As an interesting area of academic research, election forecasting has developed significantly during the last three decades. Although this interest created a highly illuminating literature, it is heavily concentrated on the elections in established democratic countries like USA, UK and France. For certain other democracies, little or no election forecasting has been carried out. Trying to fulfil this gap present work focuses on forecasting elections in Turkey. Accordingly the aim of this work is to develop and test a new forecasting model for general elections in Turkey. While doing so, this study will base on three theoretical premises: first, the voters reward or punish parties according to their performances related to the macroeconomic conditions; second, the popularity of the political parties in Turkey are highly connected by their performance in local elections, and third, the centre-periphery distinction affects the fortune of the political parties in Turkey. The contribution of this analysis is the introduction of an explicit model, which can forecast the impact of economic and political variables across the elections in Turkey by using reliable, public and macro level data.
