Income Tax, Household Consumption in South Africa: A QARDL Approach

Abstract

This study empirically investigates the effect of Pay As You Earn (PAYE) tax on household consumption in South Africa by employing the Quantile Autoregressive Distributed Lag (QARDL) model, using annual time series data from 1994 to 2023 obtained from the South African Reserve Bank (SARB). The QARDL analysis reveals that, in the long run, the PAYE tax is statistically insignificant across the lower, middle, and upper quantiles. However, household disposable income is statistically significant and positively associated with household consumption in all quantiles over the long run. In terms of unemployment, the results show that it is statistically insignificant in the lower quantiles (0.10 and 0.25), while it becomes statistically significant in the middle quantiles (0.40, 0.50, and 0.60), as well as in the 0.75 upper quantile, with a negative coefficient. However, in the 0.90 upper quantile, unemployment is statistically insignificant. In the short run, PAYE tax is statistically significant in the lower quantiles (0.10 and 0.25), middle quantiles (0.40, 0.50, and 0.60), and the 0.75 upper quantile. This implies that an increase in PAYE tax leads to a decrease in household consumption. In contrast, at the 0.90 upper quantile, PAYE tax is statistically insignificant. Based on these findings, the study recommends that the government should consider reducing PAYE tax for low- and middle-household consumers to boost short-run household consumption. Additionally, enhancing disposable income and reducing unemployment through targeted policies can support long-term consumption growth.