man was with a group of other candidates, furiously shouting at the ARO for Sinasina, demanding that the ballot papers be taken up to the Emai polling areas on an impassible road through his own clan's territory. That day failed to get ballot papers to Emai via Koge along a barely passible road, with an unarmed police escort. Road access was blocked by people fearing another heavily armed candidate further up the road. The electoral team driver was beaten up by supporters of one candidate and withdrew to his village to recover. One police vehicle slid off the greasy road.
Most of the rest of the polling teams had left Kundiawa during the morning of 21 June, without any police escort. Teams sent to Gumine were unable to get their votes out until the rain slowed on Thursday 27 June. Ballot papers and boxes sent to Karimui by fixed wing aircraft were not delivered to polling places until Tuesday and Wednesday 25-26 June. This was done by a Defence Force Iroquois, which left the province on Wednesday 26 June. A chopper from Pacific Helicopters had delivered papers over the weekend, but then ceased operations after non-payment of the account by the Electoral Commission. Another firm, Hevilift, had taken payment from the police (using K0.5m AusAID funds) but would only carry police personnel and not election materials, to the fury of senior police.
In Kup CD, at Gamar on the Western Highlands border, it was rumoured several thousand ballot papers and associated boxes were held up by the supporters of an Open candidate. One locally-based Regional candidate reportedly told an Open candidate from the South Wall of Wahgi that he was holding three thousand Open votes and would offer them to any Open candidate who after voting appeared to have 6,000 votes and was likely to win. In NAURU tribe, some 1,800 papers were allegedly taken from an unarmed patrol by a candidate. His rival complained loudly at the police station and a further 1 800 papers were sent to his area to be filled in. He allowed a number - perhaps 100 - of these to be filled in for the man who allegedly took the original 1,800. The AROs received constant demands for additional papers from similar incidents, without the capacity to prove or disprove the claims.
Rain delayed play. The downpour on 20 June had rendered most roads in the southern Gumine district impassible and delayed polling there by a week. The two day exercise took 12 days, with K100 a day traveling allowances accumulating for each staff member, which like helicopter hire seriously blew the budget. In northern Chimbu steady light rain started about 11a.m. on 21 June and continued throughout the day, severely reducing any chance of orderly polling. Most booths did not have any police presence, although a mobile squad moved up and down the main Highlands highway. Nor did teams have rain shelter. Several started operating in local men's houses, which are often used in the campaigning, or inpermanent material houses. School buildings were not used for fear of damage.
Polling in Kundiawa town did not start until about 11a.m. on 21 June and on the surface was orderly, initially with attempts made to use the roll, and finger dye. However people were openly using household bleach to remove the dye so as to be able to vote again. One supporter stood in the polling area visibly holding a bottle of bleach. (Two supermarkets had big displays of bleach at their entrance that week!) Given the ineffectiveness of the dye under such circumstances, its use was abandoned by mid afternoon and multiple voting continued. At the Old Nurses Quarters (Ward 3) at about 5.30 p.m. a heavily armed police squad came and saw evidence of double-voting. They pointed their weapons at polling officials and voters, grabbed all the remaining ballot papers (about 2 000) and burnt them on the spot.
At Premier's Hill in town earlier in the afternoon a prominent Open candidate had pressured the Presiding Officer (PO) to allow multiple voting by his supporters. According to the official, the candidate showed a pistol in his jacket pocket and forced the clerks to accept his wishes. The distressed PO told me at the police station that he was sickened at the sight of this man.
Naregu tribal area (Waiye LLG)
I travelled with polling teams to Mintima Village and observed voting at eight polling places around Naregu tribal territory. After much confusion and shouting at Bamugl, home base of the Burukngaumo clan, they learnt they had to vote one kilometre away at Kuglame on the highway. Some sub-clans were separated from their traditionally paired sub-clans, and in one case Gamgane people were asked to vote at Nende, 3 km from their home territory. There they were refused access to vote, although all the allocated ballot papers were used.
Voting started about noon on 21 June. Polling areas were usually small roped off enclosures, 10 metres diameter, or less, located in old food gardens or house yards. Polling teams were lucky to have a folding table, and borrowed what they could from villagers. Some teams used the cardboard voting screens as tables. None had rain shelter, although during the day many borrowed umbrellas or used the cardboard booths to protect the papers.
At nearby Nokun there were 2 teams in a large open area, swamped by anxious voters. A local candidate moved within a metre of the voting area. For a few minutes, attempts were made to use the rolls, and also the finger dye, creating a log-jam. Light rain soon started , which increased the pressure from the excitable crowd. There were dozens of people inside each roped off area, including many supporters of candidates. Each voter marking 'X' on their papers was observed by 3-4 men at once. I saw no scrutineers' name tags all day. Someone picked up the bottles and threw them into the bush. Soon voters' names were simply being marked off the roll in sequence, and men and women were passing through the booth in a steady procession.
At Kugame 1.5 km up the highway there were two booths operating, the first one well closed off with people queueing in an orderly way; polling officials were apparently marking names off and using the finger dye in some cases. There were about 12 people in the polling area at any one time. Voters' LLG ballot papers were taken by an officer, a few words were exchanged and the paper marked and put in the ballot box. There was little interest from scrutineers or supporters in these ballots. For the Open and Regional papers the voter usually expressed a wish to a different poll clerk at a different voting compartment, who duly marked the paper witnessed by three supporters.
The second team at Kuglame operated independently some 30 metres away. There the process was much more pressured with about 20 people in the area, as well as voters. Finger dye was not being used, and there seemed little if any attempt to find names the rolls. Some women told me after they left that their fingers were unmarked and that they would go and vote again down the hill at Bamug. Like naughty schoolgirls they showed some delight at beating the system. Others from the same clan declined the opportunity. At this place papers were simply taken from the voters with few if any words exchanged and were marked by supporters, again watched in each case by several people. Officers put the papers into the boxes. When a police truck arrived the polling area almost emptied and only poll clerks 'assisted' the voters in the 10 minutes the police were there. The Catholic Bishop of the Mingende Mission came by and observed that finger painting was not occurring, and that a few village women in the general area had been seen removing pigs, a sign of fear of impending trouble.
At Nende, half a kilometre further west up the road, another candidate stood greeting the voting public in heavy rain. One booth 50 metres up the hill was in a men's house, with open windows and people watching from outside, another in a fibro-cement house. Hundreds of people queued in a very restricted area. At the men's house officials were holding people back with a waist high stick, and people came out from each end of it into the building. Officials called names, followed by 'man' or 'meri'. Only a sex tally sheet was being used to indicate numbers of people voting, not the names. People would rush forward to enter the men's house. Often two women went forward at once, and one had to be pulled back. People saw some humour in this. I saw several under-age adolescents - about 13 years - enter the booth.
Inside the building there was a queue of people and a real production line, with the presiding officer shaking his wrist every few minutes as he churned away signing the backs of wads of papers for the dozens waiting beside him. The voters, especially women, simply handed their papers over to young men (only some of whom were poll clerks) who duly marked the papers, without a word being exchanged. This was clearly not a free individual vote. At the second booth, a 4 by 5 metre room with one door only, officials merely called out (in Kuman language) 'man ' and 'woman', and again marked off the sex tally sheet but not the roll. Voting ceased when I entered this polling place.
I then walked 4 km East to Wandi village where there had been a minor disturbance and women were leaving the area. At about 4.15p.m., on a slight ridge beside the highway, two police were organising hundreds of people to line up in same sex queues in the voting area . Two teams were operating side by side. Names were not being called, nor was finger dye being used. Names were being crossed off the roll in sequence as people moved past the officials who gave out ballot papers. A policeman pulled one woman from the queue, loudly accused her of multiple voting, and gave her a long lecture before letting her go.
By this time it was about 4.30 pm, and the rain got heavier. There were no voting compartments in use, and most people simply crossed their papers with the aid of a friend or candidate's supporter. One Presiding Officer told people he would suspend polling while he asked the RO in Kundiawa for permission to continue voting on the morning of 22 June. Two Open candidates walked around within the voting area, which was not roped off. When I joked 'Candidate!' to one of them, he said he was there to vote in his own voting area. I left at 5.20 p.m., walked back 2 km to Nokun.
An eye witness the next day told Bamugl villagers that at about 6 p.m. on 21 June the same candidate had chased and angrily grabbed a voter, to whom - he shouted - he had given a lot of support and even an overseas trip. Now the ungrateful voter was supporting the other local man. He held the voter in a bear hug, and one of the candidate's supporters chopped the voter's foot with a bush knife. (One version, which I think is accurate, says he lost three toes.) Police did not hold the assailant but rather arrested the wounded man, whom they said was a much-wanted criminal, and for days he was held in the lockup at Kundiawa. Voting continued at Wandi well after dark (6.20 p.m.) and recommenced on 21 June until mid afternoon.
Back at Nokun, just on dusk, it was still raining. The rope fences had collapsed and there were several large groups of officials, voters and supporters, standing in the mud, huddled together under umbrellas or cardboard. For each team one official was marking names off the rolls and working through them, another was signing ballot papers en masse, and these papers were handed over to a third group (with a few poll clerks visible) desperately seeking shelter as they marked scores of ballot papers and stuffed them into the ballot boxes. About a 100 people watched. No rain stopped this play. I was later told this process continued, still with no marking of fingers, until all ballot papers were finished (about 10 p.m.). I left at 6.30pm, in the dark, and returned to Kundiawa 8 kilometres away, where it rained all night.
The entire process, because of the rain, had none of the festive atmosphere of polls in previous years. It became a matter of getting people as rapidly as possible through some sort of ritual of voting, apparently according to group consensus. The procedure was largely supervised and carried out by tough young men who were the candidates' supporters, as much as by the polling officials. Freedom of choice and privacy of the vote were impossible in these circumstances. Procedures such as checking the electoral rolls and using finger dye to prevent double-voting were abandoned within an hour, either willingly or under duress. The Bomaikane clan of Kamanegu tribe, based at Kurumugl in Singganigle valley of Waiye Local Level Government, is one such case of a clan with two candidates, which has already led to fighting.The only limit to the number of votes was the supply of papers. That was a relatively easy process.
Elsewhere, in both north and south Chimbu, many teams were stranded up side roads five days after the poll, along with the local candidates and their supporters. An ARO said they were being threatened that if they travelled down to the main roads their base votes of ballot boxes would be seized by the groups located nearer the district centre, and they needed to be extracted by helicopter. At Yobai in other areas, truckloads of supporters reinforced the police and followed their local heroes' precious cargo to town. Returning officers pleaded for more helicopter time, but by 27 June many teams were still in the bush. In Nomane district the PNGDF helicopter was swarmed by 18 officials, which would have been unable to take off at that altitude. The well-armed crew told me they had had to use some diplomacy persuading the hitch-hikers to leave, and barely took off laden with ballot boxes. Some security staff were disgusted by the entire operation. Senior police officers said that they were running the election now, due to the incompetence of the electoral officials and the boisterous public.
One police riot squad driver told me he had had to use his automatic rifle to shoot his way through the roadblock at Koge village, a crucial road junction for the road to most of Sinasina District and the entire Nomane district south across the Wahgi River. He also said that many electoral officials appeared to be in the pay of candidates, and, he suspected, some police were as well.
Gunpoint voting, and post election battles
At Emai, in Sinasina, candidates had been saying for days that one of their rivals was heavily armed. He had already intimidated polling teams on the Saturday before voting. Hundreds and probably thousands hundreds of ballot papers were handed over to various candidates' supporters, with the polling officials signing them in a locked house away from the designated voting area. These papers were then marked 'X' by two groups of supporters, by the hundred, for particular candidates. One woman candidate publicly withdrew from the poll under some pressure after a man, in his police uniform, had spoken in the polling area saying that another candidate had many thousands of votes already (the rumour is 7,000). The policeman, said that he wanted to avoid fighting within Dingga. This woman leader mentioned that she had spent K37,000 on her campaign and asked that this be taken into consideration by the leading male candidates in the race. Another woman candidate refused to withdraw and demanded 150 papers for her immediate supporters and family. One hundred signed papers were handed over and filled out. She herself could not vote.
At two other locations in Sinasina, according to women's activists, votes were sold, literally, for K5 at one location and K10 at another. These are very cheap votes, considering that some candidates will have spent K50 per potential voter, and will still lose the election.
In Koglai, in Waiye LLG, there was group discussion on 21 June about how the community would vote, living as they do along an access route controlled by a volatile politician renowned in the last ten years for his use of firearms. I was told they decided to give him some of their votes, but most for their own candidate. They all voted, but did not double vote, for the Regional seat. They then found they had 451 Regional ballot papers left over and discussed how they would allocate them. They knew that individual boxes are tallied in public in Simbu, a necessary step towards the transparency that can reduce suspicions of fraud by polling officials. This group then allocated blocks of 100 votes each to the sitting member, to a local candidate who controlled their access road, to one candidate seen as likely to succeed, and to another who was their brother-in-law, and the rest to a few other individuals who had held pig feasts and would want some obvious return. This was subtle Simbu diplomacy, a form of insurance against recriminations by losing candidates (or even the winner).
Simbu informants express concern at the many firearms in the rural areas. On polling day in the Singga Valley from Pari through Kurumugl and Koglai area, Waiye Team 81 at Kuman saw the following semi-automatic military weapons: M-16, SLR, AF-15 as well as an old .22 rifle. Personally I have not seen guns in rural areas this year, but from late June I heard shots on a nightly basis in Kundiawa town
Unfortunately, there were many gunshots in their Singga River valley that night, and gun fighting broke out on the next day, allegedly started by a prominent candidate from Kurumugl village. Then, I was told, the next day his people were attacked from both west and east, from Koglai by Siambugla and from the east by part of their own Bomaikane clan within Kamanegu tribe, who had their own candidate. Many houses were burnt, including several high covenant houses . Village sources say three men were killed, one from each party to the fight, although the official figure is two deaths. The defeated group took refuge at Mogoma over the range above Kundiawa. On 27 June there was a confrontation between two Bomaikane-owned motor vehicles at the road junction in front of the Kundiawa police station. Shots were fired which in turn led to a shooting spree by some police, and to the incarceration of both Bomaikane candidates, who were subsequently released on K500 bail each.
There were other incidents in the period after polling, although to date not the major battles which Simbu people expect after the count. At Gumine district station the supporters of two candidates exchanged gunshots. Sadly, a schoolgirl was killed in the cross-fire. The sitting MP was arrested for carrying a shotgun in a public place, then also bailed for K500 (Wantok, 27 June 2002, p2). In Dingga, a candidate led a machete-wielding charge on a group of terrified men, women and children who were walking on a steep old road away from the bad situation further up the mountain at Emai village.. A woman candidate bravely stood her ground, and her opponent then apologised. This story was quickly exaggerated into reports that two women had been pack-raped. To increase police powers of search and arrest, a Tribal Fighting Zone was declared over the entire province on 27 June - rather than the town, which is what several candidates had requested of the Provincial Police Commander the previous day. This was a repetition of the events in the 1992 Chimbu poll (Standish 1996).
In Kundiawa itself, pending the count, the main action remained at the Police Station. As noted above, candidates started coming in after the vote saying their ballot papers had been hijacked on the road and were seeking replacements, or that there were not enough papers. By Monday evening 24 June, at the request of some AROs and candidates, thousands of spare ballot papers were burnt in a huge bonfire in the police yard, a process which continued over the next few days. This took some heat off officials, while giving some warmth to returning polling teams. Candidates claimed there had been 40-5, 000 additional papers in the province. Several expressed to me a concern that some police would insert additional boxes of marked papers into the count, as some alleged had been done in the 1999 by-election for the Regional seat, thus determining the winner. As the delays grew, so did popular suspicion of the entire electoral process and the hunt for scapegoats.
Some final impressions
Candidates, citizens, police and officials alike say this was the 'worst organised election ever', and 'the worst in history'. The roll is useless, they say, and many are suggesting other ways of organising voting. They want effective voter registration, and asked how it is done in Australia and America. Tatooed numbers and even implanted microchips were dismissed as unsuitable possibilities. As mentioned above, people are considering identity cards, or a franchise based on tax payments: no representation without taxation.
In Chimbu, the women's activists and candidates are most articulate, saying this election was a farce, that this is not democracy, there is no freedom of choice in this ballot, this was 'ganpoin vot' ('voting at gunpoint'). They say young male thugs have enforced the vote, bought off by rich male candidates, and that there was no choice, especially for women voters. Women and other candidates and leaders are discussing grounds for planned appeals to the results of the election, which they say has been irrevocably corrupted by bribery, cheating and intimidation.
Some thoughtful observers say there has been free and fair voting in past elections, but this year the Highlands does not have normal democracy. They say that what we are seeing is the move towards a 'Melanesian democracy ' or 'democracy Chimbu-style', following the collective consensus of the haus lain. Perhaps that is all we can hope for, they say, somewhat wistfully.
The enforcement of group consensus - or indeed its absence - shown here matches the pattern which Gelu describes as authoritarian political behaviour, with Papua New Guinea (or at least the Highlands) becoming a 'non-liberal democracy'. In essence, he argues that this style of rule-breaking, rough house politics more closely reflects Melanesian political culture than the introduced concepts of liberal democracy (Gelu 2000).
From the Prime Minister down, there have been calls for an investigation into the running of Election 2002. Sir Anthony Siaguru, the political commentator, described the national election as a 'debacle', and said that the country needs to ask 'whether the electoral process has been so seriously compromised because of a combination of disenfranchisement and manipulation that democracy in Papua New Guinea has been put at risk' (Post-Courier, 28 June 2002).
Back in Chimbu, counting will not start until all ballot papers are returned, once funds have arrived from Port Moresby to pay polling teams and the police their field allowances . The LLG counts will be conducted first. Temporary shelters are being erected at the Kundiawa football ground, Dickson's Field, and for security reasons the counting will only be conducted in daylight hours. In previous years counting (day and night) has taken 10 days, and is a very tense period with high levels of frustrations as most candidates - the losers - face humiliation.
Widespread fighting is expected after the count, as tensions mount and recriminations commence. In the NAURU area, one candidate reportedly slaughtered four pigs, and a tribeswoman on 26 June stated that he had wiped his firearms with the pig fat, a traditional preparation for war. (This was an area of intra-group fighting after 1997 election.)
The situation was very tense when I left Kundiawa on 27 June, with the town very crowded, and pressure building up as candidates sought to keep their supporters together for the delayed count. There was some relief for policemen and townspeople during the Australian State of Origin football match on PNG TV on the night of 26 June, but it was not clear if the gunshots heard that night were celebrating football goals on TV or part of the local game.
At each national election this province's administrative capacity for elections seems to decline, while the levels of fear and tension rise. While decrying corruption and violence, Simbu people take their politics seriously, and most abandon all pretence of good governance in seeking to promote the renown of their clan, right or wrong. There is little room for national interest in such circumstances, or any common good apart from the collectivity of the clan - a point reiterated by experienced Simbu observers. Even though people know they are breaking the electoral rules, the Simbu group ethic becomes the overwhelming motive.
The state, the object of these separate power struggles, is politically weak. State personnel are sucked into local politics and the state is utterly ineffectual in controlling the mobilised citizens of different clans and tribes. The abuse of the electoral system is made possible by a combination of weak provincial administrative capacities, bad policing, poor planning and budgetting, inadequate funding, and a repetition and worsening of problems which have been seen in the national electoral system over several previous elections. These are all compounded by rugged terrain and degraded physical infrastructure and - on this occasion - a week of appalling weather.
Diamond, L.J., Linz, J.J. and Lipset, S.M., 1988 'Preface' in Deomcracy in Developing Countries: Volume 2, Africa, Boulder, Co., Lynne Reinner
Gelu, Alphonse R.M. 2000. 'The emergence of a non-liberal democratic political culture in Papua New Guinea', Point No 24, 'Politics in Papua New Guinea: Continuities, Changes and Challenges', ed. by Michael A. Rynkiewich and Roland Seib: 87-119.
Papua New Guinea Electoral Commission 1997, Report to the Sixth Parliament on the 1997 National Election by the Electoral Commissioner Reuben T. Kiaulo, July.
The National. (Port Moresby)
The Papua New Guinea Post-Courier. (Port Moresby)
Standish, Bill 1996. 'Elections in Chimbu: Towards Gunpoint Democracy?' in Yaw Saffu ed., The 1992 Elections in Papua New Guinea: Change and Continuity in Electoral Politics, Canberra, Australian National University, Political and Social Change Monograph: pp 277-322.)
Refs: Wantok, Port Moresby
I thank the polling officials in Chimbu, and my Simbu informants, for their full co-operation with this study. I also thank the Melanesian Institute in Goroka for providing facilities to prepare this interim report, and its Director,Fr Nick de Groot, for his comments on the draft.
Bill Standish is a Visiting Fellow at the Faculty of Arts, ANU, and is a participant in the ANU-University of Papua New Guinea 2002 election study. Partial funding for field trips in April and June 2002 came from the Australian National University's State, Society and Governance in Melanesia Project, which is assisted by AusAID, as part of the SSGM project on policy-making and administration in Papua New Guinea.
© Copyright 2001-2006 Peter Salmon and other web site content contributors. All rights, whatever they may be, reserved.