Forecasting Tax Revenue and its Volatility in Tanzania

Forecasting tax revenue and its predictability is important for government budgeting and tax administration purposes. This study used monthly tax revenue data for a period of 182 months spanning January 2000 to February 2015. The study applied ARMA and combined forecast models, and GARCH models to forecast tax revenue and its volatility, respectively. Tax revenue was found to increase steady over the period, although with a persistent volatility which increases over time. The observed volatility was found to be associated with taxes from bases (income) which have high volatility. Based on various forecast accuracy evaluation criteria, the study recommends combined forecasts and GARCH(1,1) models for forecasting monthly revenue and its volatility, respectively. The study further recommends enhanced diversity of taxes through widening consumption tax base within the existing tax portfolio so as to enhance its contribution to revenue collection and reduce volatility.


Issue Date:
Jan 01 2017
Publication Type:
Journal Article
ISSN:
1821-8148
Language:
English
Published in:
African Journal of Economic Review, Volume 05, Number 1
Page range:
84-109




 Record created 2017-10-30, last modified 2017-10-30

Fulltext:
Download fulltext
PDF

Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)