 Let me start at the end of the process I want to talk about by saying what Election Day was like. There were really big logistical problems through the election. So South Africa is the richest country in Africa. It had an established public administration that had been around for decades. What went wrong? And that's what I want to talk about today. South Africa had a very long history of racial discrimination. South Africa became an international pariah because of the way its system worked. They withdrew from the Commonwealth in 1960. They were subject to international sanctions from the UN. They were subject to sporting boycotts. They came to be very isolated. They found themselves fighting border wars in the states that were surrounding them on the border between Namibia and Angola and when Mozambique was decolonised. There was conflict in the region. There were Cuban forces supporting the Angolans in their war with the South Africans. It was a very unstable and in the long-term untenable situation. But out of that stress, eventually the point was reached when the South Africans realised that something had to change, that the system could no longer be sustained in the way in which it had been put into place. And there were a couple of key elements to that. One was when Namibia, which had been similarly ruled as part of South Africa up until 1989, had a successful transition to majority rule and independence in 1989 and 1990, encouraged by the UN. And that set an example to the South Africans that change was actually possible if it had been properly managed. And so in 1989, the very hard-line former president of South Africa, P.W. Borter, retired and was replaced by F.W. DeClerc, who was a much more enlightened character in some ways. And that set the path for a political transition to majority rule in South Africa. Nelson Mandela, who'd been in prison since 1964, was released from prison. The apartheid system was formally dismantled, though the way in which the country was operating was still very discriminatory in many ways. And election came to be on the agenda as the critical moment when that transition would take place. So 1993 and 1994 were the very big years of the history of change in South Africa. And it was decided around about August 1993 that an election would be held in around about April 1994, which was not a very long lead time, only about eight months. Now, even though that date had basically been set, there was no electoral commission that had been created to actually run the election. And that was also subject to negotiation. And eventually the commission was legislated for only in October 1993. So a couple of months has already been lost. And the chair and members of the commission weren't appointed until December 1993. So you're now getting down to four months before the election, rather than eight months. And they really, because they're rather good at observing public holidays in South Africa, they really didn't get down to work until January 1994. So you're now looking at a time period between then and the election, which is about the same as the popular consultation here in Timor-Leste. But there were some very big differences. And the biggest one of the lot was the increase in the size of the electorate. Because everyone suddenly was going to be able to vote. The number of people who voted in 1994 went up from the previous election when it had only been 2.6 million. It went up to 19.5 million. So that was a seven and a half fold increase in the number of people who were voting. Now, no other country that I can think of has ever faced something like that. In the old days when women didn't get the vote, if you enfranchised everybody, that doubled the size of the electorate. But a 7.5 fold increase was really unprecedented. And because of the way the apartheid system worked, most of the new voters were living in areas on the edges of the cities. And they were very poor areas with inadequate infrastructure. And they were areas where electoral administrators had not been before. They had never had to provide polling facilities in these places. So there was this vast sort of area of the country that was really unknown territory to the people who were going to get involved in running the election. And that then brings me forward to talk about... I want to talk about all of the things that went wrong and all of the challenges that they faced. But I wanted first to set the scene by saying this is really not a criticism of the people in the Independent Electoral Commission in South Africa because they were handed what many of us would find an almost impossible task. And they did their very best in the circumstances, but it was still difficult and it's important to learn lessons from those difficulties. First, the political situation was inherently unstable. There were people in the white community who did not accept the notion of a transition to majority rule and they were violent. There were also what they called homelands, which had been set up nominally as independent entities within the country which had their own quasi-governments. And they were people who really were not accustomed to the idea of change either because they had set themselves up quite nicely. So they were not cooperating with the process very well either. And that was notably the case in some big areas of territory in the country that created problems. But one of the big problems also was that the Independent Electoral Commission didn't really appreciate the magnitude of its task. At one level everyone said, oh, it's a big job. But in some of the circles you were talking about, particularly people from the Home Affairs Ministry who had previously run elections, they seemed to have a sense that it was like a previous election, but just a bit bigger. And they didn't really get the heads around the notion that this 7.5-fold increase in the electorate created a whole new world. Also, Home Affairs had been the department that administered most of the most hated administrative elements of the apartheid system. So they had a really serious problem with political credibility. Going into that situation with a faith in systems that you'd used before and a belief that they would work again was not really justified. And a lot of the problems actually flowed from that. There was also a problem and a challenge associated with the fact that a new commission had been created. Most of the people who were in that commission were highly respectable, incredible people from a political point of view, but they hadn't run an election before. And there were 15 commissioners, some of whom were international ones who arrived later. When you are appointed as a commission and you don't have any staff, really, your natural reaction is to think you've got to do everything yourself. And so the commissioners started to work as if the election was actually being run by the 15 of them. And they gave much less attention in the short run anyway to the fact that at election time they would need hundreds of thousands of staff out in the polling stations and the field staff to manage that and that sort of thing. So there was a very considerable pattern of not thinking at the beginning of delegation, which decisions had to come to the commission, which decisions could be made by administrators in Johannesburg, which decisions should really be left to the provinces and the field people without bothering the top. So you got into a habit where all the decisions were going right to the plenary meetings. I recall actually just looking at Pakada nodding that something similar happened in Indonesia in 1998-99 with the big kapa'u there. It's a pattern that one sometimes sees in these things. And one of the, this whole issue of decentralisation or creating a field structure ultimately was dealt with far too late. There was a sense that it was important to have people who were politically credible at all levels of the organisation, but to actually do checks on people's backgrounds to make sure that they don't have a baggage that will damage the credibility of the election takes time. And so what happened was that things slowed down so much that at one point panic set in and then they just had to appoint people anyway without doing all the checks that they had in mind just to get the operation going. But that whole business of going from a centralised operation to a decentralised operation created a big problem. And they started too late, partly again because of this lack of appreciation of the need for the field staff. One of the more chilling experiences I've had in my life was to go to Port Elizabeth, which is the fifth biggest city in South Africa, three weeks before election day. And it was good Friday in Easter. And we went into the office of the provincial electoral officer. And it was a room like this, but with no furniture and just a telephone on the floor. And no one was working because it was a public holiday. And by that point the situation was objectively desperate. They were in real trouble, but they still didn't have this provincial office up and running three weeks before the election. That was really scary. And I went back to Jo Berg and reported to my boss in the UN that there's going to be big trouble at election time because it was all happening just too late. They also had a problem with identifying voting stations because they needed so many more than they had ever had before. There were some lists that were produced by the Department of Home Affairs of possible sites, but then the police went out and checked them for security reasons and they found that some of them didn't exist, and that didn't work. We as UN observers were asked not to talk to people whose buildings had been chosen, even though they were on a public list, because the IEC hadn't told them yet that their places were going to be polling stations. And this problem never went away. And so someone tried another emergency response which to say, let's get a geographic information system and see if we can use that to identify buildings. That might work now when GIS systems are pretty good and you've got Google Earth and Google Maps, but this was 25 years ago when you could hardly get a laptop with any power to it. And somehow in using the system they managed to get latitudes and longitudes mixed up, so we got a map and it had polling stations in Botswana and it had other polling stations in the Indian Ocean between South Africa and Madagascar. So that was pretty disconcerting too. There was a real difficulty with that. And the thing I remember perhaps most of all is an area called Trans Sky in the south of South Africa, quite a big, populous area. And on one day, which was a week before the election, the number of polling stations they were saying would be in Trans Sky went from 1600 down to 600 up to 1200 and up to 1600 again. That was on one day. We got four separate messages about how many polling stations there would be in there. This is more than the total number of polling stations in the whole of Timor-Leste, incidentally. And it reached the point where we said, let's not listen to this anymore because with the best will in the world, if they get 300 set up there, they'll be lucky. This debate about how many stations you are going to have there just became theoretical. The other problem that arose with the IEC in the beginning have a real project plan at all. They had a million obviously urgent tasks to do and so they just threw themselves at the task that came to the top of the pile, as it were. And that again meant that there really weren't work plans for the administration of the election until very late in the day. And that became a problem. Where they tried to develop computer systems that caused more trouble than they were worth. You can't develop a big computer system in three months. You just can't do it. You don't leave yourself any time for testing it. You wind up with all sorts of vulnerabilities. It's a very challenging thing to do. And that's how they wound up with a system that was able to be hacked. And very often, because of the urgency, they went straight to the computerised systems without actually building them on top of a manual process that had been devised before. So when the computers went wrong, there wasn't a backup. They also had big problems with internal communications within the organisation. And that's one of the first things that always goes wrong when you're in a crisis. Internal communication gets lost because everyone's so busy and they don't have time to tell other people what they're doing and how things are going wrong. And it got to the point where the IEC started issuing what they call technical updates to staff. And they issued 19 of them. My notes tell me between the 13th and the 25th of April, 1994, which was the first day of the election. Now, the problem with that is that when you're in that sort of situation, people don't have time to read the new technical advice. So there was an assumption being made in Joburg that once you issue a direction, it will instantly be followed everywhere. And the world doesn't work that way. Most of these directions, decisions, any decision that gets made in the last two weeks in that sort of situation, never gets implemented. It just gets ignored. The staff in the field, they're up to their eyeballs and problems too, and they don't have time to read the latest stuff from headquarters. You get the classic dynamic of the field not liking what the headquarters is doing and vice versa. Everyone's got their own problems and they're not interested in other people's problems. And in fact, they extended the polling for a day because of the logistical problems they encountered. And that was announced on television. It was simply a press conference by the chair of the Electoral Commission announcing we're going to have another day of polling. With hope, I think, that everyone would be watching television and would see that and get the message and come back to their polling stations the next day and be on duty again. The other problems they had were to do with political party participation. There was a party called the Encarta Freedom Party that was based in Natal, around Durban, led by Chief Mangasutra Buttelesi, who was a Zulu leader, and they had boycotted the election and this was looking to be a major political problem because there were significant political elements down in that part of the country. And a week before the election, they were persuaded to come on board. The problem with that was the ballot papers had already been printed. They had to come up with a solution to that and I'd have to say this was something they did well. They actually devised a sticker that could put this party on the bottom of the ballot paper and they printed 20 million of them somehow and they issued them to the polling stations and told the officials to put the sticker on the bottom of every one of these ballot papers before you gave it to the voter. So that was actually a desperate move, but it was one that worked. But there was a lot of panic at the last minute with distribution of materials and basically secure management of them was lost and they were actually dropping reprinted ballot papers from helicopters out in the east round without knowing who was catching them down on the ground below. The number of ballot papers that were reprinted during the election was anywhere from 2 million to 9 million depending on which story you want to believe from the South African Press Association and of course the usual sort of chains of custody of people signing for ballot papers and managing their numbers with exact precision just fell by the wayside. They did all the counting for Johannesburg, Forenican and Pretoria at one place. There were about 6 or 7 million ballot papers coming into the National Recreation Centre down in Soweto just outside Johannesburg. And unfortunately nobody had planned for the receipt of the materials from the polling stations and they got to a state where they had 170 utility trucks backed up the Soweto Highway all full of ballot boxes and someone panicked and just said bring them in and drop them and that's what happened. People came in and they dropped the ballot boxes. The trouble was they got separated from the forms from the polling stations that said what was in the ballot boxes and how many votes should be in there. So it took them at least a week to sort out that mess and at the end of it they wound up with a pile of ballot boxes probably about a third of the length of this room and about up to my eyeballs that could not be tracked back to any polling station. They didn't know where they had come from. They couldn't figure out what their antecedents were at all. They could have come from anywhere. And at that point they decided to cut corners and not try to do a ballot paper reconciliation. In fact Judge Creedler from the IEC went on television and made the famous comment national reconciliation is more important than ballot reconciliation which you can agree with without thinking that ballot reconciliation is unimportant. So that sent a terrible signal as to how things were being managed and really set the scene for the fact that they had to negotiate agreement on the part of all the political parties to accept the result. And it had been made worse by the fact that the IEC had made very optimistic statements about how things would be. They did not manage expectations downward if anything they raised expectations. So they were talking about being able to finish the counting by the day after the election which was manifestly impossible. They had big advertisements in the newspaper saying we're ready, come and vote, let's do this. And it turned out they weren't ready. And that did not help if they had actually anticipated some of the problems that came up during polling and sensitised people to the risk that this would happen it would have been better for their own reputation. Having said that there were a couple of strong areas within the organisation. One was the so-called monitoring director which was a rare innovation because so many of the staff had worked on elections had this rotten reputation of having been part of the apartheid management system. They actually set up in effect the biggest internal audit unit that any election commission has ever seen in the world as big as the rest of the organisation combined which was monitoring the way they were doing their own work. And that monitoring directorate actually did a very good job and wound up picking up the slack from the administration side of the organisation to try to get things working. And that was the strength. The other great strength was actually the polling officials themselves who did a brilliant job in very difficult circumstances because they really wanted to deliver the franchise to the people who'd been denied it so very long. So they got through and they were able to negotiate an agreement whereby everyone effectively saw power transferred but it was not the election that made that a success. It was the strength of the political agreement between all of the parties that had built up over about five years of negotiations to the point where everyone knew there was going to be a change of government. Everyone knew Mandela was going to become the president and the trick was to try to get this to be done as smoothly as possible. But if something like that had happened in the Indonesian election in 1999 or here in Timor-Leste in the same year it would have been a total catastrophe. So there were a lot of lessons to be learnt from the South African experience and one of which is you never have enough time. Time is always the scarcest commodity and if you have a short time you need to get your planning done very early so you know what you're doing and you need to get as much done as early as possible because if you throw away your slack time at the beginning you'll have no capacity to respond to the unexpected when it hits you later in the process. So that's just what I wanted to tell you of my South Africa story.