Forecasting short social time series with Theta method: an application to worldwide human development.

Abstract Forecasting short time series is essential in many economic and social contexts. A notable empirical example is human development, measured annually since 1990. However, forecasting these series presents significant challenges due to limited data, making sophisticated statistical methods less appropriate than simpler models. In this paper, we evaluate the Theta method as a forecasting tool for short time series. We benchmark the performance of the Theta method against established approaches and apply the framework to forecast global human development indicators from 1990 to 2022