This question was posted on behalf of a freelance journalist in Fiji

Due to a lack of expertise locally in Fiji, I am writing to ACE in the hope of expert opinion and guidance to discussions/academic papers on the following issues of concern:

1. Fiji is soon to embark on a voter registration process in preparation for the regime-announced election, scheduled for Sept, 2014. Shouldn't a full census and boundary delimitation be undertaken first, given the major changes with regards to population movements both inside and outside the country?

2. In-post conflict situations and with the nation under military rule, ought there not be some independence of election management bodies and those entrusted with even registration, in order not to have the integrity of the process questioned at a later date?

3. What lessons can be learned from the case of Indonesia, where elections have been managed both during and after military rule?

Summary of responses
It is noted that although a full census is required for boundary delimitation, neither is a prerequisite for voter registration, which can in theory be carried out beforehand (regulations permitting).  A robust and detailed voter registration system would, however, be necessary in order to detect and eliminate duplicate registration and voting. Another reply points out that although boundary delimitation should ideally be carried out before voter registration, it would be possible to adjust the register to altered boundaries afterwards.

On a more general note, conflict can be the catalyst for a review of electoral districts and population distribution, and Liberia and Sierra Leone are given as two such examples. The difficulty of organizing a general census in a post-conflict context should not, however, be underestimated.

The majority of replies advocate the independence of the electoral management body (EMB), but as one expert points out, “integrity is a factor of more importance than the independence.” Another stresses the need for competence and efficiency as well as independence. EMB independence is often aided by it having fair and equal representation of the parties/candidates.

One expert stresses that the real problem in Fiji is political; that although EMB independence is important, this is not the issue at hand and that the most recent census is sufficiently accurate for elections in 2014.

Lessons from Indonesia could include its use of technology in voter registration and mapping of polling stations. 

Examples of related ACE Articles and Resources
Encyclopaedia:
• Use of population data in the delimitation process 
• Boundary delimitation

• Independent model EMBs
• Voter registration: Periodic list

Regions and Countries:
• Fiji

Names of contributors
1. Kisimba Albert
2. Charles Benjamin Nsimbi-Kabugujjo
3. Carl Dundas
4. Julian Nundy
5. Vasil Vashchanka
6. Ian Browne
7. Andrew Ladley

1.       Il difficile dans un pays post conflit d’organiser un recensement général de la population, pour la simple raison que les données seront bien sur biaisées par le fait que certains de habitants se trouvent être soient des déplacés internes soient encore des réfugier dans les pays voisins et ce mouvement de la population ne donne toujours pas foi à une base de donnée fiable de la population. Seule l’identification des électeurs est à encourager dans un premier temps.

2.       L’indépendance des OGE est un sujet délicat dans votre contexte comme  dans bien des pays post conflits. Les membres viennent en grande partie des personnes qui ont une certaine influence sur le morale de ceux qui dirigent ; la conscience et l’intégrité des OGE est à souhaiter et à rechercher dans ce cas.

3.       Si nous recherchons  la démocratie des les yeux des dirigeants (militaire ou civile) nous ne la trouveront pas, il faut plutôt sensibiliser la population et ce à long terme et régulièrement sur les mécanisme et les outils dont dispose l’Etat pour les rendre « libre » et citoyens.

1.  A full census is required for boundary delimitation.  however both are not a prerequisite for voter registration.  Indeed voter registration may be the basis of boundary delimitation as the numbers of registered voters, the geographical and other physical features and accessibility will be the basis for determining polling stations.  Census and boundary delimitation are necessary to determine representation. Depending on the law governing registration (basis of registration i.e address), you can successfully register voters prior to a census and/or boundary delimitation.  You must however have a good system (biometric) to detect and eliminate duplicate registration.

2.  Ideally, all EMBs should be independent of the central authority.  However, integrity is a factor of more than the independence.  It is also a perception issue.  There are a number of EMB that are partially or entirely part of the government.  It is important in that case to have the processes conducted transparently.  The voters' register being the central (most important) component of an election is often suspected of being manipulated.  It is possible to ensure (to a large degree) that it is not manipulated by for example mobilizing the population to actively participate in both the enrollment and the subsequent exhibition.  It the register has images, then manipulation can only result into multiple registrations that can be detected during the exhibition.  issuing registered voters with cards can also improve the register's performance by ensuring that on polling day no body is turned away purportedly because they are missing on the register.

Charles

Normally population census is held every ten years pursuant to the constitution in place or by practice. Electoral districts or constituencies are reviewed in accordance with the constitution and the electoral law which generally require a lapse of 8 or 10 years since the last delimitation, or after a population census takes place. Population displacement due to conflict situations might cause both the population distribution and the electoral districts to be reviewed, for example, see the village mapping exercise by the UN in Liberia in 2004-5 to facilitate both location of population shifts and review of constituencies after the conflict; also see the Sierra Leone experience with population shifts during the conflict in the 1980s -1990s which triggered the abandonment of the FPTP electoral system for PR system due to dispersal of the population and the inability to define meaningful constituencies.

It is important to have not only an independent and impartial electoral management body (EMB), but an efficient and competent one when elections are being held under a military regime and in a conflict or post-conflict environment.

Perhaps, Indonesia is not an ideal comparison with the Fiji situation, but there are broad similarities. It depends on the political will of the military players, if they are committed to credible democratic elections, it will work out for the people of Fiji. I spent several weeks in 2006 advising the electoral authorities in post conflict Aceh, Indonesia, and it was pretty tough going electorally, but they did pull it off eventually.   

Getting full and accurate statistics at short notice is a difficult task.

It is easier, however, to establish methods that prevent people voting more than once and other forms of fraud. What sort of election observation is planned? Commonwealth, European Union, UN? Any combination of these organisations will at least ensure that shortcomings are noted and, hopefully, dealt with beforehand. The Australians (who would presumably make up a large part of any Commonwealth team) have huge experience in the conduct of elections in the region (East Timor, for example) and they have a lot to offer both in terms of electoral assistance and observation.

If no observers have been invited, this is a bad sign.

As for the independence of the EMBs, this is often obtained by ensuring fair and equal participation on them of the different parties or candidates' representatives.

Dear Madam or Sir,

A number of issues involved in part 2 of your question are addressed in the forthcoming International IDEA paper "Electing change:  Electoral management in transition". This paper includes Indonesia as a case study. I will share it with you as soon as it is approved for publication - which I believe will happen in the next two weeks.

What do you mean by a full census? Is this to include both persons eligible to vote as well as those who are not? If the census is only for persons eligible to vote and it is done in a comprehensive manner, then the work in a voter registration process can be considerably reduced or even eliminated. In addition the results of the census should also give enough information to suggest the extent to which any boundary realignment is required. 

The EMB in any electoral process should remain independent

Perhaps an introductory point, which I assume the person who put the question knows very well but I understand the constraints, of course. I have found plenty of expertise in Fiji to answer all these questions - since independence in 1970 there have been many elections, under different systems, all conducted by Fijians. The factors behind Fiji's current constitutional problems under the military-controlled government, have not been due to lack of administrative and technical electoral capacity. They have been, in a word, political. And under current conditions, not all the people who have previously conducted elections are likely to offer their services until they have more confidence in the integrity and direction of progress regarding the restoration of elections. The last 'Election Supervisor' (the local term for Chief Electoral Officer) was appointed ostensibly to prepare for elections to return to civilian rule in 2009. But the abrogation of the constitution (and hence of all constitutional offices, including elections) shortly after that appointment stopped all official electoral preparations. 

As regards the details of the question, it is hard to prepare for an election currently scheduled for sometime in 2014, with absolutely no indication of any of the parameters: i.e. no electoral law, constitution, electoral commission, boundaries for constituencies, age of voters, forms of representation, etc. That said, at least some of the basics of the process are already in place, but they will need updating.

Ideally, defining the constituency boundaries precedes a fresh, or revived, voter registration. But it is possible with good records (including voters' addresses) to adjust voter registration to altered boundaries. To get approximately equal voters in each constituency, a recent census is generally used. This alters boundaries, adding 'clumps' of voters from over full areas (normally urban) into less population-dense areas, to get balances. The results can divide streets down the middle etc, and census figures thus tend to be quite finely mapped in 'blocks' that allow this to take place.

In Fiji, the last census was in 2007, I think. Even though this is a few years ago, it will still be reasonably accurate - and is the best available data. These details will be available to be deployed as soon as there is any indication of basic electoral details and as soon as the necessary institutions are established. For example, a Boundaries Commission will need to know how many constituencies there will be in the next parliament, and whether registration will be by race - as in previous Fijian constitutions - or not.

As soon as the boundaries are established, a new Electoral Commission would register voters in the constituencies. 

If voter registration commenced before the boundaries were settled, it would in theory still be possible to allocate them to constituencies afterwards, by going through all the physical addresses. But this depends on the adequacy of the geographical information captured in the registration, and the computer systems that are used. As ever, bad information capture/systems will mean that the allocation into constituencies is very messy. Ideally, voter registration could be done using the same 'geographic blocks' that were used in the census, in which case the reallocation would be quite easy. 

Short answers to your questions:

1. The real problem in Fiji is the lack of any law or election details. A recent census is available. It is possible to reallocate registered voters, but the ease of doing so depends on the accuracy of the capture of details on registration.

2. Yes, it is always better to have functionally independent bodies to conduct electoral activities. But again, the real issues are not about election bodies in Fiji.

3. There are of course often lessons to be learned from other experiences. Indonesia's experience and technology in registering >120million voters down to very accurate geographical areas, and having polling stations such that on average only 300 people vote at every polling station, is interesting. The votes were counted at polling stations upon closing (quickly!), and the tallies then gradually passed up the line to be accumulated (this took much longer!). The military authorities in Fiji are, I gather, interested to learn if Fiji could use any of the technology for registration, particularly. The answer, alas, is not simple. As one of the ACE practitioners previously remarked, technology tends to make already-efficient systems work better, and inefficient systems even worse. The specifics need to be evaluated and tested, with no assumption that what works in one place will automatically work in another. In Fiji, again, the real election problems have been political, not technical. That said, even getting ballot boxes made out of plastic, not wood, could make Fijian elections, transportation, storeage etc, much simpler.

This question is posted on behalf of a member of an international organisation working in Fiji.

Based on news reports provided in Fiji, the UN is currently assessing the overall support it will give to the upcoming process leading to the elections scheduled for 2014 to the Government of Fiji, in line with a request from the Government. Those discussions are ongoing and hence, the scope of support is not yet known. Besides this, the international community, the Pacific Island Countries's Contact Group (including the major regional countries Australia and New Zealand) have undertaken a recent assessment mission to talk to stakeholders and assess the progress made to date and decide on support options. Fiji is currently suspended from the Pacific Island Forum and hence the contact group has made an official visit to talk to various stakeholders. Based on those two processes, the international community is planning its support. The EU and USA have also publicly expressed support to the democratic process.  

At present there are preparations for the electronic voter registration ongoing, with help of two international advisers: one from New Zealand and one from Australia. The voter registration equipment was bought through donor support (I think from China but not sure, Australia and New Zealand also offered). The planned voter registration will be for six weeks and will lead to all registered voters receiving voter cards - which was not the case before and should minimize fraud, as stressed by the Attorney General / Minister for Elections.
 
The Constitutional commission members have been appointed and discussions are to start soon. The government plans civic education and stresses everybody can make any submissions. 
 
Discussions on election systems have not really taken place in public but the Atorney General has stated on the news that a plurality system will be used and the previous second ethnic vote will not be used.