by Velko Miloev

Introduction

The opportunity to support individual candidates on an open party list within the proportional representation system is widely considered to be a major democratic achievement in both Bulgaria and Bosnia and Herzegovina. However, ‘preferential voting’ has led to some undesired consequences in each of these countries, both of which this paper will consider in turn.

In Bosnia and Herzegovina (hereinafter, BaH or Bosnia for short), open lists have been used successfully since 2000, but key political actors have advocated for a return to closed lists for several years. Following a new wave of allegations of extensive abuse of the preference system in the 2014 elections, the issue is likely to continue to fuel the ongoing debate surrounding electoral reform.

In Bulgaria, like in BaH, public support for open lists is linked to the widespread disillusionment with political elites. However, unlike in BaH, where open lists were initially easily introduced when post-war elections were still internationally administered, in Bulgaria it took years of pressure from civil society until open lists were finally regulated in a manner that gave voters a practical opportunity to reorder candidates in 2014. The results led to renewed calls on behalf of some parties to reconsider their use.

In BaH, a voter may mark as many individual candidates as there are names on the selected party list without providing a preference order. Possibly the most serious problem encountered on Election Days is the addition of fake preferences by corrupt poll workers during the count. In essence, candidates organize the stealing of votes within a party list.

In Bulgaria, the ballot paper includes the numbers and names of all political entities along with boxes with digits to tick the list number of one preferred candidate if a voter chooses to do so. In 2014, many voters made a politically meaningful use of this option, although various landslide victories of hardly known candidates lucky to have the same number on their party list as the ballot number of the party itself marred the 2014 elections.

This case study will examine a number of issues including legal framework and ballot design, the roles of political parties and civil society in the electoral debate, campaign techniques, and voter education.

The Bosnian case – a democratic achievement marred by fraud?

Legal framework and history of the open lists

Bosnia and Herzegovina currently uses a proportional representation (PR) system with open party lists for the lower houses of the national parliament and the parliaments of the two semi-autonomous entities, the cantonal assemblies, and the municipal councils’ elections.[1]

A voter may mark an unlimited number of candidates within a selected list. All of their names appear on the ballot paper, which makes it relatively large (up to A2 size), but not necessarily difficult to manage. Mandates are allocated in multi-member electoral districts first by ‘political subjects’ (parties, coalitions, independent candidates) using the Sainte-Laguë formula. Then, within each winning list, seats are assigned to candidates starting with those with the highest number of preferences as long as they constitute not less than 5 percent of all valid votes for the party.[2]

Figure 1 - Detail from a training material produced by the Central Election Commission of BaH showing how to mark a ballot

The Bosnian public has warmly welcomed the concept of the open list since its introduction. In 1999 this author, as project manager at one of the BaH offices of the International Foundation for Electoral Systems, had the chance to supervise locally hired IFES trainers supporting a country-wide OSCE led public information campaign on BaH’s new Electoral Law. Face-to-face community activities clearly confirmed that the concept of voting for individuals was well understood and appreciated. Comments from citizens indicated growing dissatisfaction with political parties’ underperformance and lack of accountability quite early in the beginning of Bosnia’s democratization process after the 1992-1996 war.

The open lists were first tested in local elections in April 2000 and, subsequently, in the general elections in the autumn of the same year.[3] While the author of this study was unable to find any statistical data on the implementation of the open lists in the country, a quick review of the electoral results in 2000 highlighted the success of the open lists – approximately 20 percent of the new municipal councilors would not have been elected without preferences.

Ten years later, Bosnian civil society had to defend this system, while at various points over the last 5 years at least four major political parties have advocated closing the lists.[4]

The pros and cons, real or fake arguments for closing the lists

A public statement by the Party of Democratic Action (SDA) made in January 2010 stated the following reasons for abandoning the open lists:[5]

- Securing representation of all constituent peoples and the others in all BaH parliaments;[6]

- Securing the representation of women and youth;

- Equal territorial representation in multi-member electoral districts;

- Satisfactory qualification levels of elected representatives;

- Preventing electoral fraud through adding [fake] preferences by the polling station committees (PSCs).

‘Eliminating electoral engineering’ and ‘better minority gender representation’ were again arguments presented in an SNSD-SDP agreement on legislative action from 2012.[7]

Open and closed lists were debated in the run up to the 2010 and 2014 general elections, and the issue has also been on the agenda of the official working groups tasked with proposing changes in the Electoral Law.[8]

As of this writing, the candidates’ lists in BaH are still open for several reasons, including the general inability of Bosnian political players to reach parliamentary consensus on any reform (including electoral reform) and pressure from civil society groups like NGOs and the media. Civil society qualified the eventual closing of the lists as “another step towards ‘partocracy’ and decision making limited to closed centers of power,” which would lead to “further limitation of citizens’ influence over political developments.”[9] Other observers or party representatives noted that closing the lists would be a step backwards from a democratic achievement and would negatively impact voter turnout.[10]

Indeed, most of the ‘closing’ arguments appear, to a certain extent, to be falsified and designed only to hide the parties’ desire to restore control over who will be elected as MP.

While gender representation is indeed an issue in Bosnia’s parliaments, it does not appear to lie in the center of the electoral debate. As in other countries that combine gender quotas and open lists, there is conflict between these two positive aspirations. Amendments to the Electoral Law before the 2014 elections raised the quota for the ‘less represented gender’ from 25 to 40 percent, with female names equally distributed from top to bottom on each candidates’ list. Ultimately only 19 percent of seats in the state and entity legislatures were won by women as a consequence of voters’ preferences.[11] But parties have never asked citizens whether they would be willing to sacrifice the open lists for the sake of greater female representation.

In addition, the attention given to minorities does not appear to be genuine. Bosnia’s electoral legal framework still contains ethnicity based constitutional limitations of both active and passive suffrage rights, particularly affecting those who do not identify themselves with one of the three constituent peoples. These limitations contradict a number of binding documents or recommendations of the OSCE, the PACE, the EU, and a 2009 decision of the European Court for Human Rights.  As the OSCE/ODIHR observation mission noted in its final report on the 2014 elections, “it is regrettable that this is the second election held in violation of the Sejdić-Finci judgment of the ECtHR.”[12]

Other ‘closing arguments’ are also far from convincing. Seeking better territorial representation is somewhat irrelevant in the case of relatively small electoral units in BaH, and voters using preferences would certainly consider candidates who are likely to come and talk to their constituency after being elected. It also unlikely that voters are eager to spoil the ‘qualification levels’ of MPs by backing the least qualified candidates.

All these arguments considered, the allegations of manipulating the open list in various elections since 2008 appear credible.

The abuse of the open lists – a specific BaH problem

The addition of fake preferences to individual candidates by members of a polling station committee (PSC) is probably the most worrying Election Day violation in Bosnia. This fraudulent behavior requires complicity among both PSC members and those present during the count.[13] Such complicity is reportedly not difficult to establish since the fake preferences mean stealing votes from within a list rather than from parties. Hence, a PSC member charged to count the individual votes on the stack of ballots for the party which nominated him as a poll worker may simply add X’s next to certain candidates’ names. PSCs in Bosnia are appointed exclusively on the basis of party nominations through a lottery. For example, once party A gets entitled to certain positions at certain polling stations, it may offer these to party B, which would nominate its loyal supporters (formally on behalf of party A). This can be a trade-off for cash, or A and B might “exchange” poll workers’ positions in areas where each of them has a stronger political interest to win. This practice complements parties illegally paying poll workers and intimidating or bribing observers. On paper, a PSC’s composition appears to be in complete compliance with the rule that none of the political contestants may be represented with more than one member, but in reality a given PSC might be in full control of one or two political parties.

A consensual crime is difficult to observe and prove independently, and the Central Election Commission of Bosnia and Herzegovina (CEC BaH) does not publish election results at the polling station level that are detailed enough to allow a thorough analysis. Thus, reports about modifying preference votes are mostly anecdotal. There are some clearer indications of fraud, such as when a candidate that is neither a resident nor popular in a certain area wins hundreds of preference votes at a certain polling station and zero preference votes at another polling station in the same community. It is also highly suspicious when a candidate receives preferences tallying close to the total number of votes for their party. Finally, many marginal political subjects, particularly independent candidates and some ‘phantom’ parties, receive almost no votes. Observers believe that they registered for the elections solely to become entitled to positions in the PSCs that they would subsequently trade off.

For the first time in a decade, the general elections in October 2014 saw a large domestic non-partisan observation effort. A coalition of seven NGOs named “Pod lupom” (“under the magnifying glass”) covered roughly 25 percent of all polling stations based on a random sample throughout the whole   country.[14] Only 0.6 percent of the coalition observers reported witnessing fake preferences added during the count. This can be seen as evidence that independent observers can deter fraud.

In the same time, in a number of official and informal interviews conducted by this author during late autumn 2014 in Sarajevo with members of the electoral administration and high ranking party representatives, the abuse of the open lists was indicated again as a widespread problem at polling stations unattended by non-partisan observers or at least an issue of major concern. Some interlocutors proposed the closure of the lists.

Saving the open lists – possible solutions

Many stakeholders from the electoral administration and civil society insist on replacing the current political model for appointments to PSCs with a mixed political-professional model. This implies that the chairperson and some members are selected by the municipal electoral commissions from among public workers.[15]

In addition, Bosnians could use open lists throughout Europe and the rest of the world as examples, paying particular attention to the number of preferences each voter is allowed as well as the design of the ballot.[16]

Limiting a voter’s preference to two or three candidates would likely fall within the realistic scope of choice that an average voter may wish to have and would make any other additional marks on the ballot paper easier to detect. If only one personal vote within a party list is allowed, any additional marks will make the ballot valid for the list only.

Using a stamp as a ballot marking tool will make fraudulent activities easier to control as pens and pencils are – contrary to the rules in BaH – usually available on the counting table.

A simpler ballot design without candidates’ names would allow PSC members and observers to more efficiently control each other during the count and would also resolve other technical issues Bosnia has with printing, packing, and controlling the ballot papers during the process. Such a ballot would only contain boxes to mark the list number(s) of the preferred candidate(s).

Figure  SEQ Figure 2 Fragment from a ballot paper used for the 2013 municipal council elections on Kosovo. Note that the boxes with the list numbers of the candidates start from #2 - no preference marked automatically counts for the #1 on the list.

Furthermore, a write-in ballot (e.g., with just two boxes to enter digits) would be very unusual in Bosnia but would almost fully eliminate the options for changing the voter’s preference.

More importantly, it is hoped that continued and widened presence of domestic non-partisan observers will further limit various types of fraud committed at polling stations. More detailed results published by the CEC would enable observers to use statistical tools to detect any signs of fraud, including suspicious deviations in numbers of preferences for candidates.

By mid-September 2015, the usual inter-agency working group for amending the legislation will be just about to start its work. It remains to be seen what new proposals will be put on the table. Much depends on the willingness of politicians to accept the recommendations from civil society, electoral administration, and the international community.

The electoral reform in Bulgaria - citizens vs. the political establishment

History of the debate over the open lists

Bulgarians voted with open lists for the first time in the European Parliament (EP) elections in May 2007, just a few months after the country became a member of the European Union. For a candidate to move up the list, they needed to receive a number of preferences that equated at least 15 percent of all valid votes for the party. Few people believed that this process could work and indeed, it did not. Since then, the size of the threshold required for preferences to take effect has been a key issue in the electoral debate. Moreover, since 2007, the laws provide for only one optional preference.

In 2011, the then ruling centrist GERB party[17] passed the first ever Bulgarian Electoral Code, which replaced the four laws that previously regulated different types of elections. The threshold for the preferences to count was nine percent for the National Assembly and six percent for the EP elections. However, before the national elections in 2013, GERB and a majority of MPs agreed to delete the words ‘preferential voting’ from the Electoral Code. Apparently, the politicians feared the unpredictable impact of the implementation of open lists for the first time in unequally-sized electoral districts[18] and were concerned about ‘brother-against-brother wars’ between candidates of the same party.

But there was also a key technical issue – the MPs did not know how to design the ballot. The assumption at the time was that all candidates’ names must be printed on an open list ballot paper. With some 20-30 parties and coalitions normally running in national elections, the ballot would have become uncommonly large compared to what Bulgarians were used to. Thus, the code envisaged that there would be separate ballots for each party placed in ‘modular boxes’ inside the polling booths. An electronic device would issue just one ballot once the voter pressed the button for the selected party.[19] It turned out that the drafters of the code did not know how to design and produce these devices.[20] Civil society organizations and some politicians tried to defend the open lists but lost this round of the debate.[21]

Following the 2013 elections, a new parliamentary majority formed around the Bulgarian Socialist Party (BSP) opted for drafting a new Electoral Code. In an effort to demonstrate better responsiveness to the citizens’ requests, they formed a Public Council to the Interim Parliamentary Committee for developing the code. An official report, summarizing the various proposals, highlighted that NGOs as well as high-ranking independent experts almost unanimously backed open lists in all PR elections with low or zero thresholds.[22] A smaller number insisted that it should be mandatory for voters to mark several preferences (each of which would have equal weight as is the case in Bosnia). The most common phrase in this report, also evident in the Bulgarian electoral debates over the last few years, was ‘majoritarian element.’

Technically speaking, a ‘majoritarian element’ involves introducing a mixed electoral system with a certain number of MPs elected from open lists and others elected in a plurality or majority voting system in single-member constituencies. The general political context in Bulgaria shares some attributes with the situation in Bosnia and Herzegovina. In Bulgaria there are concerns that parties are accumulating too much power and use it to facilitate non-transparent political deals in support of various economic interests. Such concerns explain why citizens desire selecting their representatives through open lists. The political establishment, on the other hand, prefers to choose their own disciplined members of parliament.

One striking instance of unprincipled, ad hoc parliamentary majorities making decisions against the prevailing public opinion took place in 2015, when the parties in the National assembly cancelled the planned referendum on the introduction of a mixed proportional-plurality/majority electoral system. This cancellation happened despite the fact that two of the parties had previously helped to collect the 500,000 signatures needed to initiate the referendum.[23]

Looking back at 2013, the debate over the threshold for the open lists also clearly showed the parties’ desire to retain control over the people’s future representatives. Parties are legitimately concerned about the fact that with a very low threshold, small organized groups would be able to change the majority of voters’ will.[24]  Nonetheless, they rejected the possible solution of making preference voting mandatory, and some insisted on a 20 percent threshold despite the failure of a 15 percent threshold in Bulgaria. The strongest opponent of the open lists, the Movement for Rights and Freedoms (DPS), requested a 50 percent threshold.[25]

The new Electoral Code was passed in February 2014 and kept the PR system in place with open lists and one optional preference. It provides that for preferences to become effective, a candidate must receive at least 7 percent of all valid votes for the list in an election to the National assembly, 5 percent in elections to the EP, and 7 percent from the municipal electoral quota in elections for local councils.[26] This was a significant disappointment for many who wanted a ‘majoritarian element’ and lower thresholds.[27] A further disappointment was the rule that if no preference had been marked on the ballot, it would count as one for the top candidate.[28]

The Sudoku ballot paper and its consequences

Despite these dissatisfactions with the legal provisions, the preferences worked in subsequent elections. Civil society had succeeded not only in achieving non-excessive thresholds, but also in introducing a new ballot design. At the Public Council, the Head of the Parliamentary committee demonstrated a patchwork of ballots from previous elections (with all candidates’ names) pasted on an A1 size sheet in an effort to demonstrate that the open list ballot paper would be unmanageable. A member then proceeded to pass to the Council a sample ballot paper from a Kosovo election, which this author had observed (see Figure 2). The design was accepted with a mixture of enthusiasm and skepticism.

Figure 3 - The initial generic design of the ballot for the 2014 Bulgarian election to the European Parliament.

The section on the right – the boxes for marking individual candidates with respective list numbers - became known as ‘Sudoku.’ Subsequently, the DPS proposed that the squares should be replaced with circles so that parties opposing the open lists could more easily instruct their supporters to not bother with preferences.

According to different exit polls and observers’ reports, between 26 and 42 percent of Bulgarian voters marked a preference in the May 2014 EP elections. Four out of the 17 Bulgarian MEPs were elected as a result of these individual votes.[29] This included a case of particular political importance for the newly established centrist coalition, named the Reformist Block, in which the second candidate on the list defeated the top candidate and won the only RB seat. Even more astonishing was the case of the BSP, which as a party had the ballot number 15. The then hardly known young candidate with the list number 15 received 34,124 preferences while the party’s leader, placed on top of the list, only received 28,039. This became known as the ‘15/15 effect.’ Such results happened again at the National assembly elections in October of the same year (as 9/9, 11/11, 18/18 etc.). A number of factors contributed to this undesired phenomenon.

First, the party campaigns frequently emphasize the list number of the party. This can lead to much confusion, particularly among rural, less educated segments of the electorate. Second, most candidates fell well short of campaigning for their preferences, while the voter education efforts of the CEC were limited and inadequate. A 38 second long, poorly produced video proved unable to familiarize the public with the new ballot design, and even well-informed observers struggled to find the official sample ballot, well hidden among hundreds of bylaws and attachments on the CEC website. Indeed, even civil society did very little to educate citizens about the new rules.

Still a success story, but an uncertain future remains

A study published by the Bulgarian NGO IPED used results protocols from multi-member districts to show that in October 2014, 35 percent of voters marked a preference, with these voters unevenly distributed among parties and regions. Twenty out of 240 members of the National assembly were elected because of preferential voting.[30]

Even a quick, random review of official results reveals many cases in which low ranked candidates with list numbers the same as the party ballot number received a surprising number of individual votes. According to IPED, the ‘15/15 effect’ accounts for 12.3 percent of all preference votes. Several winning candidates declined their seats, admitting they were undeserved.

Still, in 2014, the open list worked for the first time in the country despite five and seven percent thresholds, low expectations, and political opposition.[31] Some public statements indicated that parties were taken by surprise by the impact of the open lists. Some unclear and unsubstantiated allegations that preferences led to large numbers of invalid votes suggest the use of the ‘15/15’ phenomena as an excuse to try to close the lists.[32]

The history of electoral reform in Bulgaria has seen examples of how making amendments to legislation in order to meet public demands on one issue can be used to reverse achievements on another. Civil society should remain on high alert and fight to hold onto previously won accomplishments in official formats for debates. Furthermore, CSOs could take the responsibility to implement CEC sponsored voter education programs.  A more aggressive approach to attracting citizens with high quality research and proposals will only aid in bringing about electoral reform.