Vector Space Modeling for Aggregate and Industry Sectors in Kuwait

Trend growth in total factor productivity (TFP) is unobserved; it evolves continually over time. That assumption is inherent in the use of the Hodrick-Prescott or other bandpass filters to extract trends. Similarly, the Kalman filter/unobserved-components approach assumes that changes in the trend growth rate are normally distributed. In fact, some changes to the trend growth rate of total factor productivity are normally distributed but some may not be, especially in the presence of random technology shocks. The latter distribution may be asymmetric with large outliers. Allowing for those outliers, the estimated trend growth rate changes only infrequently. A strength of the paper is that aggregate trends in TFP are presented, but also industry disaggregated trends estimated by the same techniques.