O’Toole and the Government on the Pension Levy

This post was written by Karl Whelan

Tonight’s Questions and Answers featured a heated exchange between Fintan O’Toole and Martin Cullen about whether the government knew how the pension levy was going to work when it was introduced. O’Toole has framed this question as being about whether the levy was applied to gross income or to net income. In his original article about this, he had asked

Is the levy a percentage of a worker’s entire income, or of that income after tax and PRSI? The answer to this question is crucial.

The article went on to heavily criticise the Taoiseach for asserting that the levy was applied to gross income and argued that this is not the case, a criticism that was repeated on Q&A.

Let me attempt to cut through the Gordian knot here and explain why both O’Toole and the Taoiseach are sure they are correct. The Taoiseach knows that the government does not double-tax existing pension-related levies (for instance those applied to post-1995 government employees like me) and would likely have assumed that people understood that this also applied to the new levy. From this perspective, let’s compare two different proposals:

  1. Pay pension levy of p% on gross income but this contribution is not also subject to income tax of t%. (Most likely, the Taoiseach’s interpretation of the proposals).
  2. Pay pension levy on net (post-PAYE) income. (Fintan O’Toole’s interpretation).

The take-home pay under the first system works out as follows. The pension levy is applied leaving workers with (1-p)Y, which is then subject to income tax, leaving take-home pay of (1-t)(1-p)Y.

Now consider the second system. Income tax is applied, leaving workers with net income (1-t)Y and then this is subjected to the pension levy, leaving take-home pay of (1-p)(1-t)Y.

So, via the wonders of algebra (commutativity of multiplication!) we see that (1) and (2) are identical. The logic of proportional taxation means there is no difference between applying the levy to the whole salary and then applying income tax to the remainder of income or applying income tax to the whole salary and then applying the pension levy to the remainder of the income.

Where Fintan O’Toole (and many others) are correct about the flawed nature of the government’s calculations, is that tax-exemption of the pension levy payments means that this measure only improves the deficit by €900 million, rather than €1.4 billion figure the government uses. As far as I can recall, I haven’t seen a satisfactory squaring of that particular circle from the government. So score this one, on points, for Finto. But the issue is not whether the levy has been applied to net or gross income.

15 Responses to “O’Toole and the Government on the Pension Levy”

  1. Frank Galton Says:

    Government statement on Saturday:

    The income to be generated by the levy will reduce the cost of the public payroll and it will do so in a way which is progressive, with higher contributions being paid by those at higher levels of income.

    The fact that the government continues to claim that the levy is “progressive” and that higher contributions are paid by higher incomes (not true once tax deductibility and its interaction with the overall tax system is taken into account) suggests that Fintan is right on both points.

  2. Kevin O'Rourke Says:

    The bit of that debate that really hit home in terms of what it says about our government was when the Minister attempted to claim that tax reliefs for landlords were necessary because otherwise capital would flee the country.

  3. karl deeter Says:

    I think the pension levy was ill thought out and a soft approach to what we really need to see which is benchmarking that works on the way up and the way down, as well as phasing out of guaranteed pensions -the forward obligations only ensure another timebomb awaits in the future.

  4. Jim O'Leary Says:

    @Karl W.
    Here’s how I’ve heard the circle squared (though I’m not sure that I’ve seen this explanation in print). The Addendum to the Stability Programme Update, published in early January, which identified the need for €2bn in spending adjustments this year, of which the €1.4bn pension levy forms part, contained tax revenue projections out to 2013. The figure projected for 2009 was €37bn (revised down from the €42.8bn projected at budget-time last October). This figure anticipated (i.e. was net of) the loss of tax receipts to the Exchequer arising from the levy. Who knows if this is true? But if not, it is an ingenious response to the point made by Fintan O’Toole and others. If Martin Cullen had had the presence of mind to use it last night, he might have got the decision from the ringside judges, at least in relation to this particular round of the contest.

  5. James Says:

    @ Jim,

    Isn’t that explanation patently absurd though?

    The government claims to have already taken account, in its pre-adjustment fiscal projections, of the loss of income tax resulting from the levy but not of the revenue from the levy itself? And this before they had concluded negotiations with the social partners over what form the adjustment would take? And when the details of the levy were only proposed at the last minute of those negotiations?

  6. Jim O'Leary Says:

    @ James
    Artful perhaps, but not quite as absurd as you might think.
    Any adjustment, be it a spending cut or a tax increase, generates a partially offsetting loss of tax receipts . So, it would not have been necessary to know the precise shape of the adjustment package in advance in order to make an allowance for its negative effect on tax revenue. Of course, in the absence of such detailed knowledge, any such allowance would necessarily have had to be approximate. In this regard it is worth noting that the effect on tax receipts of a straight €1.4bn cut in public sector pay is exactly the same as the effect of the €1.4bn ‘pension levy’. The only practical difference between the two is that the latter preserves the value (and future cost) of public service pensions; the former would not. Actually, I believe the official €500m figure for tax revenue foregone from the levy to be an underestimate to the tune of about €200m. To be honest though, I think this is a matter of almost trivial importance in the overall scheme of things.

  7. Aidan McGrath Says:

    Isn’t there a third alternative - which was i understand the rationale behind the original 1 and 2% levies. That is that the levy is on the gross (pre-tax) and the tax remains on the gross.

    That way, as with other “levies” the percentage was lower than if he had raised tax rates, and also it could not be avoided by the normal tax loopholes.
    It seemed that only because he insisted on calling it a “pension” levy rather than just a levy, left it open to the victims to claim the normal pension loopholes. Wasn’t it just an oversight on his part.

  8. James Says:

    It still seems odd (albeit possible) to me to make fiscal projections that are supposed to be pre-adjustment, but then to include one knock-on consequence of such adjustments (loss of tax revenue).

    It also seems hard to believe that they would have deliberately designed the levy so as to apply at a higher rate for some lower paid people than for some very highly paid people - the government clearly wanted it to be straightforwardly progressive but making it tax deductible reversed that in some cases.

    I agree this is in one sense a trivial dispute, but this is thus far the centrepiece of the government’s fiscal adjustment and how it came about speaks to the competence of the government. (Perhaps it is at this stage trivial to debate the competence of the government as though that were still an open question…)

  9. Niall Says:

    I have tried in vain to work out the figures. I looked at the Pay & Pensions booklet for 2008 as the best source of actual figures.

    1) Total cost of pay & pensions €19,391M

    2) Pensions €2,371M

    3) Gross Pay €17,020M

    4) Employer PRSI €1,000M (my estimate but reasonable as most are insurable

    5) Employee Pension Contributions €527M

    6) Pay liable to levy €16,000M

    7) Levy @ 10% €1,600M less (300,000 * 15,000 @ 3% plus 300,000 * 5,000 @ 6%) €375M = €1,225M

    As there is an embargo on recruitment, it seems reasonable to use the 2008 figure.

    Assuming 80% of the income is liable at the higher rate and the balance @ 20%, then the tax cost is €600M.

  10. Derek O'Connor Says:

    Do you mean (1-t)*(1-p)*Y or (((100-p)*(100-t))/10000)*Y?

  11. Karl Whelan Says:

    Quick answer to Derek’s question. By a tax rate of t, I mean that a tax rate of 33% implies t=0.33. Understood that way, the two formulae are the same.

  12. Derek O'Connor Says:

    I don’t question the identity of the two formulas. Even Maple and Maxima say they are the same.

    You said originally :
    “1. Pay pension levy of p% on gross income but this contribution is not also subject to income tax of t%. (Most likely, the Taoiseach’s interpretation of the proposals).”

    Taking this statement at face value, 3% means p = 3. Thus 1-p = -2.

    Imagine you have been hired by the Revenue Commissioners or the IRS to write a program that calculates the tax on a person’s income Y. Which formula would you use?

    (1-t)*(1-p)*Y or (((100-p)*(100-t))/10000)*Y?

    Don’t laugh. I’ve seen a similar error in two different student reports over the last 20 years: “Your program predicts that this company’s sales will increase by 4,800% over the next year. Are you sure that’s right?” — “Oh yes. Check my program if you want to.” I did. It had the statement

    p := 3.0;

    If our ‘Feds’ start using the first formula we will have negative income along with negative equity, with or without commutivity.

  13. donal Says:

    The original post by Karl misses the point.

    O’Toole was not debating that one method means more in the pocket for the lay man above the other method.
    He was arguing that the government will miss their target of 1.4bn by about c. 500 million.

    The point is how much extra does the government have after BOTH the levy and tax are applied - That is the issue here.

    If it is the same amount (as CORRECTLY demonstrated in the original article)then this is not important.

    The point is how much will the levy add to the state in net terms.

    Applying the levy to the gross results in the following:

    350,000 workers at a weighted average levy rate of 7.5% at an average salary of €50k = c. 1.4bn.
    However - as a result, this 1.4bn will now not come under regular income tax and will be lost at this part of the computation.
    => Net amount added to state = 900 million.

    If tax is applied first and then the levy, then as the original article stated, multiplication is commutative and 900 million net is again added to the exchequer.

    Either way - they are 500 million short of their stated 1.4bn target - and that was O’Tooles point !

  14. donal Says:

    Sorry - Ijust realised that the original article did not miss the point. We are in fact in agreement. (i.e. Bottom line is we are short half a billion before we’ve even started !)

  15. Karl Whelan Says:

    Not to worry Donal. Sometimes the urge to declare that “Somone is Wrong on the Internet” can just be too strong …..

    http://xkcd.com/386/

Leave a Reply