Estimating Potential Output and the Output Gap for the Euro Area: a Model-Based Production Function Approach

This paper evaluates unobserved components models based production function approach (PFA) for estimating the output gap and potential output for the Euro Area.Our main purpose is that of implementing in a consistent model based framework one of the most popular approaches to estimating those key macroeconomic latent variables, thereby avoiding the use of ad hoc signal extraction filters.We fit and validate, against a bivariate model of output and inflation, a system of five time series equations for the Solow's residual, labour force participation, the employment rate, capacity utilisation and the consumer price index; the first four equations are used to define the output gap, whereas the price equation relates the latter to underlying inflation, according to a triangle model.Several hypothesis of interest are entertained: the common cycle hypothesis, with capacity utilisation as the driving force, the hysteresis hypothesis, and we propose a model with pseudo-integrated cycles that is quite effective in eliciting cyclical information from the labour market variables, and enhances smoothness in potential output growth estimates.A rolling forecasts experiment is used to assess the out of sample predictive accuracy of the alternative models.The conclusion is that, although the PFA models cannot outperform a bivariate model of output and inflation, they can be valuable for growth accounting and for reducing the uncertainty surrounding the output gap estimates.We end with a discussion about the use of unobserved components methods to obtain a thorough assessment of the reliability of the output gap estimates.