Estimating Tax Buoyancy: Using Empirical Bayesian Technique
Author: Ridda Bibi

Tax buoyancy is an instrument used by fiscal policy makers in order to know the responsiveness of tax revenue to change in GDP, without adjusting for discretionary measures, and to design future polices. This study investigate the short run and long run buoyancy of different components of tax revenue such as PIT, CIT, SSCT and TGS, for 28 advanced, 17 emerging and 13 low-income countries over the period 2000-2016 by using empirical Bayesian techniques. Along with empirical Bayesian D- prior and G prior estimators, we have also used Engle Granger co-integration, for individual’s country, and Kao panel co-integration test for testing long run relationship. The study reveals that long run relationship happens between tax revenues components and its determinants both for all individual’s country, except Kenya, and all panels. The posterior results of empirical Bayesian D-prior, with lowest root mean square forecast error, showed that in long run CIT is buoyant for all of the developed counties and PIT and SSCT for most of them, except few, while in short run PIT and CIT and SSCT show buoyancy. The study also reveals that for emerging economies CIT, SSCT and PIT are buoyant in long run while in short run TGS and SSCT are not buoyant. Mostly for low income countries PIT and TGS are buoyant in short run and CIT, SSCT, PIT and TGS in long run. CIT is buoyant for few of the low income countries. Moreover, we also pointed out that for most of the countries, in advance, emerging and less developed nations, inflation and output volatility have negative impact on tax buoyance. In last, this study indicate that empirical Bayesian D-prior is the more efficient Bayesian technique for estimation. Supervisor:- Dr. Hafsa Hina

Meta Data

Keywords : Co-integration, Empirical Bayes D-prior and G-prior, Tax buoyancy, Unit roots
Supervisor: Hafsa Hina

Related Thesis​