Formats and related files
Staff and teams are writing in their individual capacity and the views in this paper are not necessarily a “Treasury” view. Please read Disclaimer for research and commentary publications.
This work makes use of Stats NZ’s Integrated Data Infrastructure (IDI), please also read the IDI disclaimer.
Extract from paper#
This note is the second in a two-part series that seeks to improve our understanding of effective average tax rates (EATRs) in New Zealand. EATRs measure the net effect of taxes and transfers as a proportion of a taxpayer’s income. Measuring EATRs can provide insight into some important questions with implications for the fairness and efficiency of the tax and transfer system:
- Who pays tax and who receives transfer payments?
- How much do they pay or receive?
- Where might taxes and transfers distort decisions to save and invest?
This series focuses on the first two questions and what EATRs can tell us about progressivity (defined as having higher tax rates and lower transfer rates for higher levels of income or wealth). Measuring EATRs can also help us to identify tax distortions caused by the inconsistent tax treatment of different types of income. Such distortions can create allocative inefficiencies and reduce productivity.
This work stems from a recommendation by the Tax Working Group Secretariat (2018, recommendation 69) that ‘strongly encouraged’ the Government to release more statistical and aggregated information about the tax system. A lack of data on the distribution of capital gains prevented the Tax Working Group from being able to provide precise impact analysis of extending the taxation of capital gains. Closing this knowledge gap is particularly important because capital gains can be a significant source of economic income.
Part 1 in this series presented a method to estimate EATRs that can measure multiple tax and transfer types against broad definitions of income (Ching, 2023). We presented aggregated data that demonstrated that annual accrued capital gains, while volatile, have come close to matching personal taxable income in recent years.
Additionally, scenario analysis showed how taxpayers with equal taxable income can have very different EATRs when factoring in transfer payments and untaxed sources of economic income, such as capital gains and imputed rents (ie, the benefit of home ownership or notional rents that owner-occupiers pay to themselves to live in their own house).
This note presents the results from our prototype extension to the Tax and Welfare Analysis (TAWA) microsimulation model that estimates these more comprehensive EATRs across the income and wealth distributions. The intent is for these results to measure the progressivity of the tax and transfer system when we consider a more comprehensive definition of income. We describe our experimental EATR measures as ‘more comprehensive’ because data limitations prevent us from including all taxes and income sources.
We do not comment on the implications of these results for the fairness of the tax and transfer system and leave these value judgements to individual readers. Governments face trade-offs when setting tax policy which might justify a variation in tax treatment for different types of income. Decisions about whether to tax a particular type of income would require a broader consideration of a government’s objectives and established tax principles (including efficiency, administrative cost, integrity, coherence, and equity).
We find that the tax and transfer system appears progressive when measured narrowly against taxable income. However, as we iteratively broaden the definition of income to include transfers and untaxed forms of economic income, the trend of EATRs is significantly altered across the income and wealth distributions. While the overall trend still appears progressive, we find a dip in average EATRs between the 19th and 20th income ventiles (the top two 5% population groupings) once accrued capital gains are included in the EATR income base.
When we view EATRs across the wealth distribution they appear less progressive, or regressive from the middle of the distribution once capital gains are included in the EATR income base, demonstrating the imperfect correlation between income and wealth. However, we caution strong conclusions from our data because these apparently regressive trends may partially reflect the fact that a greater proportion of tax paid by the wealthy is in the form of company taxes, trustee income and trustee taxes that are not included in our model.
This is due to limitations in the data we have relied upon in the Household Economic Survey (HES) and the Integrated Data Infrastructure (IDI), which currently cannot link individuals to companies or trusts for the purposes of allocating these entity taxes back to the individual. However, we do still count any obligations that would be met with imputation credits as personal income tax.
The results in this note can be compared with those estimated by Inland Revenue’s 2023 High‑Wealth Individuals Research Project (the Inland Revenue Project), which crucially does include company taxes, trustee income, and trustee taxes in their measures. The Inland Revenue Project investigates EATRs from a cohort of New Zealand families identified as having high net worth for New Zealand, who are unlikely to be sampled by HES and, therefore, are unlikely to be included in our EATR estimates.
Together, these two projects will create a more complete picture of EATRs across the income and wealth distributions. Our HES-based modelling will provide EATRs by different income and demographic groups, including the median New Zealand family. By comparison, the Inland Revenue Project will provide insight into EATRs for the wealthiest New Zealand families. We find our most comparable population median EATRs to be consistently higher than those calculated for the high-wealth population in the Inland Revenue Project.
-  We implicitly include most retained earnings, to the extent that they are included as accrued capital gains. However, this may not capture all retained earnings in trusts because the capital gains are estimated off the asset types held in the trust, rather than the trust itself.
-  Where the family owns more than 10% of the company.
-  https://www.ird.govt.nz/hwi-research-project