29 September 2012

Taxing families, or how 1.5 equals 1

Mitt Romney's now infamous comments about how 47% of Americans did or didn't interact with their tax system reminded me of a topic I had intended to cover on this blog - net taxpayers.  That sounds incredibly boring (and I suppose to any sane person it probably should be) but it intrigues me because of what it might be saying about the design of our tax-transfer system.  If you persist to the end of this post you might conclude our system is largely incoherent, or perhaps it's deliberately preferencing some household types, or a heady mix of both.  And you'll find out what I'd like for my birthday.

First up, let's clarify some of the terms and concepts I'll be using, starting with "taxpayer".  There are lots of different kinds of tax in the Australian tax-transfer system but I'm focusing on what is commonly called "income tax".  Taxes like the GST won't feature, so a taxpayer in this context will be someone who pays income tax and/or the associated medicare levy.

As at 1 July this year, the tax free threshold (the amount of income a person can have before becoming liable for income tax) became $18,200 a year.  On the face of it then, a person becomes a taxpayer when their income exceeds $18,200.  However, people on low incomes can attract a tax concession of up to $445 - the low income tax offset (LITO) - which, as the name suggests, offsets the tax otherwise payable.  The effect of the LITO is to increase the income at which income tax liability actually kicks in to $20,543 a year.

This gives our first glimpse of the fact that the amount of private income on which a person begins to pay income tax is not simply the tax threshold, but the result of the interaction of different parts of the system.  A huge part of the system is transfer payments and it's the interaction with those I'll look at next, starting with Newstart allowance.

In the usual case, a single person with no income and who meets the eligibility conditions will be able to get a Newstart allowance (NSA).  It's currently $492.60 a fortnight.  There's also a "clean energy advance" which is part of the household compensation measures brought in with the introduction of carbon pricing.  Together these provide an income of $13,029 a year to a person whose private income is $0.  Importantly, at least for this story, $12,807.60 of this is regarded as assessable income for income tax purposes.  So, our person with $0 private income already has $12807.60 taxable income, which appears to leave them with only an extra $7735 private income before they will have reached that magic $20,543 at which tax becomes payable.

It's not that simple, of course.  NSA is income tested, which means the more private a person has, the less NSA they get.  That means as taxable income rises from one source (eg, earnings), it falls from the other (NSA).  The combined effect of these ups and downs is that the combination of NSA and private income reaches the point at which tax becomes payable when private income is $15,698.

Let's pause for breath and look at how that has panned out so far.  Our single person becomes a taxpayer at an income of $18,200.  But, that's modified by a LITO which actually enables them to have $20,54.  But, such a person usually gets NSA which makes the private income at which tax starts $15,698.  Let's put this on a chart, which will also give us the introduction to the net-taxpayer concept.

Chart 1

(click to enlarge)

This is showing how much transfer payment (in this case NSA plus clean energy advance) a single person gets as private income increases - the blue line.  It also shows how much tax they pay - the red line.  As per the earlier discussion, tax starts at a little under $16,000 and steadily increases with rising private income.  At a private income of $22,773 the two lines cross, and from this point on the person is paying more in tax than they receive from NSA.  That income point is often referred to as the net tax threshold, and it's the point at which the person becomes a net taxpayer.

From this we have three interesting numbers for any given household: the maximum transfer "package" paid at zero private income; the private income from which tax becomes payable; and the private income from which the person becomes a net taxpayer.

As an aside, there's a fourth interesting number that I won't be covering in this post - the private income at which transfer payments cease.  It's interesting because over the income range from which tax starts to be paid up to the eventual cessation of transfers, a person is both receiving transfers and paying tax.  This "churning" is an oft-criticised aspect of tax-transfer interactions.  (If churning interests you, you'd do well to look up Peter Whiteford's articles about it in your favoured search-engine.)

Now we can look at how a couple of other household types look in relation to the three numbers.  Chart 2 shows the situation for a single parent with two children.  (And before you react, the scale in this is completely different to that in Chart 1.)

Chart 2

(click to enlarge)
Our three numbers here are: $28,583 for the total transfer payment package; $14,401 for the income at which tax starts; and $60,019 for the net tax-payer point (or net tax threshold).  These are rather different from the single NSA results, but you'd expect that given this is a three-person household (or would you expect that?  Hold the thought....).

Next is a single income couple without children, both getting NSA (where relevant).

Chart 3

(click to enlarge)

Now the three numbers become: $23,525 for the total transfer payment package; $18,812 for the income at which tax starts; and $35,642 for the net tax threshold.

Although these charts help illustrate the concepts I've been discussing, it's hard to compare the different results across the households in this form.  Given there are really only 3 numbers of interest in each case, a chart that focuses on them allows for more than one household at a time to be shown.  And so to Chart 4, which compares the 3 numbers across 6 different households.

Chart 4

(click to enlarge)

At first glance you might think this doesn't really reveal any underlying rationale in the tax-transfer design.  Don't overlook the fact that not all the households are the same size though.  Perhaps if we look just at the couples without children we'll find there is some consistency - oh wait, there isn't.  Perhaps the couples with children?  Not really.

Perhaps we need to equivalise the results?  Equivalising is an attempt to adjust the numbers to take into account the various household sizes.  For example, we could simply divide by the number of people in each household (a per capita result), but this overlooks the generally agreed finding that there are economies of scale associated with living together (eg, the idea that a couple doesn't need twice as much income as a single person to get the same standard of living), and that young children don't cost as much as adults.

Unfortunately, equivalising is a very murky field of study indeed, and there's no agreed scale.  One that is commonly used is the OECD modified equivalence scale.  The Australian Bureau of Statistics uses it when it compares households (they have a handy explanation of equivalising here, with a link to a more detailed paper should the subject interest you).  The OECD scale says, in effect, that a couple needs 1.5 times the income of a single person to attain the same standard of living as the single person.  It also says that each child under 15 counts as 0.3 of an adult in terms of extra cost.

Armed with our modified OECD equivalence scale we can adjust the chart above to take into account the differences in household size.  If the Australian tax-transfer system is built around the following principles:
  • the assistance available to households with no income should provide those households with the same standard of living taking into account family size;
  • the income at which a household becomes a net taxpayer should take into account the number of people in the household so that liability for tax commences at the same standard of living
then all the results should look the same, after equivalising.   Fingers crossed...

Chart 5

(click to enlarge)

Well, they are closer together than before but I think it would be difficult to say the overall results support the idea that household composition is being considered.  The transfer payment totals are probably the best fit to the principle, but the tax outcomes are the proverbial dog's breakfast.

Actually that's not a surprise.  The OECD scale isn't the one used for setting the relativities between single and partnered rates for the income support payments shown here (NSA).  They are (more or less) used for pension purposes, and have been since the 2009 pension increases.  That's why the single pension went up so much compared to partnered rates - the relativities were reset.

The relativities haven't been reset for the non-pension income support payment types (or parenting payment for singles), and Australia doesn't use the OECD scale for child payments (ours are typically more generous),  In any event, what relativities that did exist between child payments and adult payments have not been maintained due to the practice of making ad-hoc payment changes.

The result is that different household types are hit with tax liabilities (become net taxpayers) at wildly differing income levels.  But this is in some ways deliberate too.  It's been a feature of tax changes in Australia to (increasingly) ignore household composition in favour of an individual assessment model.  Some people prefer this (for example, those who favour the argument that says why should the community subsidise the decision to have children).

So, we have a transfer system that, at zero private income, embraces the concept of equivalisation but haphazardly and inconsistently, and a tax system that wants to be individual.  To me that's a bit like mixing fire and ice - where they meet it's just slushy mess.

Personally (I can say that now given my impending retirement) I'd like to see more attention paid to equivalence for assistance rates up to the point at which the household becomes a net taxpayer.  However, this requires, among other things, an equivalence scale that is accepted for such purposes.

And that's what I'd like for my birthday - an equivalence scale that works.

No comments:

Post a Comment