Original question:

This question was posed by an ACE user through the "Ask a question" function in Electoral Advice. He/she is a member of an international organisation.

I am hoping you will share your expertise on the issue of absentee voting facilities allowing voters to cast their ballot away from the area they are registered to vote.

Introduction

Bhutan will hold its first National Assembly elections in 2008 for a maximum of 75 seats. The elections will be conducted in two stages: first in what is called a Primary Round in which all registered political parties are eligible to participate, followed by what is called a Second Round (or General Election) where only the two parties with the largest number of votes from the Primary Round will be able to compete. The General Election will take place in single member districts – in which the candidate with the highest number of votes in each district is declared elected (First Past the Post). The party whose candidates win the most of the seats will form the government, while the other party will sit in opposition.

A mock election was held in March to May 2007 in which voters were asked to choose between the Yellow, the Red, the Green and the Blue Druk (Dragon) Party. The Yellow Dragon Party won all seats except one in this election.

Summary of responses:

Posted on 13 September, 2007

Absentee voting allows citizens to vote at a location other than where they would normally be registered for voting. A general overview of absentee voting can be found in Absentee Voting  in the ACE Encyclopaedia. Absentee voting facilities (often in combination with facilities for early voting) are valuable tools in increasing the accessibility of the voting process, by inviting citizens who would otherwise be unable to cast their ballots (for example, due to working situation, security reasons, economic constraints, illness or disability) to participate in the election. On the other hand, the provision of extensive absentee voting facilities is likely to increase the cost and complexity of the electoral process; it is likely to delay the announcement of the results and might even raise the risk of real and alleged electoral fraud. Some basic issues to consider before introducing absentee voting procedures are brought forward by Antonio Spinelli and David Henry below. These include:

Is the necessary legislation available to allow for absentee voting? Does the legislation need revision before the introduction of absentee voting provisions, and in that case, is there time for such a revision? In addition to the issue of legislation relating to the absentee voting provisions Roger Hällhag also says that a purely technical discussion of the upcoming elections might risk bypassing fundamental problems in the legal and political framework in Bhutan, and argues that there is need for a wider debate, addressing issues such as political rights and freedoms.

Can the details of the absentee voting provisions be worked out well in advance of election day? How does the electoral system affect the absentee voting provisions? Is the problem primarily affecting a few specific constituencies/areas or the whole electorate? What is the size of the electorate?

Will there be time and resources available for the massive voter education campaign needed to inform the voters of their rights to cast absentee votes and how to do it?

What is the capacity of the national, regional and local electoral administration? Will they have enough resources to deal with the implementation of the process? Is it worth the extra cost?

When these issues have been considered, some basic design issues need to be addressed, such as:

• What conditions, if any, must a voter satisfy to be eligible for an absentee vote (see Qualifications and Eligibility Procedures)?

• Where may a voter have an absentee vote - for example, only within the electoral district within which the voter is registered, or in some wider area (see Locations for absentee voting)? Should all ballots be available everywhere, or sent to specific stations only? What are the administrative, cost and logistical implications of this choice, and how is the choice affected by the electoral system? (Can one ballot be used everywhere or do specific ballots need to be sent to specific stations?)

• How is it determined at a voting station if a voter is eligible for an absentee vote (see Qualifications and Eligibility procedures)?

• Should the ballot format for absentee votes be the same as or different from that used for normal votes (see Absentee ballot forms)?

• How should the absentee ballots be singled out in order to be counted in the right district? (see Issue and Casting of Ballots) Sean Dunne, David Henry and Richard John all mention below the possibility to use a “double-envelope ballot” where the absentee ballot is put in an envelope marked with a code that identifies the consistency of the voter.

• What arrangements are appropriate for transporting absentee votes to and from the voting stations and what arrangements should be in place for counting the absentee votes? Dr Noor Mohammad points out that the transporting of ballots could prove to be costly not only in terms of logistics but also since it might challenge the legitimacy of the elections, raising opportunities and suspicious of fraud. This is especially true in small constituencies, where a large number of postal votes could actually determine the outcome of the election. Antonio Spinelli adds that transporting ballots might raise security and transparency concerns and are likely to delay the tabulation of results.

Several specific solutions which could allow voters absent from their voting stations to vote in Bhutan are brought forward in the replies below and in the ACE Encyclopaedia. They can be categorized into three groups:

Replies were received, with thanks, from:

• Ray Kennedy

• Sean Dunne
• Dr. Noor Mohammad
• Richard John
• Debashis Sen
• Antonio Spinelli
• David Henry
• Maria Gratschew
• Roger Hällhag
  

Individual answers

The Australian system of placing a quantity of ballots for each parliamentary constituency at one designated polling center in every other parliamentary constituency might be workable, particularly given the small size of Bhutan. After the election in Australia, the ballots are moved quickly from where they were cast to where they need to be included in the count.
Costa Rica used to subsidize the political parties in transporting people back to their constituency of origin but discontinued the system in the 1990s due to the high cost.
Mexico's system of placing a limited number of national ballots at a limited number of special polling stations in each state enables some people to vote for national (but not their own local) races, but generates a lot of controversy when more voters turn out than there are ballots.

For operational solutions they might consider :

• a special registration for out-of-constituency voters, to enable them to identify and pre-register for their preferred polling location.  This would permit a known number of ballots to be prepositioned for these voters, and the polling staff to have a roll of eligible voters for this purpose.  This would at least minimize some of the risks of double-voting that might otherwise arise (if they're not using indelible ink), and would provide logistical information on whether a special polling station and staff would be required to handle the number of voters anticipated.
• A second approach (which might be used with the approach above) would be the use of double-enveloped ballots at polling stations for out-of-constituency voters.  A code could be placed on the external envelope indicating the voter's constituency.  Transportation and counting of these ballots would then only need to occur if there were more out-of-constituency envelopes than the difference between votes for the candidates or parties within that constituency.

In a country going for poll for the first time, the complex solutions become difficult operationally. While preparing the civil register, normally the home and the current addresses are taken. If the law provides for voting at the place of residence only, a print module can be prepared for printing the voter lists, which can cull out the names on the basis of the place of residence of the voters.

In case the law does not provide voting compulsorily at the place of residence, options could have been take at the time of registration as to where the person would like to vote and voter lists could have been printed accordingly. It looks none of these applies in the present case.

In the first election, the process should be the simplest - and I believe that a huge number of ballots being transported from one constituency to another after the poll may raise legitimacy issues and allegations of fraud, and should be attempted only with great caution. There may be cases when a candidate wins only on the basis of postal votes or votes transported from another place. In such cases, it might be difficult to deal with complaints of fraud.

There are good precedents from the Bosnia and Herzegovina (BiH) Municipal elections in the 1990s. Many voters in BiH had been displaced by the conflict and were not able to or were too scared to return to their original municipality.  To overcome this problem, voters had to register in their municipality of residence for the specific ballot that they required from their original municipality – or municipality of birth.  The 1991 Census was used as the base data for proof of original municipality.  A special commission was established for those that claimed a different municipality from their present municipality, or that listed in the 1991 census.

Whilst it is recognized that the level of inclusiveness was increased through this policy, the burden placed upon the registration process and the logistical systems of printing and especially packing ballots was increased many fold and should not be ignored. In an attempt to keep the administrative and logistical burden within reasonable parameters, special “Absentee” stations were prepared.  Enough ballots for registered voters had to be made available, with options from 147 different municipalities.  In addition enough spare ballots had to be made available.  It should be noted that when a polling station committee has to check voters against a specific municipality and manage the issue of 147 different ballot types, the voter throughput in a station is dramatically reduced.  This should be allowed for when planning voter numbers per polling facility. Where there were large populations of displaced persons from one specific municipality a special absentee station was set up to manage that specific element of the electorate.

Decision will need to be made with regard to the use of secrecy envelopes (with ballot code) and the location where the ballot is counted.  Due to the special nature of absentee stations, there is a danger that the loss of secrecy of vote can occur.  However, what ever system is used, it must be auditable and transparent.

Without going into the details of the pros and cons, I'd add two ways, other than postal ballots and e-voting, in which the matter of remote voting may be addressed:

(1) Proxy Voting: Intimation must reach the Returning Officer well in advance that voter X would like her friend Y, a resident of her native village, to cast vote on her behalf. Suitable markings are made on the voter list of the village as directed by the Returning Officer. Y casts proxy votes on behalf of X at X's village though X stays elsewhere temporarily (This method is used in India for Army personnel). Proxy voting can be a possibility in Bhutan.

(2)  Duty Certificates: In certain local village elections, officials on election duty are given, upon application, "EDCs" or Election Duty Certificates. Holders of such EDCs can, in course of their duty, enter any polling station and cast their votes by surrendering the EDCs. This method can be suitably expanded in Bhutan if required.

For some time now, from the neighbouring Nepal, I have been following with interest the issues relating to permanent and civil registry residence in Bhutan and of low participation of voters in urban areas in the mock election exercises. The query recently submitted to the ACE Practitioners’ Network provides much more clarity on the problems that are currently being encountered.

With first-time ever elections only four months away, the Election Commission of Bhutan (ECB) is facing the challenge of having to devise, in quite a limited period of time, some effective measures to allow Bhutanese voters to vote at their place of actual residence instead of the one assigned to them through the civil registration records.

Time needs also to be factored in to inform the electorate about the adoption of any extraordinary measures and on how (and when) such measures could be effectively used by them.

Before being able to more clearly identify a solution that is both suitable and feasible, it would be important to know:

• how wide is the extent of the problem. Is this voting residence problem primarily affecting only two specific urban areas? Or will there be voters facing similar constraints also in other parts of the country? This would be crucial to determine whether (a) any measure devised to correct this problem should only apply to a restricted number of voters, or (b) if instead it applies to the whole electorate;

• what is the total size of the Bhutanese electorate;

• if the adoption of the required legal/procedural amendments is possible at this late stage, and how easy (or otherwise complex) would be to introduce the such amendments to the existing legal framework of the elections and, finally, an estimation of the time that a possible amendment process would entail.

Putting aside the options of postal voting (which was ruled out by the Bhutanese Government) and of proxy voting (which could be a bit too hazardous for a first-time ever democratic election), perhaps, given the unique circumstances of Bhutan, the most viable option might be to adopt “special voting” provisions.

As already mentioned by others who have contributed to this query, several “special voting” options could be adopted.

In my opinion, the Mexican “model” by which “special” polling stations are specifically established for voters unable to vote in the constituency in which they are originally registered, would not be advisable because it doesn’t allow to clearly estimate the number of voters that could turn up on election day and, for this reason, it could result in creating more problems than actual benefits. In Mexico, in 2006, I have personally witnessed situations of great confusion, serious tension and disappointment by many voters at all special polling stations visited, where much larger numbers of voters were turning up than those whom a single special polling station could have handled (if I am not wrong, the Mexican electoral law sets a maximum number of 500 special ballots for each special polling station).

It is also important to note that the introduction of a “Mexican” special voting system would require to be backed by a mechanism that efficiently and speedily re-allocates the special votes from the constituency in which they were cast to the respective constituencies of origin, in various parts of the country where such votes must be accounted for. The (actual and perceived) integrity, transparency, reliability and security of such a vote-transmission system should also be carefully weighed.

With these considerations in mind, depending on the available time, on the legal requirements of the electoral process in Bhutan and, most of all, on the number of voters expected to use the “special vote” provision, perhaps, a better option could be to create supplementary voters’ lists for “special voters”.

This would mean that any eligible Bhutanese voter - willing to vote in a different place than the one that was originated from the civil registry records -  could apply for a special vote to be cast in an existing polling station located in the area of his/her permanent and actual residence. These voters would apply for a “special vote” well in advance of election day, during a prescribed period of time, at specifically designated places, possibly in the same polling station in which they intend to vote. The application form would record relevant information about the polling station/constituency of origin and the new polling station/constituency designated by the voter as the one of permanent residence, so that the ECB could ascertain enough in advance how many special votes were requested, where these votes have to be made available and where to re-assign each special vote after it has been cast.

An important element would be to introduce an efficient and reliable verification system, by which each application for a special vote could be accurately verified before being approved. Possibly, the offices receiving the special vote application should be able to easily and quickly verify whether or not the applicant is effectively registered at the constituency/polling station stated in the application form and the new registration will not be validated until confirmation is received that the name of the concerned voter has been removed from the voters’ list of the polling station of origin.

Ideally, some requirements could be introduced to ensure that only those voters actually residing in a place different from the one of registration are able to use special voting provisions. For instance, it could be considered the obligation for a voter to produce some documentary proof of residence (e.g. utility bills, tax invoices, etc.) to unequivocally establish the intended new voting residence within the new constituency where such a voter effectively resides.

On election day, supplementary lists would be used together with “ordinary” voters’ lists in “ordinary” polling stations.  Special votes, once voted, would be inserted in double envelopes (to protect their secrecy) and, at the end of the day, transferred to the exact constituency/polling station to which they pertain.  

There are no doubts that the special voting exercise would require setting up and running a considerably complex logistical operation to be designed and implemented in a relatively short period of time. Furthermore, it has also to be considered that special voting provisions are likely to delay the final tabulation of the election results.

Finally, if all the above is deemed too complex from an operational point of view, or legally unfeasible, or for any other reason it is considered as unsuitable to the context of Bhutan, then, the Government of Bhutan could consider a “milder” and simpler option.

The Government could create incentives for urban voters to travel to the villages where they have been registered to vote, either by:

• providing actual transportation - free of charge - for these voters (special buses for special voters); and/or

• offering partial – or preferably, full - reimbursement of travel costs to those voters who will be able to demonstrate to have travelled to their villages of origin and to have voted.

To ensure that such incentives are adequately understood and utilized by the electorate, all relevant information should be widely disseminated well before election day through an extensive voter information campaign.  

I will describe in some detail how New Zealand does it but will leave the questioner to determine if the situation in his country could sensibly  allow a similar system, especially in the time available.

1 any registered voter can vote at any polling place (although voting at a polling place in the home electorate is naturally encouraged by various means).

2 The voter's list (roll) held by the polling place is that for the electorate only (or if close to a boundary a neighbouring electorate). NZ does not have a general civil registration system or a national identity card.  However most voters will have received though the post 7 days before polling day  a simple EasyVOte card showing their personal details and their electorate.

3  A voter from outside the electorate cannot therefore be immediately checked against the roll. The voter is thus required to complete a detailed declaration stating, amongst other things, where s/he is registered as an elector. If they have their EasyVOte card with them this will help. The voter is then given a ballot paper for the declared elector together with an envelope with 2 sealable pockets. The claimed home electorate is written on the outside of the envelope. The first pocket contains the signed declaration. The voter marks the ballot paper in the normal way and then places the ballot paper in the second pocket, seals it, and deposits the envelope in a special ballot box.

4 At the close of the poll the special ballot box is opened and the number of envelopes is counted and the claimed electorates recorded in the central computer system. The envelopes are then dispatched to the respective home electorates. The votes -known as special votes- are thus not included in the preliminary results which are released on election night. But the number dispatched for each electorate is known.

5 When the envelopes are received in the home electorate the declaration only is extracted at stage 1  and checked against the voter's list for that electorate. Where the check confirms the voter is registered for that electorate the ballot paper is then extracted and the vote is included in the official count released 10 days after polling day. (NZ's proportional voting system allows voters to vote separately for both a party and an electorate candidate so there are other procedures to allow the nationwide party vote if the voter is registered in an electorate other than the one which s/he has stated in the declaration.)

6 In the 10 days between the election night " preliminary count" and the official results NZ conducts elaborate manual and computerised checks post election to identify dual voters and this includes checks to ensure voters using the special vote declaration system I have described have not voted twice.

7 The special voting system generally accounts for about 10% of the total votes cast. (The main use is by people temporarily outside their electorate on polling day.) The system raises additional security issues regarding the distribution of the envelopes around the country, though these are not major in the NZ context. The system is slower for the voter and much more difficult and costly to administer (by a multiple of at least 5) than the ordinary voting system but NZ considers it worthwhile overall from a public policy perspective.

8 Finally I should add the rather obvious point that the special voting system is authorised by legislation and detailed subordinate legislation (regulations). And the system is supported by good information delivered directly to the voter and by public information programmes.

I would like to draw your attention to handbook that International IDEA is producing on the subject of voting from abroad.

While the question above relates to remote voting within the country, this handbook focuses on voting that takes place from abroad. Many of the principles (such as secrecy vs cost considerations,)  and practical solutions (voting methods, logistical arrangements etc) are however as relevant to remote voting from within a country. The formal name of the handbook is "Voting From Abroad - The International IDEA Handbook". This will be available in electronic and paper format in October. The contents include theoretical chapters, case studies, a global overview and other information, the table of contents of the handbook is listed below for your information.

Maria Gratschew, International IDEA

Introduction

Chapter 1. External voting: a comparative overview

Chapter 2. The history and politics of external voting

Chapter 3. The legal framework and an overview of electoral legislation

Chapter 4. Entitlement to vote

Chapter 5. The implementation of external voting

Chapter 6. Host country issues

Chapter 7. The political rights of refugees and displaced persons:

enfranchisement and participation

Chapter 8. The political rights of migrant workers and external voting

Chapter 9. Observation of external voting

Chapter 10. E-voting and external voting

Case studies

Afghanistan’s 2004 presidential election: external voting for a large displaced population

Bosnia and Herzegovina: post-war trends in external voting

Botswana: disappointing results of external voting

Brazil: compulsory voting and renewed interest among external voters

Cape Verde: a large diaspora and low turnout by external voters

Colombia: representation of emigrants in the Congress

The Cook Islands: seat for overseas voters abolished

The Dominican Republic: political agreement in response to demands for

the right to vote from abroad

Estonia: more options for external voting

Honduras: a decision based on political calculations

Indonesia: a long-established system for external voting at diplomatic missions

Iraq: a large diaspora and security concerns

The Marshall Islands: a high proportion of external voters

Mexico: safeguarding the integrity of the electoral process

Mozambique: a system that is too subjective

The Portugal: extended voting rights and decreasing participation

Philippines: the first experience of external voting

Senegal: a significant external electorate

Switzerland: external voting in a federal state with direct democracy

Zimbabwe: highly restrictive provisions

Annex A. External voting: a world survey of 214 countries and territories

Annex B. Glossary of terms

Annex C. References and further reading

Annex D. The cost of external voting: some examples