We establish the causality between government revenue and government expenditure using Granger causality tests within the Vector Auto-Regressive (VAR) framework. The estimated VAR model includes gross domestic product, exchange rate and Treasury Bill rates. Granger causality tests found unidirectional causality running from government expenditure to revenue. This work is founded in economic theory of public choice and the underlying causality of budget deficits. Knowledge of the revenue spending nexus will shed more light on the nature of the intertemporal relationship between government revenue and government spending and help shape the political economy of fiscal policies.

The results are augmented by forecast error variance decompositions which suggest that a one standard deviation shock to government expenditure will only explain about 15% variation in government revenue. The absence of social crises may explain the weak relationship indicated by forecast error variance. The spend and- tax hypothesis is based on the premise of large-scale social disturbance, crises, or war, which make it easier for a political system to decide how much to spend and then adjust tax revenues. In the context of Zambia, these results may indicate that the political system makes spending decisions in the midst of peace and gets to adjust revenue policy to finance the budget deficit.

Zambia must emphasise policies that control or place limitations on government spending to effectively manage budget deficits. Policymakers, politicians, and civil society will have to prepare themselves and the general public on the importance of bringing the government budget in balance.