Though generally seen as the arcane province of election experts, systems of representation matter—they shape electoral processes and results. In December in Iraq and January in the Palestinian Authority, this lesson was reiterated. In both cases, one political party (coincidentally, both were Islamist) won a larger share of seats in the newly elected assembly than of the popular vote. In Iraq, a coalition of mainly Shiite Islamist parties won more than 41 percent of the vote and more than twice as many legislative seats (128 of 275) as its closest competitor—not an outright majority, but a strong plurality. The Palestinian results were even more dramatic, with the Islamist party Hamas (considered a terrorist organization by the United States and the European Union) winning a small plurality of the popular vote (about 45 percent) and an outright majority of seats (almost 60 percent) in the new Palestinian Legislative Council (PLC).

In both elections, the Islamist parties benefited from disproportionalities in the system of representation. This does not suggest that their victories were illegitimate. Election results are often disproportionate to the popular vote. For example, in the 2005 British elections, Labour won 35 percent of the vote to the Tories’ 32 percent, but Labour holds 55 percent of the parliamentary seats and the Tories only 31 percent. This is the case because, while proportionality is one important attribute of systems of representation, they are often used to achieve other goals as well, such as ensuring a strong link between a geographic community and its representatives or guaranteeing representation for minority groups.

To understand the current political situation in the Palestinian Authority and Iraq, it is worth exploring the systems of representation used in both elections. This essay will present a brief overview of these systems, their effects on their respective elections and some legislative history of their adoption. In the case of the Palestinian Authority, its system was adopted for reasons grounded in local politics, and one party was better able to compete under that system. The result is a dramatic political shift as Hamas takes power on the basis of a narrow plurality of the popular vote. In Iraq, the system of representation was adopted in a more haphazard way and has had more haphazard and less dramatic effects.

The Palestinian Authority
The Palestinian Authority uses a mixed system of representation to elect members of the PLC. Half of the seats (66) are elected using a system of list proportional representation (List PR). Under this system, each voter selects a single political party, and the parties win a share of the seats roughly proportional to their share of the votes. Hamas and Fatah were closely matched in this part of the election, with Hamas winning 44 percent of the vote and 29 seats to Fatah’s 41 percent and 28 seats. These totals are probably the best reflection of how popular the two parties’ policies and platforms are with the voters, and they reveal a closely divided electorate.

The other 66 PLC seats are elected from 16 districts using the bloc vote. In this case, each district is assigned a number of PLC members based on itspopulation—the smallest have one, the largest have nine. In each district, a voter can choose a number of candidates up to the number of available seats. This system is not widely used internationally, especially for elections in which there are well organized and highly competitive political parties. It can exaggerate the share of seats won by a large and well organized party, especially if other major parties are poorly organized.

This is exactly what happened in the Palestinian elections. Counting both official party candidates and affiliated independent candidates, Fatah had more than three candidates running for every seat. In contrast, Hamas offered approximately one candidate (official or affiliated) for every seat. Fatah’s inability to agree on a single group of candidates caused its supporters to divide their votes among many different options.

The consequences were tremendous. Fatah and Fatah-affiliated candidates won 44 percent of the vote while Hamas and Hamas-affiliated candidates earned 45 percent, but Fatah won only one-third as many seats. Table 1 presents the List PR vote totals, the actual results in seats, and two counterfactual scenarios. In the first, all 132 seats in the PLC are distributed using List PR. In the second, the votes for Fatah-affiliated candidates are redistributed to official Fatah candidates. The first scenario reveals that Hamas’ small plurality in the popular vote would be reflected in a small plurality in seats, but Fatah would place a close second and might be better positioned to form a coalition government. The second scenario shows that Fatah might even have won a small plurality of the seats despite placing second to Hamas in the popular vote (such an outcome derives from the structure of the constituencies and some quirks of calculating the counterfactual scenario).

The effects of the system of representation used in the PLC elections can only be described as overwhelming. Palestinians voting on January 25 were evenly split between Hamas and Fatah, evenly split between a platform of intransigence toward Israel and a platform of continued efforts to negotiate a two-state solution. Why was this system of representation used? Why did Fatah (as the ruling party) adopt an electoral law capable of damaging its own party interest?

The answer lies, in part, in legitimate legislative objectives. Given the traditional patterns of authority in Palestine and the interests of its various regions, Palestinians expressed a strong desire to vote for individual representatives linked to geographical communities, and the bloc vote (with candidates tied to districts) accomplishes such an objective. In addition, personal corruption of senior officials is widely felt to be a central problem in the Palestinian Authority. As a result, many Palestinians wanted a system that would make legislators specifically accountable to the voters. The bloc vote, which unlike List PR lets voters select specific candidates, achieved this goal of accountability. Forty-five members of the sitting PLC ran for the 66 constituency seats; only six won.
Of course, there are other ways of achieving similar objectives with less risk of such extreme disproportionality. To pick two of many possibilities, PLC members could have been elected on the basis of proportional representation in constituencies using open candidate lists to allow voters—not parties—to determine which candidates would take seats. Or, they could have been elected using first-past-the-post (candidate who wins the most votes wins) in constituencies but with compensatory seats used to balance the outcome, rather like the German system, which ensures a reasonable level of proportionality in the final assembly.

Although discussed when the electoral law was debated in the PLC, these options (and others) never gained momentum. The bloc vote was used in the 1996 elections for all PLC seats (88 at the time), and the members of the sitting PLC were afflicted by incumbents’ general reluctance to reform the system. They had been elected under the bloc vote system once and believed that they could win again. However, Hamas did not participate in the 1996 elections, and sitting legislators were not able to appreciate the difference between using a bloc vote system with only one major party and using it with two major competitors. Some did see that the bloc vote system could harm Fatah, but none fully recognized to what degree.

Iraq
Iraqis have tested different systems of representation in the post-Saddam era. In the January 2005 election, they used a system of proportional representation in a single national constituency. Although the results of that election were widely accepted as credible, some Iraqis felt this system gave too much power to the political parties, neglected Iraq’s very real regional differences and possibly underrepresented the Sunni minority, which had not been able to vote in as large numbers as the Shiite and Kurdish communities due to the security situation in their region.

Following that first election, members of Iraq’s Transitional National Assembly (TNA) began discussing other options. However, their choices were severely constrained by the operational realities of conducting elections on very short schedules and in the midst of military conflict. For example, the nature of the voter register required that Iraqi constituencies follow the existing administrative boundaries of the governorates (it would not have been possible to conduct new voter registration or redistricting prior to the December 15 election). Given these constraints, TNA members quickly turned their attention to a List PR system within governorate-level constituencies. This solution was seen to address the two most pressing political complaints about January’s single national constituency in that the smaller constituencies recognized the federal aspirations of the Shiites and the Kurds and also guaranteed a minimum number of seats to the predominantly Sunni governorates, regardless of the turnout in those areas.

The main Iraqi concern about the proposed system was that it would disadvantage minorities that might have enough support nationwide to win a seat but not enough support in any particular governorate. In addition, many Iraqi legislators expressed concern about “wasted” votes in the constituencies. These concerns are part of a broader concern for proportionality in the system of representation as List PR divided by constituencies is likely to be less proportional overall than List PR in a single national constituency. To address these concerns, international election advisors proposed the German model of compensatory seats. Under this model, most seats would be awarded to winning lists in each constituency, but a small number would be awarded at the national level to bring the results closer to “ideal” proportionality with the popular vote.

In the end, the system adopted was a strangepseudo-compensatoryhybrid. Overall, 230 seats were apportioned to governorate-level constituencies and 45 were reserved as national and compensatory seats. The latter category was awarded first, but only one party won enough votes nationwide (without winning any seats in a governorate) to qualify for a seat. The remaining 44 “national” seats were awarded to the parties that had won seats in the governorates—in proportion to their share of the nationwide vote. However, the allocation of these 44 seats was calculated without considering the allocation of the 230 constituency seats, and therefore the process did not compensate parties that were disproportionately underrepresented in the constituency seats.

Such an arrangement rewards parties whose support is strongly concentrated in particular constituencies rather than spread across a wider area. Predictably, this led to the overrepresentation of the main Shiite party (whose support is highly concentrated in the south) and to the underrepresentation of the Kurdish parties (because Kurdish voters live in almost every governorate). Unlike Hamas, the Shiite Islamist party in Iraq earned a disproportionate share of seats not due to its organization, but rather because of its geographically concentrated support.

Table 2 shows the popular vote totals and the number of seats won by the various parties. It also shows three counterfactual scenarios: one in which the results are recalculated using the single constituency used in the January 2005 election, another using a genuine compensatory seat system and a third in which all 275 seats are apportioned to the governorate constituencies with no national compensatory seats. As this table demonstrates, the complex system of compensatory and national seats adopted in Iraq achieved very little. Under the actual system, the Kurdish parties won roughly nine fewer seats than they would have won in a fully proportional system; the Shiites won about eight more. If the seats had been fully apportioned to the constituencies, these disproportionalities are only marginally changed. The very smallest parties—a main concern of the Iraqi decision-makers—are also not helped by the system adopted. In each case, the smallest parties would have won as many seats or more under a real compensatory seat system or even under a system with no national compensatory seats.

The main disproportionality—a shift of roughly 10 seats from the primary Kurdish party to the primary Shiite party—has had few political consequences thus far. The Shiite and Kurdish parties have formed the center of Iraq’s governing coalition since its return to sovereignty, and so the Kurdish party is unlikely to challenge the results. However, over the longer term, the disproportionality of these results may have more pernicious effects. Kurdish leaders have long lobbied to allow Kurds to register as voters in the majority Kurdish governorates regardless of where they live in Iraq. Such a system would legally sanction the already serious fissures between Iraq’s ethnic groups. A compensatory seat system has been suggested by international experts as a way to ensure proportional representation for the widely dispersed Kurds without an ethnic voter registration. Now that the current pseudo-compensatory seat system has failed to deliver positive results, it may be harder to dissuade the Kurdish parties from pursuing ethnic registration.

Why was this pseudo-compensatory seat system adopted in lieu of a real compensatory seat system? This question is difficult to answer in full, because international advisors were not present during the TNA’s final deliberations over the electoral law. The problem, though, seems to lie not with any clear policy rationales on the part of Iraqi decision-makers, but with disorganization among the international advisory team. Despite the clear Iraqi desire for some kind of geographical constituencies to replace the single national constituency used in the January 2005 elections, some leaders of the international advisory team were absolutely opposed to change or even to full discussion of the options with Iraqi decision-makers. As a result, the Iraqis received incomplete technical advice. Technical confusion is clear in the text of the electoral law, which is ambiguous or unclear in many details—a dangerous situation, especially if results might be contested.

In drafting the complex language necessary for Iraq’s system of representation, it was this incomplete technical advice rather than clear policy objectives that guided the law and so the electoral process. The system of representation used in the Palestinian Authority led to a more disproportional result and has politically more dramatic consequences, but it was at least adopted to achieve an understandable set of policy objectives. The Iraqi system of representation is more objectionable because of the poor process leading to its adoption.

Conclusions
Under different systems, and for very different reasons, Islamist parties have a greater share of assembly seats than of the popular vote in both Iraq and the Palestinian Authority. The lesson drawn from these two experiences is one learned too often: participants in the electoral process must take the time to understand the ramifications of systems of representation when the electoral legal framework is under debate. More specifically, the Iraq case highlights the danger of attempting to influence the adoption of a system of representation by limiting decision-makers’ access to information. With limited information, decision-makers may be guided toward the advisors’ preferred outcome or, as was the case in Iraq, they may simply keep their own counsel, adopting a system with serious deficiencies and few policy rationales.

Jarrett Blanc is a post-conflict transition and governance specialist. He is IFES’ chief of party in the West Bank and Gaza.