Thursday, 17 December 2015

Swinney: Confused or Aiming to Confuse?

Last night (16/12/2015) on the BBC's Scotland 2015 programme John Swinney was asked why he had chosen not to use the Scottish Rate of Income Tax (SRIT) to mitigate the pressure on public spending driven by Westminster's austerity policies.

If you watch from 13 minutes in, you'll see he's pressed as to why he didn't increase the SRIT by even just 1p in the pound. The following is a verbatim transcript:
"Well what that would have done would've put a disproportionate impact, and had a disproportionate impact on the incomes of people in low income households. It would have in fact been about double the effect on the taxable income of individuals at the lower income thresholds rather than people at higher income thresholds and I don't judge that to be the right way to deploy any tax raising changes"
Similar statements are now being parroted around social media. By seeding this misunderstanding Swinney shows that he either doesn't understand the effect of Personal Allowances on effective tax rates or that he is consciously trying to deceive the Scottish electorate.

Let me explain.

If all income was subject to tax then the statement above would make sense. If for example low earners paid 20% on all income and higher earners 40% then adding a flat 1% additional tax to both would indeed be "disproportionate". The lower earners' tax burden would rise by 1/20 = 5.0% whereas higher earners would rise by 1/40 = 2.5%.  I think this is how most people intuitively think about tax; it's an intuitive misconception that Swinney either shares or wishes to exploit.

Maybe Swinney is unaware of this thing called the Personal Allowance, the threshold above which tax has to be paid?

To illustrate with realistic (but rounded) numbers: say the Personal Allowance (PA) is £10k and we look at a low earner on £20k and a higher earner on £40k.  The 1% additional tax is only paid on income above the Personal Allowance, so

  • The lower earner pays the additional 1% on £10k (£20k income minus £10k PA) so has an additional tax bill of £100
  • The higher earner pays the additional 1% on £30k (£40k of income minus £10k PA) so has an additional tax bill of £300
So the lower earner is seeing a further £100/£20k = 0.5% of their total income being lost but the higher earner sees a further £300/£40k = 0.75% of theirs.

The simplest explanations is this: The Personal Allowance means that the higher an individual's earnings, the larger the proportion of their income that is subject to the additional 1%.

I've used round numbers here to try and make the illustration clear, but they're pretty close to the actual figures. For those who like to see the actual numbers (and a graph) I include them at the end of this post.

But this is about more than arguing over how to interpret percentages. These additional taxes would be directly available to the Scottish Government to alleviate spending cuts, to protect the public services on which the most needy in society depend.

Let's take our worked example further: what does the Scottish Government do with the extra £400 they've raised from our two tax payers? Even if they couldn't find a way of spending it to disproportionately help the lower earner (as they surely could), both the earners in our scenario could benefit from £200 of additional public spending (their "population share" of the additional £400 raised).

The net effect would therefore be that our lower earner would be £100 better off (because they pay £100 more tax but they benefit from £200 more public spending). This would have been achieved by our higher earner making a net contribution of £100 more (paying £300 more in tax but only receiving back £200 by way of public spending).

Of course there are non-earners who could benefit from those additional taxes, which would be even more redistributive than the example above.

SRIT may be a blunt instrument, but it undeniably has the potential to redistribute wealth from the richer to the poorer.

So now let's return to Swinney's statement:
"It would have in fact been about double the effect on the taxable income of individuals at the lower income thresholds rather than people at higher income thresholds"
It appears to me that this statement is demonstrably untrue. If I'm right then Swinney is either an incompetent Finance Minister or a liar.

***** Clarification / Correction *****
When I first published this post I offered that if anybody could demonstrate that I'm wrong - that his statement could be justified with reference to robust factual analysis - then I would of course post an appropriate correction and apology here. So here we go ...

It appears Swinney's figure comes from taking the %age increase in the %age effective tax rate (which is a bizarre way to interpret "proportionate impact" as the explanation above I hope proves - taking percentages of percentages nearly always leads to confusion).

In our worked example (see table below) the effective tax rates are 9% and 15% respectively. If we take the increase in effective rate and divide it by the effective rate, our £20k and £40k earners both see a 5% increase (the numbers are rounded in the table;  0.47/9.40 and 0.74/0.147 respectively). At £60k the increase by this measure becomes 3.69% (0.82/22.34), at the higher threshold (£150k) the figure becomes 2.8%. So 5.0/2.8 = "proportionately 1.8 times more" i.e. double in Mr Swinney's rhetoric (I'm not quibbling abut 1.8 vs 2.0).

If you've understood the worked example above, I hope you'll agree using this figure to argue that lower earners are disproportionately impacted is disingenuous at best.

I'm happy to apologise for suggesting Swinney might have been lying. I maintain however that he either doesn't really understand what the figures mean or he is trying to confuse people

**** Clarification Ends *****

The Smith Commission powers being delivered through the Scotland Act 2015 will give the Scottish Government the ability to flex all income tax thresholds and rates. Let's hope that by the time those powers are delivered our SNP overlords have managed to understand how taxes work and/or start being honest about the choices they are making.

Is that too much to ask?

The following graph shows the effect of the personal allowance (note the personal allowance gets phased out between £100k and £121k) and changing tax rates at different thresholds. Note how the red line (adding 1% to SRIT) subtly diverges from the blue line - this is the graphical representation of the dynamics this blog describes.

I'm indebted to Andy Wightman for pointing out the existence of this incredibly misleading graphic (which simply ignores the Personal Allowances diluting impact for lower rate tax payers). I'm genuinely shocked to find it comes from this Scottish Parliament Information Centre (SPICe) briefing paper. Holyrood: we have a problem.


Ed Macfarlane said...

Is there anyway of including a 'net effect' line?

Anonymous said...

Purely my opinion, but I believe the reason why the SNP will not touch tax rates (which can only go up) this side of the Scottish Elections should be self-explanatory. To create disparity of earnings from rUK means he runs the risk of alienating people who would likely 'jump ship' if the trend of higher taxes in Scotland sets in. Equally, he should have used the tax-raising powers while he still has the UK's block grant topping up our deficit gap. No such cushion would be in place if we became independent and almost guarantee that tax on income would rise to cover the shortfall.

Of course, none of the above excuses the pitiful reason given.

Garve said...

Sticking with your policy of using round numbers to demonstrate...

If we have 2 million lower rate taxpayers and 500,000 higher rate ones then

Lower rate taxpayers will pay a total of £200 million extra. (£100 * 2 million)
Higher rate taxpayers will pay a total of £150 million extra. (£300 * 500,000)

So in the hypothetical situation where the Scottish Government couldn't spend the money to disproportionately benefit poorer families, higher rate taxpayers would be subsidised to the tune of £50m by lower rate ones.

That is a hypothetical situation of course, and lumping taxpayers into the two groups is a dubious methodology.

However, it's a lot more sensible to consider the effect on poorer households rather than the raw numbers. In general poorer households have lower disposable income than richer ones. So the effect of a 1% rise in tax on someone on £20k can be far more than a 1% increase to someone on £40k, even though they're paying a lot more in actual pounds. So a sledgehammer tax increase such as SRIT will have more of an impact on households with low income.

I imagine the Scottish Govt considered the impact of £100 lost to a poor family compared to £300 lost to a well off one, and that may well have been the impact JS was talking about.

Sheumais said...

I suspect Swinney's thinking may be an increase from 20 to 21 is 5% more, whereas 40 to 41 is only 2.5% more. Yes, it does require a fundamental lack of understanding of the application of taxation, but I have seen no evidence from Swinney to suggest he is suitably equipped to cope with finances.

Kevin Hague said...

Garve - I don't disagree with that - it is a blunt tool. That's why I agree with tax rates tiering as they do - the incremental income being taxed at higher bands is less vital to maintaining basic standards of living.

My point still stands though - SG could definitely use SRIT to benefit less well off (non-earners of course but also per the illustration) - unless you believe cuts to public spending doen't hurt the poorest most (I believe they do).

Note someone on NMW would only be £34/year worse of - surely the taxes raised across the board could be used to benefit them by more than that?

Neil Lovatt said...

Just a note on your clarification Kevin. You would not take a percentage of a percentage because it defeats the purpose of a percentage in the first place, all you are doing is deliberately gearing the effect.

Garve - has a fair point in that the impact of a small increase in disposable income would have a bigger effect on a low paid fairly compared to a well off one.

However the trouble with both these arguments is that it means that in which case you should cut SRIT and I haven't seen anyone arguing that.

Anonymous said...

Talk percentages as much as you like. The person on 20K loses £100, this is £100 more than they can afford and no amount of extra public spending will put that £100 back into their pocket. If you look at public service workers that would equate to almost no wage rise as theirs has been limited to 1% for each of the past 6 years.

Edwin Moore said...

A poetaster whines at the receding prospect of any free dosh from ‘Creative' Scotland -

I think that I shall never see
A grant of money bound for me.
But efter all it isnae odd,
The purse is held by Swinney Todd

Douglad (Scotland) said...

Yea. Because local councils will transfer that benifit to the least well off.

Douglad (Scotland) said...

In reality, control over all tax rate bands is required. Raising the % for the higher tax rate bands would not inadvertently hurt the least we'll off.

The UK had such a complicated tax system overall that inevitably the least well off end up paying the most.

ndcalvert27 said...

You were right first time. Being that disingenuous is effectively being dishonest.

Anonymous said...

A thought occurs to me, Kev - isn't this the same guy who was advocating a 'penny for Scotland' back in the day, which basically meant a penny rise on all bands (as he did not have the power to differentiate)? He should be getting called out for this flip-flopping claptrap.

I have never, ever rated the guy, and whenever I see him stuttering and simpering on TV, I am reminded of Kevin Spacey talking about Keiser Soze in The Usual Suspects, whom I would paraphrase: the greatest trick the SNP ever pulled was convincing people John Swinney was ever anything other than thoroughly inept.

David GREEN said...

The last Anonymous comment is wrong. The person on 20K doesn't really lose £100. The £100 is placed in Mr Swinney's hands, and, trust me, he knows what to do with it much better than the person giving it up. Remember, Swinney's fighting austerity. For example, by giving free tertiary education to EU students. The idea of leaving purchasing power in the hands of those who generate the money is very last century.

Neil King said...

Kevin, pardon my ignorance but is your clarification (which I didn't understand) to make the point which Sheumais made in his comment above (which I did)? If it is, then Swinney may not technically have lied but he's guilty of the equally heinous crime of being shamelessly disingenuous in a deliberate attempt to mislead the public.

Neil said...

Clever, but how do you fit all this into 30 second sound bite or a newspaper headline.
For that reason flat rate tax, flat though they may not be as you point out, are political non-starters.

cabsandy said...

I can see where your going but what you dont take into account is the amount each higher rate taxpayer earns-so a HRTP on £45,000 pays a lot less tax than one on £100,000! For your argument to work completely, you'd need to break down (say in blocks of £10,000's) how many higher rate taxpayers pay what. Same argument for lower rate taxpayer. So I'm told, 20% of Scotland's tax take comes from 1% of the workforce.If even a small amount of those people take flight, then a hole starts to appear....

Anonymous said...

Thanks for the insight. It would be useful to have a graphic to go along with your clarification. To capture it in ordinary language, one might say that the rate of increase in the proportion of tax paid as a percentage of salary is greater for those on lower salaries than those on higher salaries - is that correct?