Revenue Buoyancy Analysis: Tax Collection Efficiency across Economic Cycles in Nepal

Abstract

This paper estimates the long-run buoyancy of Nepal’s tax system and tests whether revenue responds symmetrically to expansions and contractions in economic activity. The analysis uses annual data from 1990 to 2023 within a five-variable vector error correction framework that includes total tax revenue, nominal GDP, imports, the consumer price index, and a measure of financial depth. All five series are integrated of order one, and the Johansen procedure identifies one cointegrating vector. The estimated long-run buoyancy coefficient is 1.07, indicating that tax revenue grows slightly faster than GDP in equilibrium. Imports contribute a meaningful additional effect of 0.24, reflecting the importance of customs duty and import-stage VAT in the Nepali revenue structure. An asymmetric specification distinguishing between positive and negative GDP changes finds an expansion buoyancy of 1.18 and a contraction buoyancy of 0.84, with a Wald test rejecting symmetry at the one percent level. Sub-period analysis shows that buoyancy rose markedly after the introduction of VAT in 1997 but fell sharply during the 2020–2023 period of COVID disruption and balance of payments stress. The error correction term in the tax revenue equation is -0.42 and significant, indicating that about 42 percent of any deviation from the long-run revenue path is corrected within a year. The findings have direct implications for fiscal sustainability, the design of countercyclical revenue policy, and the management of revenue volatility in a small open economy. Building on these estimates, the paper projects the forward path of collection efficiency under alternative growth and reform scenarios and links the observed asymmetry to the behaviour of taxpayers and the revenue administration, whose choices ultimately determine how much of the statutory base is realised.