I have written about the potential future for indirect presidential elections in Estonia on the pages of this blog several times over the last years. After the minor constitutional crisis in 2016, when both parliament and electoral college failed to elect a president in five rounds of voting, it appeared as though a reform of the system was imminent. Although plans for the introduction of popular election were quickly shelved in favour of a reformed indirect system, no changes were decided until the end of the Riigikogu’s (parliament) legislative period in March. Two weeks ago, a new coalition took office that included the introduction of more direct democracy – including the popular election of the president – in its coalition agreement. Although it is unclear whether there will (ever) be a constitutional majority in favour of popular presidential elections, it is timely to look back at Estonian presidential elections to date and analyse what factors shaped party strategies and electoral outcomes. This blog post summarises some of the key findings from my article “The effects of majority requirements, selectorate composition and uncertainty in indirect presidential elections: The case of Estonia” published in East European Politics.
Since 1996 Estonian presidents have been elected indirectly in either parliament or an electoral college that acts as a (frequently used) failsafe if parliamentary parties cannot muster the required 2/3-majority over three rounds of voting. Nevertheless, the electoral college – consisting of members of parliament and local councilors – does also not stipulate a plurality runoff or other safeguard against cyclical preferences. Furthermore, as the partisan composition and control of the selectorate differs greatly between parliament and electoral college, parties face various uncertainties and cannot always predict if their candidate will be successful. Consequently, the varying majority requirements, selectorate composition, and uncertainty have been the key factors influencing party strategies in Estonian presidential elections.
Only once since 1996 has a president been elected during the first round of voting in parliament (re-election of Toomas Hendrik Ilves in 2011); in 1996, 2001, and 2006 presidents were elected in the respective last round in the electoral college (usually with razor-thin majorities), and in 2016 in a sixth round in parliament after no candidate received the required absolute majority in the college and the election was handed back to the Riigikogu. Throughout this process, it is clear that the requirement of (super)majorities provided a major obstacle for parties. In particular, it minimized incentives to make serious attempts at electing a president in parliament. Rather, parties chose to face the uncertainty of the electoral college. However, in the electoral college local electors who did not belong to any of the parliamentary parties dominated, so that parties had less control over the result. Even those electors belonging to the larger parties often felt torn between voting in line with local or national-level interests (and usually chose the former).
To assess the degree of uncertainty faced by parties in the electoral college through shifts in selectorate composition I gathered data on all 500+ MPs and over 1,000 local council electors that took part in the elections. Interestingly, I found that although overall indicators suggested that shifts in selectorate composition became less dramatic – in part thanks to national parties’ attempts to bring local politics under their control – this proliferated rather than reduced uncertainty over the electoral outcome. The more equal distribution of parties’ strength in parliament and electoral college reduced their capability to present candidates with a credible chance of capturing the presidency – the electoral college merely mirrored the deadlocks and problems of national politics. Paradoxically, the greater permeation of local politics by national parties meant greater power in the hands of electors representing local lists and alliances who subsequently acted in accordance with a different set of preferences and proposed their own candidates.
Although the overall design of the Estonian system may be somewhat unique, it still combines a number of common characteristics that can be found across other parliamentary republics. Unfortunately, there is as of yet very little comparative research on indirect presidential elections (Csaba Nikolenyi’s study is a notable exception here) that would allow us to reliably gauge the wider applicability of the findings from this study or make informed recommendations to policy-makers as they seek to reform the Estonian system. Nevertheless, what appears to be clear it appears likely that if parties fail to instate any type of run-off / simple majority, deadlocks and cyclical preferences are likely to return.
This blog post is based on a recently published article: Köker, Philipp. 2019. The effects of majority requirements, selectorate composition and uncertainty in indirect presidential elections: The case of Estonia. East European Politics [Online First: DOI:10.1080/21599165.2019.1604339]