Making Blank Votes Count: An Agency Theory of Electoral Boycott and Thailand’s 2006 Parliamentary Election∗ Kraiyos Patrawart† December 2009

Abstract This paper develops an agency model to examine the political agency problem when the opposition and voters have options to boycott and to cast blank vote in an election with political uncertainty. This model shows the extend to which electoral boycott and blank vote casting can determine the accountability of politicians and the quality of the government through the democratic election. The main findings show that the degrees of adverse selection and moral hazard problems between politicians and voters depend primarily on incumbent’s reputation and comparative strength between incumbent and opposition. Electoral outcomes in perfect Bayesian equilibrium can vary from perfect electoral accountability and efficient public goods provision to the situation where corrupt politicians face no electoral accountability and inefficient level of public goods is provided. The evidences from Thailand’s unique boycotted parliamentary election in 2006 with 33% of blank votes share strongly support adverse selection and moral hazard arguments. Strong performance of opposition’s candidate in preceding competitive legislative election in 2005 associates with voters’ strong supports from both sides for boycott in terms of blank voting. Losing a quarter of total uncontested seats to blank votes alone had put the legitimacy of TRT party into serious doubts which had forced the alleged corrupt interim Prime Mister to decline the post of Prime Minister after the Parliament had reconvened. JEL classification: D72, D73 D82 and N45 Keywords: Election Boycott, Agency Model, Blank Vote, Political Accountability, Thailand ∗

I am indebted to Cecilia Testa for precious discussions throughout the course of writing this paper. Financial support from Royal Holloway is acknowledged. † Department of Economics, Royal Holloway, University of London, Egham TW20 0EX, U.K.; [email protected]

1