by Ben Goldsmith

Increasingly, elections are viewed as a crucial peace-building tool in post-conflict states. While the transition from non-democratic to democratic political traditions is rarely smooth, it is particularly challenging when the transition follows a war. However, post-conflict elections can jump-start the democratization process by offering all citizens the opportunity to be involved, and their involvement is critical to giving elected institutions the legitimacy they need to govern in often difficult circumstances. For this reason, post-conflict elections are often held earlier than some think they should be, sometimes even when widespread insecurity means that they cannot be genuinely free and fair.  

One key decision officials make when organizing a post-conflict election—like those held recently in Iraq and Afghanistan1— is how to deal with a country’s external populations. These populations may include refugees who fled a conflict for neighboring countries and members of national diasporas (who may have left the country for political or economic reasons, have since settled elsewhere, but retain ties to their homelands). Will they be eligible to vote? In the post-conflict elections held in Bosnia, Kosovo and East Timor, refugees were allowed to vote.  However, setting up out-of-country voting (OCV) for external populations is a lot more complex than organizing in-country polling, and taking on this logistical and financial burden in the  challenging environment of a post-conflict election is rarely without risks. In the end, it could even endanger the credibility of the wider electoral process.

Beyond financial and logistical concerns, the decision to organize external voting can also be politically difficult, especially when the external population is substantial in number. A large external population could change the outcome of the election, which may not be politically acceptable in-country. While it is rarely possible to enable all of the external population to vote, politically it may be easier to exclude all rather than include only some.

Despite these challenges, there are often strong pressures to include external communities in post-conflict elections. Given the ongoing instability in the period immediately following the end of the conflict, those who wish to return home are often unable to do so even though they may have a great desire to be involved in post-conflict developments. Diaspora and refugee communities have often suffered greatly; in addition, they may maintain significant influence in their country of origin.

Including external populations in post-conflict elections can also serve to re-establish ties between them and citizens in-country, which can be very beneficial for the reconstruction process. Diaspora populations have often developed financial resources and professional qualifications in exile that can be useful during rebuilding. Engaging them in the political process makes more likely their return and active participation in reconstruction.

(The complexity and financial burden of external voting also affects the sustainability of such programs. While many post-conflict elections are heavily supported by the international community, once the initial elections have been conducted, this support diminishes. This can leave post-conflict countries financially and institutionally challenged to continue delivering external voting programs. However, it can easily be argued that external voting is more important in the immediate post-conflict elections and that subsequently, when a real opportunity for return exists, there is less reason for such external voting options. This reality mitigates the arguments on sustainability, although once a decision has been taken to enfranchise the external  populations, it may be politically difficult to retract later on.)

The OCV operations conducted in the recent Afghan and Iraqi elections were, in many ways, very different, but the experiences of these voting programs serve to identify many of the issues surrounding the conduct of OCV programs in post-conflict elections.

Timeline

Both the Afghan and Iraqi electoral administration bodies made the decision to enfranchise their external voting populations very late in the process. In each case, these complex programs had to be set up in approximately 80 days—an extremely tight timeline. With the Afghan Joint Electoral Management Body and the Independent Electoral Commission of Iraq already challenged to run their respective in-country elections—and there being no other domestic organization with the capacity to conduct an OCV program—the decision was made to contract a strategic partnership between the International Organization for Migration (IOM) and IFES to conduct both OCV programs. Due to the time constraints, both programs were led by large numbers of international electoral experts, which unfortunately gave Afghans and Iraqis little opportunity to build their capacity to manage the process themselves in future elections.

The budgets for both OCV operations were large (US$92 million was budgeted in the case of Iraq), and in many ways money was used to mitigate the time constraints on the operations. Security was a concern in both in-country elections, and the threat of violence extended to the OCV operations. A considerable part of the budgets for both operations went to efforts to ensure security at the polls.

Scope of Operations

The short timeline impacted the scope of OCV operations as well. Due to the dispersed nature of the Afghan and Iraqi external populations, it was not possible to provide polling services for everyone. Instead, OCV operations focused on the largest concentrations of expatriates. This meant that a significant part of the diaspora community (in both elections) was not provided for, but it was determined that to enfranchise some was better than none at all.

While it was very difficult to estimate the exact number of expatriates living outside of either Afghanistan or Iraq, both countries have sizeable populations of eligible external voters. Large Afghan refugee communities exist, primarily in Pakistan and Iran. Best estimates indicated that there were 500,000 eligible refugees in Iran and 1.2 million in Pakistan (all Afghan nationals over the age of 18 were eligible), although government and UNHCR estimates varied. Similarly, many Iraqis had fled the regime of Saddam Hussein and sought refuge in countries around the world. Iraq’s election commission asked the IOM to operate in 14 countries with significant Iraqi diaspora populations. 2 Again, population data was difficult to come by as the Iraqi diaspora was in many cases well integrated into the host countries; often its members had been there for many years and had been granted citizenship. Best estimates indicated that there could be as many as 1.2 million eligible Iraqis living in the chosen 14 countries. However, from an operational perspective, it was important that the higher estimates were used for planning purposes in order to be prepared to deal with all potential voters.

The tight timeline had a particular impact on which Iraqi expatriates were able to vote. The Iraqi diaspora was particularly spread out, and a number of countries outside of the designated 14 claimed larger populations of Iraqis than those living in the 14—including Switzerland, Norway and Lebanon. In addition, the distribution of Iraqi expatriates within the 14 countries made locating the registration and polling centers there a controversial issue. For example, in the United States, a large community of Iraqi Kurds lives in Nashville, Tennessee— which did have a voting center—while a large community of Iraqi ChaldoAssyrians (who are Christian) lives in San Diego—which did not have a voting center (though there was one in Los Angeles). Many groups demanded that centers be established in many additional cities in the 14 countries, but operationally it was not possible to meet all of these demands.

After establishing the territory in which OCV operations would be held, the next hurdle involved negotiating with governments of the host countries to conduct external voting programs in their territories. In the Afghan case, memoranda of understanding (MoUs) already existed for both the Iranian and Pakistani governments that covered the activities of the OCV program. Despite this, a number of issues arose. Visa approvals in Iran were problematic, and as a result, many staff arrived very late in the process, leaving them little time to implement the program. The Pakistani government had a number of interests in the OCV operation. Among other things, it wanted access to the voter registration data so that it could identify subversive elements who might attempt to disrupt the electoral process. However, the OCV program was unable to support this request as it would have been in violation of data protection principles and contrary to international external voting standards.

In the Iraqi case, no MoUs were in place with any of the 14 countries. In some instances, this MoU negotiation process was smooth; in other cases negotiations were difficult, with considerable demands being made on the OCV operation. Some countries were reluctant to allow in-person registration and voting because of security concerns but were persuaded when it became clear that there was not sufficient time for the alternative: postal registration and voting. Ironically, it was most difficult to conduct the OCV program in some of the more established democracies as agreements between central governments did not necessarily translate into support from local government institutions. In more authoritarian countries, once the central approval for the operation was obtained, all levels of government actively supported the OCV program.

During Afghanistan’s 2004 presidential election, refugee voting services were provided through eight field offices in Pakistan and seven in Iran. In total, 180 international staff were recruited and deployed; 16 separate offices established (15 field offices and a headquarters); 260 core Afghan, Pakistani and Iranian staff employed for the duration of the program; and over 20,000 registration and polling staff employed and trained over a short period before Election Day. For Iraq’s 2005 parliamentary elections, more than 400 international elections staff and more than 7,000 local staff were employed during the registration and polling periods. Registration and voting operations took place in 74 centers in 36 cities across 14 countries.

Determining Eligibility

In both operations everyone involved accepted that the pre-registration of voters was essential to the integrity of the process. In any election, it is very important that only eligible voters be allowed to participate. In the context of external voting, identifying eligible voters is a particular concern—given that you are operating outside of the country holding the election— as you cannot assume that residence indicates eligibility. At pre-registration, potential voters demonstrate their status using approved documents. However, to combine determining eligibility with the act of voting would make the transaction time in the polling station too long, slowing the processing of voters on Election Day and limiting the number of people who could vote as a result. In addition, the list of registered voters generated by the pre-registration process can be reviewed by members of the voting community and examined for possible errors or the registration of ineligible people. The opportunity for community review is an important element of the electoral process and one possible only with the pre-registration of voters at a time prior to Election Day.

The short timeline of both Afghan and Iraqi elections meant that registration had to be conducted quickly. It was not possible to conduct this process of registration and voting by mail, even in the countries in which a secure postal system existed. Instead, a short period of intensive registration was conducted in the same locations where polling would take place. Afterwards, hard-copy voter lists were assembled, made available for review and challenges, and remained onsite for polling on Election Day.

This was the procedure in all OCV locations for the Afghan and Iraqi elections with the exception of Iran in 2004. In Iran, many Afghan refugees had already been registered and issued an Amayesh card (indicating their refugee status), and initially OCV officials decided to accept this card as proof of eligibility to participate in the Afghan election. However, the Amayesh database proved to be out of date as many refugees had returned home since registering and some still in Iran had never collected their cards after registration. In the end, it proved necessary to allow Afghans living in Iran to use other forms of documentation to prove their voting eligibility. In Pakistan, as no register already existed, eligible Afghans went through pre-registration.

As no register of eligible Iraqis existed in any of the 14 designated countries, a registration process was required that took place over nine days.  Initially, Iraq’s electoral commission required that eligible Iraqis provide two documents to prove their identity, their nationality (eligible voters must be Iraqi citizens, entitled to reclaim citizenship or eligible for Iraqi citizenship), and their age (eligible voters must have been born before 31 December 1986). As the registration period began, it became apparent that many Iraqis were struggling to provide the required documents and, in the end, certain single documents that proved all aspects of eligibility (such as an Iraqi passport) were accepted.

Assessment of OCV Operations in Afghanistan and Iraq

Overall, the OCV programs in Afghanistan and Iraq should be considered a success. Given the incredibly challenging timelines and significant operational complexities, the fact that they happened at all is evidence of considerable success. That functioning institutions—institutions that were able to plan and implement complex electoral processes in challenging environments—were created from nothing in approximately 80 days represents incredible achievements.

Approximately 850,000 Afghan refugees participated in the election of their president through the OCV program in Iran and Pakistan, which represented about 10% of those voting in the election and about 50% of the estimated eligible population. Much of this success can be attributed to a comprehensive voter education campaign conducted through local community structures to inform Afghan refugees about the election.

Participation rates in Iraq’s OCV program were much lower, with only 265,000 of the estimated 1.2 million eligible Iraqis living in the 14 designated countries taking the opportunity to vote—about 22% of the estimated eligible population. Turnout was particularly low in countries like Jordan, Syria, the UAE and Turkey.

In analyzing this outcome, a number of reasons for this lower-than-expected rate of participation were discovered. In the listed countries, the Iraqi diaspora is subject to fears of repatriation, and many felt that participation in the election might be used to return them to Iraq. In addition, many saw the election as a U.S.-sponsored event and, as a result, did not want to participate.

Others complained that they did not know many of the more than 100 political party or candidate entities on the ballot and therefore could not be expected to make a choice among them. Voter disinterest was also a major factor; many Iraqis had left Iraq decades ago and were well integrated in their new homes. This fact limited their interest in returning to Iraq and hence in the electoral process there. Given this context, the lower-than-expected turnout is more understandable.

Conclusion

Irrespective of the advantages and disadvantages of enfranchising external populations in post-conflict elections, the decision to do so should not be left as late as was the case in these two recent OCV programs. While ultimately both OCV programs were implemented successfully, the compressed timeframe made the possibility of failure too great, and such failure could have had negative consequences for the perceived success of the wider electoral processes. Nevertheless, the symbolic value of including such populations cannot be underestimated. The joyful celebrations that were seen amongst the Iraqi diaspora and Afghan refugees when given the opportunity to be part of the post-conflict electoral process cannot be denied, and this human dimension to external enfranchisement is a powerful and abiding memory for many of these electoral events.

Ben Goldsmith is an international election expert who has been engaged in elections in the UK, Bosnia, Kosovo, Georgia, Iraq and Afghanistan. He is currently the Deputy Chief Electoral Officer of Afghanistan.

Notes:

1 In October 2004, Afghans elected their president and, in January 2005, Iraqis elected a National Assembly, Governorate Councils and a Kurdish National Assembly.

2 The 14 countries selected were Australia, Canada, Denmark, France, Germany, Iran, Jordan, the Netherlands, Sweden, Syria, Turkey, the United Arab Emirates, the United Kingdom and the United States.