22 August 2013

4 Parliament comparison - single earner couple with an invalid partner

This is another in my occasional multi-Parliament comparison series.  This time the subject is single income couples where one is unable to work due to a disability – an invalid spouse as the tax law puts it.  I’ve been meaning to do this one for a while because the combination brings up quite a few interesting issues around policy and system design. 

The 4-Parliament comparisons are all about the pictures, so without further ado here’s my usual opener, all 4 Parliaments on one chart.  Messy, but useful for comparisons.

Chart 1

I’ll look at each Parliament separately in more detail, but putting them all together on the same scale makes comparisons easier and puts a bit of context around the relative gains (or losses).  It also shows at a glance that Parliaments 41 and 43 are the outliers with biggish moves in disposable income.  In contrast, the action during Parliaments 40 and 42 is rather more restrained.

Now to each Parliament in more detail, starting with Parliament 40.  Here’s a chart showing the overall change in real (ie, CPI adjusted) disposable income, broken down into the variations in the tax-transfer system components that brought it about.   Note too that this blog is about the tax-transfer system, so this and the other charts are showing changes attributable to the tax-transfer system only.  Things like wages growth and interest rate changes aren’t included.

The change in disposable income is calculated by comparing households with the same real private incomes (ie CPI adjusted, and in July 2013 dollars) at the start and end of the periods.

Chart 2

This household’s combination of disability support pension (partner 1 – shown as P1 DSP) and Newstart allowance (partner 2 – shown as P2 NSA) shows the rather disparate treatment handed out to pensioners compared to those on social security benefits.  At zero private income – where households get maximum rates of income support – DSP increased by almost 3% in real terms, whereas NSA essentially remained unchanged.  This reflects the different mechanisms used to maintain the rates of these payments – DSP is benchmarked to wages growth, whereas NSA is adjusted by reference to increases in the consumer price index (CPI).   The former is supposed to be a proxy for keeping up with community living standards whereas the other is an adjustment for inflation.

At slightly higher private incomes we can see that NSA fell in value, which is due to the fact that the income test for that payment was unchanged in nominal terms (ie, no CPI adjustment) over the Parliamentary term.  For DSP the result was positive all the way – the DSP income test is adjusted with upward movement in the CPI.

Tax cuts at higher incomes are also noticeable, but check the scale before getting excited.  These were small beer overall, particularly compared to subsequent Parliaments.  Again, Chart 1 provides a better indication of the comparative size of these changes.  Overall, if extensive movement from the no-change (0%) line suggests an action-packed term of office, then Parliament 40 was pretty bland.

Parliament 41, on the other hand, was an entirely different beastie.

Chart 3

Tax cuts, tax cuts, tax cuts!  A veritable sea of red!  Remember, these charts are showing changes after taking inflation into account, and tax cuts still show up everywhere.  The cuts in that Parliamentary term were somewhat skewed to those on higher incomes, but in isolation were almost as large in percentage terms for some low income households too.  (By “in isolation” I mean divorced from the effects of interactions with income support payments.  An earlier 4 Parliaments post using single people makes this a bit more obvious.)

The tax cuts at the lower income end in this chart are somewhat subdued as a result of another significant change – the relaxation of the Newstart allowance income test.   This increased the amount of Newstart being paid and, as Newstart allowance is taxable, had the effect of increasing taxable income as well.  In turn a higher taxable income means more tax payable, somewhat offsetting the tax cuts otherwise provided.

The income test relaxation was applied only in respect of income over $142 a fortnight.  The income test free area remained unchanged at $62 a fortnight and consequently declined in value in real terms.

The relaxation of the Newstart income test raised, and still raises, a bit of a policy dilemma.  You’ll notice that increases in Newstart appear at incomes up to around $45,000 a year.  That exceeds full-time wages for many people (it’s around 150% of the Federal Minimum Wage, for example).  If you think of Newstart allowance as being an unemployment benefit, the income test relaxation brings about the odd situation where the gains from relaxing the income test fall on those who, at face value, aren’t actually qualified for payment – they’re not unemployed.

The use of Newstart as an “in-work benefit” has only increased since then, with the relaxation of its income test for single parents perhaps the most recent example.  The tension introduced by using an “unemployment benefit” as a device to deliver payments to full-time workers has never really been addressed, or had a proper public airing.  But that’s probably a topic for a separate post.

I mentioned that the increase in taxable Newstart allowance payments had the effect, in this type of household at least, of reducing the apparent size of the tax cuts.  In a similar vein it also increased the amount of income taken into account for the medicare levy, which is why the medicare levy “take” increased in real terms for incomes in the approximate range of $15,000 to $35,000 in the chart.

Overall though, for this household type Parliament 41 provided real increases in income over the entire income range covered here.

Tax cuts were obviously seen as a bit of a winner electorally, as both major parties went to the election for Parliament 42 with very similar tax cut proposals.  In the event Labor came to power, and the next chart shows what they did.

Chart 4


This chart reminds me that I should explain the way these household comparisons work.  I’m comparing a household – in this case NSA-DSP – at the election date that kicked off the Parliament in question with the same kind of household at the following election date.  It’s not the same household though, just a similar one.  That’s an important distinction as you’ll see shortly.

The chart itself has the tax cuts I referred to above but, more intriguingly, has something strange going on with DSP.  At zero private income (where maximum rates are paid) we can see that DSP has increased in real terms by a little over 4% but then falls away quite markedly as private income increases.  The partner’s NSA has declined in real terms due to the non-indexation of the income test.  There’s also a very small decline in the maximum rate, but this is really a product of the timing of the elections compared to the dates on which payments are increased under the indexation arrangements. 

The increase in DSP, and its decline with rising incomes, flows from changes to pensions made in September 2009.  Some of the increase is due to the usual wages-based benchmarking, but some reflects an ad-hoc increase in pension rates that was part of a re-setting of the ratio between partnered and single pension payment rates.  Both partnered and single pensioners got an ad-hoc increase, but it was much larger for singles (a look at the Parliament 42 chart for single age pensioners will show how much more significant this was for singles). 

The decline in DSP is because the income test for pensions was tightened up, with the rate of reduction going from 40 cents in the dollar to 50 cents.  This tightening meant that, for this household type, the partner’s earnings have a considerably greater impact than before, so much so that at incomes above around $30,000 the gains from the rate increases are wiped out.

Importantly, no one actually experienced the reduction in the way shown.  Again, I’m comparing two different (but identically constituted) households at the beginning and end of the period.  In that context the reduction is as shown.  But households already in the system when the income test change occurred did not have it applied if it would make them worse off.  Instead, their access to wages linked pension increases was turned off.  They got to keep the old income test, but with a pension that increased only with inflation and cost of living increases, not wages growth.  That slower rate of growth in the pension rate means that it is eventually overtaken by payments made under the new income test where wages linked rate increases are still provided.

Still, the tax cuts in that Parliamentary period were generally larger than the reductions caused by the tighter pension income test, except for private incomes in the approximate range of $62,000 to $70,000.

The other notable feature is the gains caused by relaxations in the medicare levy rules.   At low incomes this flows from the introduction of a lower medicare levy rate in the so-called shade-in for low income families.  At higher incomes the change comes from the introduction of a higher income point before the household become liable for the medicare levy surcharge.

Overall, each of the 3 Parliaments looked at so far have provided real increases in disposable income via tax-transfer changes over almost all the charted income range.   That comes to an end with Parliament 43.

Chart 5

Here we can see a division into what could be two different worlds.   To the left of the $70,000 point there are gains in income support payments which, at zero private income, exceed 6% in real terms.  To the right, a sudden reduction of over 4% caused by increased income tax.  What happened during Parliament 43 to create this divide?

Zero private income is the best point to start because, uniquely among the 4 Parliaments, there’s a real increase in Newstart allowance.  That’s almost entirely due to the household assistance measures introduced as part of the carbon pricing arrangements.  This was intended to provide a boost to income support rates that would exceed the impact of carbon pricing on the CPI.  And it did, at least up until 1 July 2013 which is the end of the period covered by the chart.  In addition, the Government also introduced an income support bonus for some income support recipients.  That enabled the package of assistance for the Newstart partner to increase the total household income by a little under 2% in real terms.

Carbon pricing compensation was also put into the DSP payment, and it plus the usual wages growth based payment increases gave the household an over 4% increase in their total income.  Together the DSP and NSA package increased by over 6% in real terms for those with zero private income.

At incomes from roughly $10,000 to $25,000 some benefit appears from tax cuts, but this turns negative at higher incomes, primarily as a consequence of bracket creep.

When we cross the $70,000 divide things really hot up.  The decline in disposable income is due to increased levels of income tax.  Again, part of this is bracket creep, but mostly it’s due to the loss of access to the spouse tax offset.  Therein lies a tale.

Getting a tax offset in respect of a partner has become rather more difficult over the course of the last two Parliaments.  First, an income test was introduced under which entitlement was lost if the main earners income hit $150,000 (a similar arrangement was introduced for Family Tax Benefit Part B).  Then access began to be removed entirely via an ever increasing age limit for qualification, except in respect of partners who were invalids or carers.

The household shown in the chart is one that would attract the spouse tax offset, so why is entitlement seemingly disappearing at $70,000 instead of the supposed $150,000 income test limit?

When the invalid partner category was introduced, eligibility was based on the invalid partner either getting DSP, or having some appropriate certification that they were an invalid.  The latter category was necessary to accommodate those whose working partner’s income was too high for DSP to be paid, but still under the $150,000 limit.  That second category was discarded with effect from 1 July 2012 leaving only the “is paid DSP” test in place.

Given that DSP is not payable where partner income exceeds roughly $70,000, this means that the $150,000 test has very little application. It might apply in some cases, for example involving part-year entitlements, or where differences in the definition of income between the tax and social security systems come into play, but for those with earnings it seems a new, more severe income test has been introduced by stealth.  Perhaps I’m wrong on this, but the revised taxation legislation seems pretty clear.  Still it would be kind of nice (if a little embarrassing) to be able to add a footnote to this post that I was wrong and these households have not been treated this way.

The linking of the spouse offset with DSP receipt causes a drop in disposable income at the point where DSP entitlement is lost under the income test.  It’s not the only sudden change to occur there though.  You’ll notice that the end of the DSP payment is associated with a sudden, rather than tapered, reduction in the amount paid.  That’s due to the “sudden death” withdrawal of the pension supplement, a part of the overall DSP rate.  Not shown on the chart is the fact that DSP loss also triggers a loss of pensioner concessions, a further financial hit.

These three losses are all stacked together, potentially making an exit from DSP on income grounds a financially painful scenario.  A perverse incentive to remain on payment is created with this type of design, but, as with the earlier comments on unemployment, this is probably a topic for a different post.
There's a bit of a gain showing up at roughly $165,000 due to the increase in the threshold at which the medicare levy surcharge kicks in.  Like DSP rates, these limits are linked to wages growth, not CPI, and this leads to gains in disposable income as the surcharge ceases to apply in the affected income range.

I’ll feature this household again when (if!) I do a final update on Parliament 43 once the CPI for the September 2013 quarter is available.   That will mark the “proper” end of that Parliament at least for this type of presentation.   Presumably, it’s then on to Parliament 44.

No comments:

Post a Comment