Prediction Models for High Versus Less Performant Economies in the European Union

Abstract

In the current unstable economic environment the European Union countries seem to be facing real challenges that distinctly affect their economic performances. Although there are several attempts in the international literature in building efficient macroeconomic prediction models, the subject still remains of great relevance and it is mostly believed that automated correction in any decision process should be based on proper prediction models. Therefore, we draw on the main macroeconomic performance indicators, such as economic growth, current account balance and labour market indicators, such as labour productivity, employment and average net earnings for the year 2013 to propose several prediction models for the European Union countries. Thus, by applying both econometric analysis and classification trees methodology we will attempt to extend the empirical research in the field. A distinction between high performant and less performant European Union economies will be highlighted, and several CHAID classification trees will be elaborated, followed by a sensitivity analysis. The main findings of the study will consist of several efficient prediction models for which the prediction ability will be tested and compared.