Linear Autoregressive Distribution Lags (ARDL) Versus A Non- Linear Autoregressive Distribution Lags (NARDL) Analysis of Tax Revenue-Growth Nexus in Nigeria

Abstract

This study examines the impact of tax revenue on economic growth in Nigeria. Annual time series data from 1985 to 2022 collected from various sources were used. The results of the short-run ARDL Bounds Test showed that PPT, PPT(-1) and VAT(-1) are positively and statistically significant with economic growth proxied by RGDP; VAT(-2) is negatively and statistically significant with it while CIT, CED and VAT are insignificant. The results of the long-run ARDL Bounds Test showed that while PPT(-1) and CED(-1) are positively and statistically significant with RGDP; CIT(-1) and VAT(-1) are insignificant. The results of the short-run NARDL model where the variables are divided into positive(POS) and negative(NEG) changes show that ?(PPT_POS), ?(CED_POS), ?(PPT_NEG(-2)), ?(CED_POS(-2)), ?(VAT_POS(-1)) and ?(VAT_NEG) are positively and statistically significant with economic growth proxied by RGDP; ?(PPT_NEG), ?(CED_POS(-1)) and ?(CED_NEG) are negatively and statistically significant with it while ?(PPT_POS(-1)), ?(PPT_POS(-2)), ?(CIT_POS), ?(CIT_NEG), ?(CIT_NEG(-1)), ?(CED_NEG(-1)), ?(CED_NEG(-2)) and ?(VAT_POS) are insignificant with RGDP. The results of the long-run NARDL model reveal that PPT_POS(-1), PPT_NEG(-1) and VAT_NEG are positively and statistically significant with RGDP while CIT_POS, CIT_NEG(-1), CED_POS(-1), CED_NEG(-1) and VAT_POS(-1) are insignificant with it.

Keywords

References