Search

When examining citizens’ behaviour in elections to the European Parliament (EP), second order elections (SOE) and economic voting (EV) have overlooked each other as alien traditions. However, they should not be seen as incompatible: voters may use EP elections to support national incumbents but, at the same time, this support may be the result of an economic assessment. This paper blends both frameworks and finds that, even though EP elections are not intended to evaluate domestic politics, voters use them to reward or punish national incumbents (i.e. SOE) for the state of the economy (i.e. EV); a behaviour referred to here as ‘second-order economic voting’. The predominance of second-order EV is relevant for EU politics. If EP elections enhance national incumbents’ accountability while the European executive remains unacknowledged, this could be disturbing for European economic institutions.

In this article, we analyse patterns of turnout and electoral choices of Cape Verde’s citizens in different types of elections, looking at all legislative, presidential and local elections held between 1991 and 2016, and testing four hypotheses derived from the second-order election model about differences in terms of turnout, number of spoiled/blank papers, results for the incumbent party and the electoral success of smaller parties. Our results show that, in what regards turnout and electoral behaviour, local elections present the features of second-order elections much more clearly than the presidential elections in this semi-presidential regime. However, this pattern does not necessarily mean that voters look at the latter as less second-order, but since they often took place in the honeymoon period of the legislative electoral cycle, it may only mean that there were lower incentives to punish the incumbent and/or disengage from political participation.