Opinion: Why the tax treatment of savings is a mess that needs urgent attention before fixing Working for Families

By Michael Littlewood*

A new Official’s Issues Paper has just been released by the Inland Revenue and the Treasury (It's available here).

The paper suggests “...changes to improve the fairness of access to social assistance programmes”. It was foreshadowed in the 2010 Budget.

The Issues Paper proposes to extend the definition of ‘income’ when someone may be entitled to an income-tested benefit such as Working For Families, Student Allowances etc.

Income that is not directly earned by the applicant (or the household) will now be included.

There is no doubt that the way income is defined needs changing because it is so easy to ‘hide’ income to maximise entitlements to income-tested benefits.

However, the problems don’t just apply to “social assistance programmes”. They apply to the calculation of income tax as well.

It’s just as easy to avoid tax by splitting assets and income across different tax vehicles. It’s even possible to ‘shelter’ employment income from full tax.

What we have at present with particular regard to different saving vehicles and instruments is actually a bit of a mess and is the result of decades of patches, compromises and the creation of new saving vehicles (most recently PIEs) with their own tax treatments.

An Retirement Policy and Research Centre (RPRC) Working Paper by Michael Chamberlain (of MCA NZ Limited) and me (available here) analysed all the definitions of ‘income’ that now apply to the taxation of investments.

Taking the example of a single Australian-listed share, the paper showed 11 different ways that single share could be owned either directly or through combinations of unit trusts, registered superannuation schemes and Portfolio Investment Entities (PIEs).

There are seven different tax consequences of those choices.

For a single overseas bond, there are 13 different ownership possibilities with, potentially, nine different tax consequences – that’s nine different ways of calculating tax on what is, ultimately, a single investment asset.

So the tax treatment of savings is a mess that needs urgent attention.

As the Working Paper pointed out, the problems created by the ‘silo’ approach to calculating ‘income’ also have consequences for income-tested benefits under the welfare system and the Issues Paper now acknowledges that.

However, the Issues Paper is not really about “improv[ing] fairness of access to social assistance programmes”. It is in fact about fixing the mess created by all the different definitions of income used in the various tax silos as well as their disconnections from what individuals actually pay income tax on.

Here is what the RPRC Working Paper said about how the now acknowledged mess (what the paper describes as a need for “greater integrity”) should be fixed:

“This paper recommends a broad framework to replace current arrangements that does not require the invention of artificial definitions of income. Instead, it attempts to recognise the true economic nature of the transactions involved. Adopting a principles-based framework will also make the interaction between ‘income’ and income-tested payments and levies of all kinds by the state more coherent and fairer. The [paper’s] recommended Collective Investment Vehicles (CIV) tax regime requires that investor/members are taxed on the basis that they had earned the income directly. A practical foundation that will see the income of investor/members calculated in ways that will be familiar to taxpayers is suggested.”

Instead, the Issues Paper suggests that the gaps between the different tax silos should be bridged with a new raft of artificial concepts.

In summary, it proposes widening significantly the current definition of ‘Family Scheme Income’ (FSI) to include all kinds of indirectly earned income.

Yet more complexity

Inevitably, there will be another web of complexities, gaps, administration and disconnects.

For example, the Issues Paper proposes that investment income earned through KiwiSaver will not count in welfare income tests whereas income earned through what the paper describes as an ‘unlocked PIE’ will count. It is difficult to see how that distinction can be justified.

Why doesn’t all income, however and wherever earned count?

FSI is not a new concept and already includes amounts received that are not usually taxable. Things like child support, some overseas pensions and other distributions are now part of FSI.

Also, FSI cannot be reduced by some types of business losses.

The Issues Paper proposes that life is about to get much more complicated than now: FSI will now also include “trustee income” (TI), “attributable fringe benefits” (AFB), “passive income of children” (PIC), “unlocked PIE income” (uPIEi), “Non-resident spouse income” (NrSI), “exempt income” (EI), “Main income equalisation scheme deposits” (MIESD) and “periodic payments” (PP).

So, before someone can work out what amount of income-tested benefit they might be entitled to, someone (most certainly this will be beyond the individual’s capacity) must total:

Taxable income + (TI + AFB + PIC + uPIEi + NrSI + EI + MIESD + PP + the other adjustments already made) ...in order to see how much their FSI is.

Only then can someone else (inevitably that will have to be a different person) work out what welfare the individual is entitled to.

There is a better way to do this and it isn’t conceptually difficult.

We first need to fix the calculation of income tax so that New Zealanders pay the appropriate amount of tax on all their income, including investment and employment income earned indirectly. That is actually more urgent than “improv[ing] the fairness of access to social assistance programmes”.

Fixing income tax will automatically attend to the current problems with income-tested welfare benefits. The Issue Paper’s objective must be attended to but it is actually a second-order problem.

If all ‘income’ were appropriately taxed then the Inland Revenue would automatically know whether the taxpayer qualifies for an income-tested benefit; and the income tax system itself will be fairer.

Having a fairer tax system is actually more important than “improv[ing] the fairness of access to social assistance programmes”.

Both need fixing but the tax problem is more important than the benefits one.

* Michael Littlewood is the Co-director of the Retirement Policy and Research Centre (RPRC) at the University of Auckland