Promises, lies and the accountability trap. Evidence from a survey experiment in Armenia and Georgia

Lie a little or promise a lot is a question many politicians face when campaigning before elections. This paper examines whether voters support ambiguous pre-election promises in Armenia and Georgia using an experimental design and if so, what it tells us about accountability mechanisms and a potential accountability trap. The accountability trap emerges when voters cannot hold their elected officials accountable for their promises due to their ambiguity and become disillusioned with political participation. This paper looks at how voters’ expected political behaviour changes in response to randomly assigned types of electoral promises from a hypothetical party. I find a positive effect of ambiguity: if a party makes an ambiguous promise, it will do significantly better in Georgia and at least not worse in Armenia than a party promising a specific policy option. The effect of ambiguity partially explains why parties have been poor at putting forward coherent electoral programmes in Armenia and Georgia. More broadly, my findings contribute to understanding the problem of accountability in hybrid regimes, which may lead to representation crises.