Abstract:
This paper provides the first estimates of a Quadratic Almost Ideal Demand System (QUAIDS) for New
Zealand and uses this model to investigate the distributional effects of a move to a multi-rate GST system.
The estimated QUAIDS model covers nine non-durable expenditure groups and produces highly plausible
expenditure and price elasticity estimates. Behavioural simulation results show that a multi-rate GST
structure would, on average, benefit poorer households relative to richer households – both in terms of the tax households pay and money-metric welfare. However, around 27% of the poorest decile would lose from the reform due to their particular consumption preferences, while around 19% of the richest decile would gain. Behavioural simulation results also confirm the finding from previous non-behavioural analysis that the distributional impact of reduced GST rates can vary significantly depending on the type of expenditure subject to the reduced rate. Overall, the GST system is found to be a poor mechanism for targeting support to poorer households.
