By the Numbers: The Kyrgyz Parliamentary Election and its Quantitative Shortcomings

by Ryan Weber

As widely reported (including on this blog), the October 2010 Parliamentary elections in the Kyrgyz Republic were a tremendous accomplishment and a great step forward for the Central Asian state twice-"liberated" from semi-authoritarianism. The mere logistics of organizing an election in an often difficult-to-access country, with a weak central government, a history of rigged voting, and few experienced administrators were daunting. Doing so in the aftermath of an irregular regime change, under the guidance of a President still looking for legitimacy, and with simmering ethnic tensions following outbreaks of tremendous violence, is nothing to scoff at.

Or, as the opening paragraph of the OSCE/ODIHR preliminary report puts it, "The 10 October parliamentary elections were conducted in a peaceful manner. The authorities displayed the political will to bring the Kyrgyz Republic closer to holding democratic elections in line with OSCE commitments. Political pluralism, a vibrant campaign and confidence in the Central Commission for Elections and Referenda (CEC) characterized these elections. Fundamental freedoms, including the freedoms of expression, assembly and association, were generally respected. Overall, these elections constituted a further consolidation of the democratic process. Nevertheless, there is an urgent need for profound electoral legal reform."

However, handing the Kyrgyz effort a blank check of congratulations ignores some of the less laudable aspects of the process, and obscures the areas in which the government, even the newly-adopted Constitution itself, is working against its own stated objectives. Take, for instance, the question of electoral thresholds.

Nothing puts non-experts into a catatonic stupor like discussing electoral technicalities, but in the case of the 2010 Kyrgyz Parliamentary elections, the numbers clearly demonstrate how one aspect of electoral policy can influence not only the makeup of the resulting government, but the ultimate distribution of power and future trajectory of the entire country. Additionally, electoral thresholds tell us a lot about the intended inclusiveness of the resulting government - that is, how many different voices are allowed to play a role in national decisions. Thresholds are employed in many parliamentary systems as a way to make governing more effective - removing marginal parties and encouraging compromise and coalition building - but have the notorious potential of going too far. When a threshold is set too high, or other factors are used to intentionally disenfranchise specific groups, this is recognized as an abuse of thresholds and contradictory to the productive democracy they ostensibly support.

I contend that the Kyrgyz elections, which employed a 5% threshold of eligible voters and a 0.5% "regional threshold," represent a poor implementation of the principle. Below I present figures demonstrating how different threshold policies would have resulted in different governing results and levels of inclusion for the 120-member Kyrgyz Parliament as it moves forward.

The Kyrgyz Parliamentary election system is a Proportional Representative National List system, meaning that voters select a list of candidates by party in one, nation-wide constituency. Though not accurate, American audiences could think of this as roughly approximate to checking the "All Democrat" or "All Republican" chad in a Presidential election year. Importantly, the Kyrgyz electorate is not broken up into smaller voting districts where results are tabulated, nor do Kyrgyz voters cast ballots for particular candidates. Each party 'lists' its candidates, and then fills the government from those lists with however many seats it is awarded. One odd characteristic of this system is that every party is required to field a minimum number of candidates, though ultimately every party will be awarded fewer seats than it has candidates. This can result in 'celebrity candidates' that are added to a party's list with no intention of giving that person an awarded seat in Parliament.

Once the election is concluded, officials tabulate the results by party. Any party that receives less than the threshold amount is eliminated, and the votes for the supporters of those parties are effectively negated. The remaining parties are awarded seats in the legislature based on the proportion of the valid votes they received. In a 120 seat legislature, a party receiving 25% of the remaining votes would receive 30 seats, etc. The exact mathematical mechanism by which proportionality is assessed can become very complicated, but is relatively straight-forward in the system operative in Kyrgyzstan.

There are 3 elements of the Kyrgyz electoral process that seem troubling, and in the case of the 2010 election had determining effects. They are:

1. The 5% threshold was taken as a percentage of all registered - 'eligbile' - voters in the country (approx. 3 million), not the number of citizens (5.5 million) or the number of people who actually voted (1.6 million). This is quite unusual, and resulted in a dramatically higher "effective threshold" than the stated 5%.
2. The 5% threshold is relatively high among Proportional Representative (PR) systems. Thresholds of 3% are more common, though some do range as high as 7%.
3. The 0.5% "regional threshold" was highly contentious in Kyrgyzstan's last Parliamentary elections in 2007. It requires each seated party to have a 0.5% vote share in each of the country's 7 oblasts (provinces) as well as in the two largest cities, Bishkek and Osh. In 2007, it was actually the lever by which the President's Ak-Jol party entirely excluded the highly-popular Ata-Meken party despite its national 2nd place finish. This electoral rule has continued under the new constitution, though doesn't seem to have played a major role in the 2010 results.*

* Full elections results are not yet widely available, so it is possible that this 0.5% threshold did influence results, or would have in the event of a lower national threshold.

Thesholds on 'eligible' vs. 'actual' voters

It's not hard to understand the rationale behind applying a threshold to the total number of eligible voters in a country, rather than just those who choose to vote in an election. The logic would assume that any party which couldn't motivate a certain % of the total potential electorate to support it - and to vote in such a manner - would not be sufficiently organized to govern effectively. Thus, truly marginal parties - extremists, personality cults, etc. - would be excluded from government even in the event of a very low national turnout.

But as policy, an 'eligible voter' threshold throws out the baby with the bath water. Consider the Kyrgyz results (table at right). Parties included in the new Parliament are in blue, those excluded by the 5% eligible threshold are in red. With an actual turnout of just 55% of all eligible voters, it becomes clear that a 5% 'eligible' threshold turns into a 10% 'actual' threshold in practice. Butun Kyrgyzstan ("United Kyrgyzstan") received more than 5% of the actual votes, but was denied any seats.

Another interesting result is that a total of 532,545 cast votes - 1/3 of those cast - were not represented in the resulting Parliament because they voted for a party that did not cross the 5% eligible threshold. While Majoritarian voting systems - like the US Presidential race - commonly exclude almost 1/2 the participants, these numbers are extremely high for a PR-based election.

Why 5%? Consider the alternatives

Without changing the voter turnout or tabulated results, let us instead assume different thresholds and how they would impact the resulting parliament. First, consider if the 5% threshold was maintained, but applied to actual votes, not eligible voters (table). Doing so would include the Butun Kyrgyzstan party in the new Parliament with 14 seats - almost as many as Ata-Meken. Further, Butun Kyrgyzstan is closely aligned with Ata-Jurt and Ar-Namys, meaning that these three would likely form a coalition with 60 seats, the number required by the Kyrgyz Constitution to name the Prime Minister. Of course, the other three parties would also possess a total of 60 seats, but Respublika is less closely aligned with SDPK and Ata-Meken, and might even join the ruling coalition to gain influence. If this occurred, SDPK and Ata-Meken, which were the dominant opposition parties under the Bakiev regime, would again return to the opposition. Even more shocking, an Ata-Jurt leader - such as Kamchibek Tashiev, a former Ak-Jol member and Bakiev's Minister of Emergencies, who has called for a return to the Presidential system - would be a likely choice for Prime Minister.

Extending this experiment further, the results of progressively smaller thresholds taken on the actual number of votes cast are illuminating (see table). At a 4% threshold, the Ak-Shumkar (White Falcon) party is included, and receives 7 seats. At a 3% threshold, Zamandash party joins the Parliament, with 5. No further parties would be included at the 2% threshold (not shown), and a 1% threshold essentially removes the purpose of having one.

Of course, all these seats don't come out of thin air - they are subtracted from the other included parties. Ata-Jurt, SDPK, Ar-Namys and Respublika all lose 5 representatives over the range of these examples (Ata-Meken loses 4). At the far end (3%), the 3-party nationalist coalition of Ata-Jurt, Ar-Namys and Butun Kyrgyzstan no longer has the 60 seats necessary to name the PM (only 55), but neither does a mega-coalition of SDPK, Ata-Meken, Ak-Shumkar and Respublika (50). Presumably a compromise cabinet would be formed - though the possibility of legislative deadlock, as in Iraq, is always a possibility.

Regional Thresholds

The issue that was the most controversial aspect of the 2007 Parliamentary elections (other than the emergence of the pro-Presidential Ak-Jol party and the rampant voter fraud) has been retained in the new Kyrgyz Constitution for reasons that are difficult to fathom. Originally installed under the guise of preventing one region to politically dominate the others, it was always seen as a tool for Southern-backed parties, like Ak-Jol, which had migrant support in the North, to deflate Northern parties, which didn't benefit from the same urbanization-driven regional movement. While many with familial ties to Southern Kyrgyzstan moved to Bishkek following the 2001 Soviet collapse, and even more after the 2005 "Tulip Revolution" swept Southerner Bakiev into power, the inverse has never been the case.

Notably, the ODIHR preliminary report states, "The double threshold requirement undermines the objective of a proportional representation system."

It is interesting - and worthy of more analysis when full figures are available - that the 0.5% regional threshold did not play a major role in the formation of the 2010 Parliament, but this is likely due to the exclusion of so many smaller parties. Major factions like SDPK and Ata-Meken were able to pass the low 0.5% bar, but I suspect Ak-Shumkar, for example, may have been disqualified by this condition even if the national threshold were reduced to 3%. That is pure speculation, and I look forward to having access to data to prove or disprove it.

Macro-Analysis: The effect of thresholds on Electoral Inclusiveness.

The goal of any honest democracy is two-fold: First, to produce a government capable of governing, and second to include and represent as many of its citizens in the composition of that government as possible. The relative priority is not always ordered in this fashion, but all democratic election systems strive to achieve some combination of these two. So how did the Kyrgyz system fare?

The ability of the resulting Parliament to govern is yet to be seen, but let's assume, based on historic and contemporary precedent, that with only 5 parties and no clear majority, they will be able to work something out - or at least, no less able than they would with 6, 7 or 8 parties. The eventual success or failure of the 2010 Parliament is far from certain, but in the scheme of 'ability to govern,' the common principle is that it is easier to get things done with fewer parties. By this rubric, the 5% threshold on eligible voters - as stipulated in the Constitution - is the most effective, as it produced the fewest number of parties in Parliament.

But what about inclusivity? It's not just a matter of only 5 out of 29 parties gaining seats - many of those parties were very small, even familial, and had little chance - or business - of influencing Kyrgyz politics. Look instead at the percentage of the voters and the electorate that was involved in choosing its eventual government (see table).

What is immediately apparent is that the % of eligible voters represented in the Parliament is not very high - from 37% to a maximum of 47% - mostly as a result of the 55% turnout. Eligible voters who do not vote have no chance of being politically represented in any legislature. Look instead at the dramatic shift in the inclusion of actual voters, which begins at only 67% with the official 5% eligible threshold, but which could include 80% of all voters with a 4% actual threshold, and even 85% at the 3% level.

This tells us what we already knew - that a lower threshold would produce a Parliament with a larger number of parties that represented a larger % of the politically active citizenry. The question is whether the additional burden of more parties outweighed the advantages of having a government in which more citizens had an active stake. Even at the 3% threshold, the total percentage of the population of Kyrgyzstan represented would only be 25% (up from 20% in the official results).

The Kyrgyz government has made its determination, and facing the many challenges that lie ahead, their reluctance to foster a more vibrant multi-party legislature is understandable. But is it the right path for a country racked in recent years by increasingly aloof political leaders and a populace so divorced from political power that it takes control through disorganized riots? If anything, the people of Kyrgyzstan need to take ownership of their new polity, and excluding the representatives of those that actually do make the effort to vote is hardly a step in the right direction.

These are only preliminary findings, and I invite you to access - and improve - my research and analysis (download). Any new, or contradictory, findings and opinions are welcome in the comments or via e-mail.

The relevant legal documents setting the Proportional Representative National List and 5% / 0.5% Threshold based on registered/ eligible voters can be found at the following locations:

• Section IV Article 70of the Kyrgyz Constitution (adopted by referendum on 27 June 2010), "The procedure of electing the deputies to the Jogorku Kenesh including the establishment of an electoral threshold for passing to the Parliament, shall be defined in the constitutional law."


• Article 77 of the Election Code of the Kyrgyz Republic states, "Political parties, whose lists of candidates received less than five percent of votes of voters, included into voter lists, in the whole republic OR less than 0.5% of votes of voters, included into voter lists, in each oblast, cities of Bishkek and Osh, shall be excluded from the distribution of deputy’s mandates in a single-member electoral district."