Estonia’s parliamentary elections were held on 4 March. The following discussion of the electoral system is transplanted here from something I originally wrote Sunday. Below that is entirely new text on the results.
The Electoral System
As I noted a few days ago, Estonia was, for one election, one of the few countries (the only one?) outside of the U.K. and its former colonies to have used STV. While that system turned out to be too unfavorable to party leadership for the tastes of–who else!–party leaders, the system adopted and in use ever since does have a stronger personal element than many party-list systems in Europe. That is, the current Estonian system is broadly in the family of ‘flexible list’ but the actual flexibility is much greater in practice than in others of that category (e.g. Austria, Belgium, the Czech Republic, and the Netherlands).
In Estonia, the voter must cast a preference vote for one candidate on a party list. Seats are allocated to parties according to their share of the national vote (with a 5% threshold). Seats are allocated to candidates as follows. Candidates can win outright on purely their own preference votes, but seats in the various multi-seat districts that are not filled by one of two stages where preference votes are taken into account* are filled instead in a third stage at the national level, based on their order on the party’s closed national list.
I believe the number of seats filled at the national rather than district level (and therefore by party rank rather than by preference votes) has grown over time, in part because parties are free to nominate many more candidates than there are seats in a district, and the more that parties do so–and nominate candidates with a personal following–the lower the number of candidates that will tend to have enough votes to be elected based on their preference votes at the district level. (See the comment below by Taavi Annus for new information here.)
Clearly, it takes more than just a moment to attempt to explain the Estonian electoral system (and, actually, there is much more that could be said). I should note that Estonia’s most famous expert on electoral systems, my graduate-school mentor, Rein Taagepera, did not design this system!
Results
Preliminary results show that the ruling coalition performed well in Sunday’s election. Reform won 31 seats and Kesk 29, which is a large majority of the 101-seat Riigikogu. The results represent a gain of 12 seats for Reform and one for Kesk.
The largest opposition party remains the Pro Patria and Res Publica Union, despite its loss of 16 seats. It now will have 19. (Taagepera was involved in founding Res Publica and published an academic article about its “meteoric rise.” He is no longer involved with the party, and its showing in this election has to count as a recovery considering the party’s post-election meteoric fall that Rein’s article also addresses.)
Add Estonia to the ranks of countries (all of which use PR) to have Green representation: The Rohelised party won 6 seats. (This Green party is new. However; the green movement has deep roots in the social movements of the late Soviet period. This new party marks the first representation by a Green list since 1992, before the threshold was raised to 5%. In 1999 green candidates ran on joint lists with Kesk, but none had enough preference votes to win. Thanks to Taavi for corrections here.)
Other bloggers’ commentary
Josep Colomer has a very interesting post about Estonian democracy and Taagepera’s role in its development since the country recovered its independence from the Russian-Soviet empire.
(Josep mentions having visited Estonia in October, 1991. I guess we just missed each other, as I was there at the end of that very month. Interesting experience–maybe I will tell some time.)
Another of Taagepera’s former students, Steven Taylor, notes that the election will have a considerable element of on-line voting. From the country where Skype originated, this election is apparently the first widespread use of Internet voting technology anywhere.
Alex Speaks about the election results.
_____
* In one of two ways: Either by having personally a quota’s worth of votes, where the quota is the number of valid votes divided by the district magnitude (the ‘simple’ or ‘Hare’ quota), or by having preference votes equivalent to at least .1q, where q refers to the quota.



Speaking of Estonia and Taagepera…
Matthew Shugart (who wrote the electoral studies classics Seats and Votes with Taagepera) points to a post by Josep Colomer about Rein Taagepera’s contributions to Estonian democracy.
Scion grafted by PoliBlog ™: A Rough Draft of my Thoughts — 04 March 2007 @ 22:50
Thank you for paying attention to Estonia.
You speculate that “the number of seats filled at the national rather than district level (and therefore by party rank rather than by preference votes) has grown over time.” This is actually not true – the share of “national” seats has invariably gone down in each election since 1992 (1992 – 60, 1995 – 52, 1999 – 46, 2003 – 27, 2007 – 26). One of the reasons why your guess was inaccurate is that the number of candidates each party can nominate is restricted (this was not the case for a while, but is now – district magnitude +2 is the current limit). The second reason is that a party gets a district seat also if it gets 75% of the “simple quota” (or gets an additional district seat if its district level remainder is 75% of a simple quota). This was not the case in early elections.
The Greens that cleared the threshold this time is actually a new party – it was officially registered in November 2006. The previous Green party disappeared in 1998, after it merged into Keskerakond. It is true that green movement as such has strong roots, and that its role in the independence movement was great.
Seed planted by Taavi Annus — 10 March 2007 @ 15:48