Low levels of voter turnout in the first election to the National Assembly for Wales in May 1999 brought into question both the ability of devolution to revitalize representative democracy and the legitimacy of the Assembly itself. But drawing wider implications from turnout requires that we understand why electoral abstention was so widespread. We examine three hypotheses about voter turnout in 1999: that non-participation simply reflected a general apathy towards politics; that it was based on a specific apathy towards the new Assembly; or that low voter turnout reflected antipathy towards an unwanted political institution. We find support for the first two hypotheses, but little evidence for the third. Devolution has failed to engage the interest and support of many in Wales, but low turnout has not been prompted by fundamental antagonism to the devolved institution among the Welsh electorate.

This paper examines voting behaviour in the inaugural election to the National Assembly for Wales (NAW), held in May 1999. We address two questions: (i) why did the election produce a ‘quiet earthquake’ in Welsh electoral politics, with the nationalist Plaid Cymru denying the Labour party their expected majority in the Assembly?; and (ii) what broader lessons does this case-study offer for the study of elections in the UK under devolution? Drawing on data from the Welsh National Assembly Election Study, we find that while some features of second-order election theories, such as lower turnout and a lower vote share for the governing party were manifest, contrary to the predictions of such theories the surge in electoral support for Plaid was largely prompted by Welsh-specific factors rather than UK-wide ones. The findings are argued to indicate limits to the applicability of second-order approaches to the study of devolved elections in the UK.