Why ‘Will it encourage saving?’ is the wrong question.

It’s heating up, again, pension tax reform.

The Conservatives seem to have ditched plans first proposed by the Centre for Policy Studies to remove up front tax relief for pension savings and replace it with back-end relief – when those savings are expended. And quite rightly too, because that otherwise fascinating CPS paper neglected the key question: whether such a change would encourage those who need to to save. It’s difficult to make the case that it will.

What has been mooted instead is replacing the iniquitous current system – which gives 40 of tax relief to a higher rate tax payer contributing 100 but only 20 to a basic rate payer making that same contribution – with a flat rate. Assume for the sake of argument, 33.

Such a change would provide further encouragement to basic rate payers to save – but diminish that for higher rate payers. And this has, in turn, sparked a row about whether the Financial Secretary to the Treasury, David Gauke MP, was right to tell the BBC:

“We need to ensure it is effective in terms of encouraging saving, and it is going in the right place”

And the arguments as to why a flat rate would or wouldn’t encourage saving are rehearsed in this typically thorough piece by Jim Pickard of the Financial Times.

Sadly, not surprisingly, it’s a typical political argument: under-evidenced, riven with sectarian division – and about completely the wrong thing.

The right thing is ‘what is the purpose of pension tax relief?’

If you proceed from the assumption – as I do – that its purpose is to encourage people to save for their old age so that they won’t become a burden on the State, you should really provide the greatest incentive to those otherwise most likely to become that burden. And that’s those on lower incomes.

The higher your income, the more likely you are to have surplus income – which, it being surplus, you will accumulate as savings even absent a tax incentive. And there is the closely related point – the higher your income the more likely you are to accumulate assets over your lifetime which you can liquidate during your retirement to provide for your then needs.

Putting the matter more acutely, what we shouldn’t be interested in is whether a flat rate pension tax relief would encourage aggregate savings. What we should be interested in is the change (relative to the present system) to the distribution of savings that a flat rate relief would effect. Will people more likely to be a burden on the state save more?

And the answer to that question, in the case of a flat rate pension tax relief, is indubitably yes.

28 thoughts on “Why ‘Will it encourage saving?’ is the wrong question.

  1. I don’t really see the logic in your argument. Surely the objective is to encourage pension savings by all taxpayers whether they be basic rate or higher rate?

    The idea that higher rate taxpayers will necessarily have sufficient surplus income that they will save anyway is not evident given that the higher rate kicks in at a fairly low income level compared to other countries. Depending on circumstances (mortgage, school fees, etc.), that higher rate taxpayer may not have a lot of surplus income and the last thing we need is a disincentive to saving.

    For basic rate payers there are two problems. The first is to provide sufficient incentive to saving. The second is whether, even with saving, they are able to accumulate sufficient pension funds to really make a difference.

    If there really is a public policy imperative to encourage pension saving then, logically, we go down the Swiss route (2nd Pillar) and make it obligatory along with employer matching, portability and tax relief.

  2. I totally agree with this. The only issue is about defined benefit (final salary) pensions where a move away from the current system will become horribly complicated. I can see a strong argument for moving to two separate systems: flat rate relief for defined contribution, and marginal rate relief (the current system) for defined benefit. The only issue with that is that does really further exacerbate the “gap” between those who are fortunate enough to still be in a defined benefit and those who aren’t.

    Thanks – I get a lot out of all your blogs.

  3. (To be clear, I “agree” with the original post, not Charles’ comment which appeared while I was typing!).

  4. ‘What is the purpose of pension tax relief?’

    My answer is that it is simply a way of allowing the tax on income to be deferred from the year it is earnt to the year it becomes available to the taxpayer, reflecting the fact that the amount contributed to the pension is not available until retirement – the point being to smooth income from your working life to cover retirement too.

    That is, I see a pension contribution effectively as having been diverted away from me before it becomes taxable in my hands, and held in limbo for a number of years. Using that model, you can’t muck about with the rate of tax relief as there is nothing to apply a rate to – the “relief” is simply cancelling the charge on income that it looked as if I was going to receive but, as it turns out, didn’t.

    If you say that income is received by me and taxed, and the tax is refunded when I contribute the cash to my pension, then a rate of relief makes sense. That is more in line with how cash flows, but not necessarily in line with the conceptual basis of pension funds.

    My problem with looking at it the second way is that it introduces the possibility of double-taxation, or indeed non-taxation. If you put 75 in and get 25 of relief to top it up to 100, then get tax of 20 when you take it out, then you have an active subsidy of savings, rather than a smoothing of income

    The tax-free lump sum does a similar thing, which is why I don’t really like that concept either.

    So to me the underlying question is ‘what is the purpose of pensions?’ – is it to save cash, or is it to defer income?

  5. Yes. Keynes, at least in my limited understanding of his work, overturned the idea that interest rates determine how much an individual saves. It is instead the individual’s income, not the interest (or savings) rate, that determines how much that same individual saves.

    In Keynesian economics the interest rate determines whether savings are held as money or as bonds (or other financial assets) AFTER the decision to save has been made. The decision whether to save is independent of the interest rate.

    I doubt that the flat rate pension relief, even if set at 33%, will encourage those on low incomes to save. In the majority of case, I fear low-to-middle earners simply won’t have the discretionary income (or surplus) to make that decision.

    Still, the government, as represented by a relatively enlightened Gauke, is being dragged slowly towards adoption of clever Keynes’s insights, albeit kicking and screaming. But it is still a case of the wheel being rediscovered.

    C++ to the government for effort.

  6. I don’t think you can talk of pension tax relief as simply deferring a tax charge. It also moves income from high marginal tax rate years into years more likely to be low marginal tax rate years. Indeed, with a high personal allowance it can easily move income from high tax rate years into no tax rate years.

  7. Absolutely right – only I think the change of rate is a feature, not a bug.

    The point is not to defer the tax charge itself, but to defer the recognition of the income for tax purposes to match the receipt for economic purposes – and for all tax matters (rates, allowances, payment dates, etc) to flow consistently from that.

    After all, once the cash is in the pension you can’t benefit from it, so why should you be taxed on it? Pension tax relief is simply taxing you on a cash basis rather than accruals. For an individual that seems rather fairer than having dry tax charges – or in fact a double tax charge, if you have 33% relief on contributions and 40% tax on withdrawals.

  8. But there’s a huge cost attached to the State permitting that deferral which is why it’s right to ask what State benefit attaches to that cost. And it’s pretty unlikely – if not impossible – that with the pension pot cap you’ll ever generate income from your pension pot which is taxed at 40%.

  9. Worth bearing in mind that the tax treatment of pension savings is just one piece of the jigsaw. Other factors include loss of liquidity because the cash is locked in, the extent of employer matching, portability in the event of a change of job, confidence in the pension provider and confidence in government not to ‘raid’ pensions at some point in the future or otherwise change the tax regime.

  10. Even if you get “only” 33% on way in, and pay 40% on way out:

    33% income tax relief
    At least 2% NI saving
    Years of tax free growth
    25% tax free at exit
    Balance out at 40%

    It is still quite easy to see that as attractive and certainly not penal “double taxation”.

  11. I go back to the idea that a pension is something that you get to look after you when you too old to work. From a public policy perspective, that has to be a good thing. I imagine an old-fashioned miner, coughing up coal dust who has to stop work and the pension was there to compensate them in the few years before they died.

    For some people, this is still the case. And I have absolutely no issue whatsoever of giving tax relief to encourage this. I think that the negative consent method of auto-enrolment is a great idea of helping to ensure that people can be provided with an adequate pension. While we might argue whether or not the provision of such a pension should be the sole responsibility of the state, having the ability to supplement a minimum pension is a good thing.

    For others though, pensions are a deferred income. There is an arbitrage between the tax relief (and potentially NIC relief) now versus a tax-free pre-commencement lump sum and the chance of lower tax rates later. It doesn’t take much spreadsheet skills to show that if you pay high rates of tax now but can manage your income on retirement so that you pay no, or little tax, on your other income in some years then it is better to save for a pension than over-pay your mortgage and use the tax savings to repay the mortgage later.

    For this type of person, some or all of the pension is not really a pension and I don’t think that the tax system should encourage excess pensions. A flat rate of tax relief is a great way of achieving this. Especially when presented as tax relief actually being more than the tax you would otherwise pay.

    I do struggle with defined benefit schemes in this regime though. I don’t see a practical way to move into this regime and expect that pensions will turn into a two tier systems with those employed by the state having a defined benefit scheme and the rest of us having a defined contribution one. The current system hugely advantages employees with defined benefits and, while I believe that is unfair, I think it will get worse.

    So for me, a new system of tax relief for pension has to increase the ‘valleys’ of pension savings and topple the ‘peaks’ (caused by the current system). While auto-enrolment also helps with the valleys. I can only see a new tax system dealing with the peaks.

  12. Andrew Jackson sets out beautifully the -once upon a time – true nature of pension tax “relief”, to charge the income when drawn rather than when first earned, but including the capital accumulated free of tax in the meantime. You refer to the income therefore being taxed at a lower marginal rate when finally drawn, but that is simply a reflection of the lensioner’s income being lower than it had been when he was working. I can see no reason why this ‘model’, this way of viewing it, should be in dispute.
    I also agree with Andrew’s disquiet over tax – free lump sums etc. This moves u the pension saver away from that link to deferred and away from the very nature of the pension itself. I hear my colleagues at work talk in terms of “taking it while you’re still young enough to enjoy it” and “you can’t take an annuity with you, so get a lump sum”. The idea that the USA might suit them.more doesn’t seem to occur.
    Be that as it may, however, that is where we have been headed for some time now. These features have caused us to lose the concept and I don’t see us ever regaining it. The environment we have created has made it legitimate for politicians of all hues to consider changes, upon recent change, upon previous change to the system. In that context your comment at 5.31 is significant. Talk of the “cost to the government” inadvertantly or otherwise touches on a very significant, perhaps the most significant driver in all this: the fiscal deficit. A chancellor absolutely committed to reducing thay deficit is looking for ways to do so. He will have realised the lack of appetite to reduce the welfare bill further. He will know the now inviolable political impertinent to increase the NHS budget ever more. He is committed to raising PAs and NI thresholds (quite rightly) and he will certainly not want to raise the headline rates of IT to.whay will appear eye – watering levels. So restricting, er, ‘relief’ on pension contributions will give him the big hit he needs. And there is a redistributional element thay will turn on lefty commentators no end (hello Jolyon).
    The quid-pro-quo will need to be a lower tax rate when the pension is drawn or higher and higher tax free lump sums. Of course this means he is shifting the debt burden to a future generation, but we’ll all be dead by then.

  13. The proposed 33% top-up doesn’t seem like a huge new incentive to save for basic rate payers.

    Basic rate payers, those who employer contributes to their pension (as a result of either a contribution being part of the employment package, common with automatic enrollment or via salary sacrifice) already save 33.8% (20% tax & 13.8% NI), which is much closer to the 42% (40% tax & 2% NI) saved by most higher rate tax payers, than the government is presenting.

    Different relief rates (not taking account of effect on any tax credit entitlements) would seem to be:

    20% Zero tax payer (paid to pension scheme – limited to £3,600 gross contribution pa)
    20% Basic rate personal contributions
    33.8% Basic rate employer contributions
    40% Higher rate personal contributions
    42% higher rate employer contributions
    45/47% Additional rate personal/employer contributions
    60% (child benefit reduction range / personal allowance reduction)

    But the solution here would seem to allow NIC refund for personal contributions, and remove the random high effective marginal rates although that would cost money rather than generate significant additional current receipts.

    Their are already newspaper stories of (and I am aware of this to some extent anecdotaly as well) of people declining to join pension schemes in part due to the drop in current income.

    If we moved to a position where employers DC contributions were taxable, but received 33% additional cash in the pension scheme, this would reduce the pay packet by 20% of the employers contributions. Overall this would be offset by higher cash in the pension scheme, but the immediate pay packet impact may discourage savings.

    However if viewed as a short term revenue raising measure, the proposed changes seem very effective (long term dynamic effects with impact on future tax receipts on pension payments don’t seem to be properly analysed in any government costing of current relief), although I suspect it will also raise revenues in the long term.

  14. “But there’s a huge cost attached to the State permitting that deferral which is why it’s right to ask what State benefit attaches to that cost. ”

    This is where it depends which side of the fence you’re looking from 🙂

    From the taxpayer’s point of view, if they have not received income then it is hard to see why they should be taxed on it. The state isn’t incurring a cost, any more than it incurs a cost when a self-employed person takes the day off and doesn’t earn anything – it was never entitled to a benefit.

    From the state’s point of view, if income has been received by someone (in this case the pension fund) then arguably someone should be taxed on it. Which of course they are at the moment: although the fund isn’t taxed, the pensioner is taxed on what come out of the fund.

    I think it’s worth considering whether a pension fund ought to be tax-free, but I think that if you were to tax pension funds then you’d have to look at how pensions received are taxed: to tax them as is done at present would give double taxation, but to give credit for the tax suffered in the fund would be complex (see trust tax – which of course is a pretty closely related situation).

    Coming back to the question of the benefit the state obtains, I think it is fairly plain: by allowing the deferral it encourages the taxpayer to provide for their retirement, and so it obtains a taxpayer who is self-sufficient in retirement.

    This all presumes that the state is something more than the sum of all taxpayers, of course, which is a philosophical debate for another day 🙂

  15. But there is only one side of the fence. It is their income when they earn it. The state permits the creation of the statutory fiction that it’s earned later – when the taxpayer actually receives. There is a tax cost attached to that deferral.

  16. The whole pension world was “simplified” in 2006 by the then Labour administration. Two simple concepts: a lifetime limit and an annual limit.It is shocking to me how far the recent Tory led administrations have reduced these benefits.
    The obvious anomaly (which no one wants to address) is the tax-free lump sum. Why is it tax free on the way out when it was tax free on the way in?
    Finally, each week in the Sunday Times Money section, the question is posed to a well known person-Which is better, property or pension? Most people prefer property and, looking ahead, I see that the proposed pension changes will push more people towards the property market.

  17. I agree with all of that. And especially that there is no justification at all for the lump sum.

  18. The issue it seems to me is that higher rate tax kicks in at about £45k of earnings. That just isn’t enough to ensure you’ll have a comfortable retirement – yes you are well off, but you aren’t rich enough to live on your assets for 20 years. If higher rate taxpayers stop saving then we store up a huge problem for the future.

  19. Whilst on the subject of flat rates, why not have flat rate contributions to the state pension scheme?

  20. The state doesn’t create a statutory fiction that income is received later, it creates a legal reality – you cannot get access to those funds.

    For other purposes, you are not treated as receiving income until and unless you are able to spend the cash – look at the taxation of interest, for example.

    I suspect that the problem is that pensions originated as DB, and DC schemes are trying to replicate a DB position. It is much clearer that a DB scheme gives you income later, and indeed the tax-free lump sum makes more sense (a lump sum is nice, but it isn’t really income, so don’t tax it as such).

    However, if one can argue that cash in a DC scheme which you cannot use is taxable when you earn it, then one could equally argue that the right to receive future income under a DB scheme should be taxed when you earn it. That takes one into a whole lot of complications, assumptions, and of course dry tax charges (unless you relieve them).

  21. I confess to being puzzled by what the politicians are trying to achieve here other than an easy stealth tax increase.
    Very rich people will not care as a £1m lifetime limit is practically irrelevant to them.
    Young working people on normal wages probably can’t afford to save into their pension beyond the basic requirement and,for many of them,pensions are too far off to worry about.
    So it amounts to an attack on what the press call the “squeezed middle”. People with a good job and a decent income who are probably doctors, senior teachers and policemen as well as businessmen and City workers.
    Watch for GPs and similar retiring early as they cannot increase their pensions because they will have reached the new maximum lifetime limit.

  22. No sorry Joy on, I can’t agree with this: “The state permits the creation of the statutory fiction that it’s earned later – when the taxpayer actually receives.”

    As regards the taxpayer’s contribution to the fund, taxing it.only when he draws it from the fund is only placing him in the same position as everyone receiving employment income. As regards accumulated sums in the fund that he draws, yes, there is a definite DEFERRAL of that yax. There is.no absolute tax advantage, however. We can either tax the fund.as the income and gains accrue, or.we can tax the pensioner when he draws it; anything gore would be double taxation.

    I do take Andrew’s point ( as usual) that the Sate loses nothing by the present system of ‘allowance’ now and then charge when he draws the capital. That, however, isn’t the point. Denial of relief now.is on reality a tax hike. That is what the Chancellor thinks he needs though, so…

  23. You earn it when you earn it. If you don’t put your earnings into a pension they are taxed. If you do put them into a pension fund the tax on them is deferred until you draw them out. When your marginal tax rate is usually – materially invariably – lower. But it’s a statutory fiction that you don’t earn them when you pay them into the pension fund. It’s a statutory fiction because the reality is that you do.

  24. So when do you earn the earnings you receive from a DB scheme?

  25. As you pass the year that gives you the extra integer on your numerator.

  26. Jolyon, are you suggesting that those few lucky people left in a DB scheme should also be taxed each year on the increase in the value of their pension rights (say, the amount they would have had to contribute to a DC scheme to get an equivalent increase in value, less an investment return on the value of their pension rights at the beginning of the year), at the difference between their marginal rate of income tax and your fixed pensions relief rate (say, 33%)?

    Back to the question, what is pensions relief for? Surely to encourage more people to put more money aside to support themselves in their retirement. If tax relief is largely taken away, why would someone choose to lock up money for decades in a pension, rather than say put it all in ISAs where it will grow tax free anyway but fully accessible so all of it would be available as a tax free lump sum (or indeed all of it could be spent before retirement) .

  27. I’m done Andrew. Thanks for your input.

  28. I’ll get my coat 🙂

    But I see the ABI is also advocating a flat-rate tax relief for pension contributions, pitched at say 25% or 33% so above the basic rate of income tax and flagged as a “Savers’ Bonus”. https://www.abi.org.uk/News/News-releases/2016/01/Three-key-reasons-why-the-Chancellor-should-opt-for-a-flat-rate-system-of-pension-tax-relief

    However, they do not discuss the effect on defined benefit pension schemes.

    No doubt you would welcome the distributional effects (less tax relief for higher and additional rate taxpayers; subsidy for others) but I wonder what the effect will be on the total amount contributed to pension schemes. Is a few pounds of tax relief really going to incentivise thousands of people on the minimum wage to save more? Certainly, I can see that withdrawing relief for higher-rate taxpayers will encourage them to save less. Perhaps we don’t care about what they do as that have more than enough already? But if they spend it all as they earn it, they will be back claiming benefits in retirement.

Comments are closed.