Ritchie on civil servants service companies

And, try as you might, someone working as a senior civil servant will always come out as an employee using this test.

Hmm, I wonder here, I really do.

What actually is the effect on govt finances of paying people through such service companies?

As far as I can see there are three major points.

1) On the taxes paid by the individuals. The combination of corporation tax and taxes on dividends, or income tax if they pay themselves a wage from the company, this is pretty much a wash. There\’s an NI loss if they pay themselves dividends, true, but then that can\’t be what is complained about because Ritchie did that himself. Indeed, published an article explaining how to do it.

2) The government doesn\’t have to pay employers\’ NI on those wages. This would be a fair old whack too: 14 % of so of total wages (for someone on £150 k a year the NI free allowance is almost irrelevant). That\’s money the government would get straight back of course. So the nett effect on govt finances is exactly zero. They don\’t collect it, true, but then they don\’t have to pay it out either.

3) And this is where it gets murkier. If they are civil servants, paid through PAYE, then I very much assume that they accumulate pensions at the normal civil servants\’ rate. If they\’re paid in service companies they won\’t. And while I don\’t know I would bet that this saving on subsidy to pension plans far outweighs the loss of NI in 1).

I think I would be right in saying that such a payment method, the service company, reduces, on nett, the cost to government of employing such people. The income tax being pretty much a wash, the pensions savings being greater than the NI losses.

And if that\’s true, then what\’s the problem?

And I do recommend reading the comments to Murphy\’s piece. Several dig up the pieces he wrote for the G back when he was advising people how to run service companies.

40 comments on “Ritchie on civil servants service companies

  1. I’m all in favour of people dodging or defying the state’s bullshit, thieving laws.

    When it comes to political/bureaucratic scum trying to dodge their own obnoxious laws there has to be a line drawn.

    Time to get rid of said laws–even the vermin can’t stand them anymore.

  2. Thanks to Gordon Brown they can pay CGT at a reduced rate on most of their ill-gotten gains instead of paying high-rate income tax on dividends if they “retain” most of the profits in the company for ten years. If there is any problem about cash flow to pay for their holidays in Tuscany they can borrow money (interest-free) from the company and pay it back later.
    Secondly declare income just below higher rate tax threshold for the worker and an equal amount for spouse, so get £80k at basic rate. So there is a loss of tax.
    As pointed out this was a device for New Labour to hire people and pretend they were being paid less than they effectively are, concealing the extent to which the public sector are paid more than the private sector.

  3. @Tim:

    “And if that’s true, then what’s the problem?”

    This is not about the comparable cost to government between PAYE under contract of service and payment under a contract FOR services.

    It is much simpler than that, under the civil service code there are ceilings which operate preventing excessive salaries.

    It is these salary ceilings that were being bypassed more than just the PAYE.

  4. He is getting the boot put in about his earlier advice – he must think everyone came down in the last shower if he reckons his excuse about there being no formal HMRC notification scheme is going to wash

  5. BenM

    Facepalm. It is not the TOOL that is being discredited, it is the use that is made of it. The idea that, for example, a 95% top tax rate is a major discouragement to people to bother to earn more is, I think, well established. It is the point at which tax rates turn from a positive effect to a negative effect that is disputed.

  6. I can’t imagine that any of these types are sacrificing a public sector pension in order to duck a spot of tax. I think it’s far more likely that the value of the pension they’re not getting has been loaded onto the ‘salary’ somewhere along the line (perhaps linked into John Galt’s point above).

    That said, if by paying people this way the state simply passes over the annuity/investment risk of the pension to the individual.. then it’s probably still a good deal. I’m happy to permanently exempt all public sector workers from NI if, in return, they take on their own pension risk.

  7. The Laffer curve isn’t so much a theory as a statement of the blindingly obvious. As Frances says, the debate is over the inflection point of the curve, not its existence.

  8. “2) The government doesn’t have to pay employers’ NI on those wages. This would be a fair old whack too: 14 % of so of total wages (for someone on £150 k a year the NI free allowance is almost irrelevant). That’s money the government would get straight back of course. So the nett effect on govt finances is exactly zero. They don’t collect it, true, but then they don’t have to pay it out either.”

    Could this thinking be extended to income tax as well? You could seriously reduce public spending on wages if you did.(and tax revenues would fall by an equal amount) Entirely fiddling the figures though.

    Are these people costing the departments the same as if they were employed (wage + employers NI) or just the wages part?

  9. they can borrow money (interest-free) from the company and pay it back later.

    Except that any loan over £5k and below HMRC approved interest rates is a taxable benefit and, if you are a close company, which any rational service company is, loans to participants will also attract a tax charge of 25% (if over £15k and more than 9 months over accounting period.)

  10. As Frances says, the debate is over the inflection point of the curve, not its existence.

    And the shape, dear boy, and the shape (and noting that there may well be multiple inflection points, especially if you expand the inputs to include benefits.)

    Except that you underestimate the economic ignorance of much of the left. You and I, and Frances, may consider the existence of the Laffer curve self-evident. But many refuse to accept this.

    I would also note that the study being debunked does not actually measure the tax-intake (which is all the Laffer curve applies to). It measures overall growth and doesn’t allow for the income tax vs sales tax state funding issues in the US. That being said, the debunking doesn’t seem to appreciate that people moving to your jurisdiction because you have a nice tax environment, and bringing their economic activity with them, is generally seen as a good thing rather than a bad thing.

  11. Shape as well, you’re right. What’s actually defined by the Laffer curve theory is the two end points, right? So what the nitwits are disagreeing with is that if you have two points on a graph you can connect them with a curve, which is… sigh.

  12. And, clearly, the curve goes upwards because we know that governments do indeed collect tax revenues. Given 2 end points and at least some area of the curve above zero (remembering that most discussions of the Laffer curve do not include the collection costs of the tax) means, by basic graph theory, that there are one or more maximum inflection points and at least one of them is an absolute maximum.

    Mathematically, it is trite. Economically, it is much debated. Which says quite a lot about the comparative influences of political ideology versus arithmetic competence on the general discussion of economics.

    And, frankly, given that the WGCE was an accountant before the ubiquity of desktop computing, some degree of arithmetic competence shouldn’t be too much to have asked.

  13. But the latest attempt to revive it as a mainstream tool for political engineering, as in the article, has been proven false, even opposite, to what is being hyped.

    get over it, it can never be used for anything other than to promote forever lowering taxes until all the poor die and Ayn Rand gets canonised.

    You all talk bollocks all of the time. Do you wank while typing.

  14. Arnald

    It can be, and is, used to promote raising taxes. The point is that it seeks to find the optimal point.. and that is the point at which tax is maximised. If one took the view that taxes were too low, and that the public would happily pay more of them, then one could use the laffer curve to demonstrate this.

    It so happens that the curve seems more often used to promote lowering (or non-raising) of tax. That may be because we’re already at the top, or it may be because people on the ‘more tax please’ side of the debate are so determined to shout down anything said by their ‘opposition’ that they’ve blinded themselves with their own rhetoric and so cannot see that the curve itself (setting the shame of it to one side) should be their friend because it identifies (I repeat, for clarity) the point where the MOST TAX POSSIBLE is raised.

    Isn’t that what you want?

  15. But the latest attempt to revive it as a mainstream tool for political engineering

    You are confusing (quelle surprise) the “Laffer curve” with “Arthur Laffer”. Read the bloody WSJ article. It doesn’t mention the Laffer curve or tax rate optimisation at all. It is a simple attack on state income taxes. Which, frankly, are likely to be more progressive than state sales taxes and almost certainly have lower deadweight costs and would be supported by most here given those alternatives. Although getting the LVT monomaniacs to stop insisting that “that’s the way we do it” might take a bit of effort.

    get over it, it can never be used for anything other than to promote forever lowering taxes until all the poor die and Ayn Rand gets canonised.

    You see – the dichotomy between economics and morality. Everything that doesn’t fit into your narrow moral constraints, however correct a description of reality it is, is evil and must be expunged from public debate.

    You are, quite simply, a fool.

  16. @ SE 13
    That may be true in theory (OK – it is true in theory), but in practice I see dozens of company/LLP accounts with loans to directors or “directors’/partners’ drawings” that are interest-free.

  17. @ Arnald
    Perhaps it would help if you actually read the original source material before venting your spleen on young Tim.
    The purpose of these contrivances was to disguise the effective level of pay in the higher reaches of the public sector so as to conceal the extent to which public sector pay exceeded pay in comparable jobs in the private sector. This *is* political engineering.

  18. in practice I see dozens of company/LLP accounts with loans to directors or “directors’/partners’ drawings” that are interest-free.

    Yes – but it is a taxable benefit to the individual and should be reported on the P11D. “Partner’s drawings”, on the other hand, aren’t legally a loan? More of an advance on income earned but not yet due? And are taxed as income from business profits (with income tax and Class 4 NICs).

  19. Pendantry alert: “by basic graph theory” – no! The Laffer curve is a result from differential calculus.Graph theory is something completely different. It exists as a syllogism thus:

    If:
    1) tax rates of 0% and 100% raise no revenue
    2) there exists a rate r, 0% < r < 100% where revenue is positive
    3) revenue as a function of tax rate in the interval [0,100%] is continuous and differentiable

    then:
    there exists some rate r_m, 0 < r_m < 100% where revenue attains a maximum

    If the assumptions hold, then it is no more than a statement of Rolle’s theorem, which is itself a special case of the intermediate value theorem. Given these assumptions, it can no more be ‘debunked’ than can the Fundamental Theorem of calculus itself. If they do not, however, then it is not applicable. This seems to be what the Murphys of this world are asserting.

  20. David,

    I appreciate that 3) is necessary for calculus but I don’t believe it is necessary in this instant. If tax revenue is discontinuous – the function rose or dropped instantaneously on a (any) delta change, then it would make calculating the differential function “interesting”. But it wouldn’t invalidate the underlying point.

    At some point in the system there will be one or more tax points that raise maximum revenue.

    I believe that the Murphrist heresy is arguing that this tends to maximum as the marginal rate tends to 100%, although most of them (with a drink or two inside them) will agree that actually at 100%, it is 0.

  21. @ SE 13 et seq
    In several cases there is a note that the loan had been repaid since balance sheet date – so if loans for less than 9 months post balance sheet are exempt this is tax avoidance rather than outright evasion.

  22. @John77,

    Loans for less than 9 months post balance sheet are exempt the 25% (but recoverable) company payable tax but not the employee payable “benefit in kind” tax.

  23. Hugo and ChrisM
    I think you will find that “Arnald” is not only Lawrence Aegerter, but is also surely “Fast Robert” in the Guardian thread?

  24. Beyond the Laffer maximum revenue shrinks because the economy shrinks. This unavoidably means that at the Laffer maximum growth is zero. Funny how nobody mentions that.

  25. Roue le Jour

    We probably take it as read that beyond the turning point it isn’t worth doing more work, so the economy is bound to shrink as productivity falls. At 100% taxation nobody would bother to work at all.

  26. Frances Coppola

    My point is that you can’t both tax at the maximum and have growth. Since we want growth, the the maximum is something to be avoided, not targeted.

    Tim adds: Umm, yes….but growth will only disappear if all of the economy is taxed at that Laffer Maximum. And no one’s really arguing that at all.

  27. Excellent pendantry David, the mathematician in me agrees (at least until I figure out how he managed to get rid of the gag). However in the real world I think those assumptions are beyond dispute.

    1) tax rates of 0% and 100% raise no revenue

    The 0% case should be self-evident, especially as we’re considering a single tax only. The 100% point is the weakest part of the argument, but requires logical contortions to dispute. You could argue that, for example, unpaid community service is effectively taxed at 100%, but that’s veering into sophistry.

    2) there exists a rate r, 0% < r < 100% where revenue is positive

    This isn’t necessary – if all points between 0% and 100% give negative revenue, then the maximum revenue points are at 0% and 100%.

    3) revenue as a function of tax rate in the interval [0,100%] is continuous and differentiable

    In the strict mathematical sense this is a necessary assumption, but the human scale world is inevitably continuous and differentiable. I submit that if you’re measuring the economy with sufficient precision to detect quantum effects, you have bigger problems than worrying about the Laffer curve. Even a discontinuity in real life is continuous – the sampling points in the tax domain aren’t infinitely close together so the slope will be extremely large but finite.

    You could say that every sample represents a point where the first derivative is discontinuous and therefore undifferentiable, but that’s realistic only if you have infinite tax bandwidth. The fact that I can’t believe I just typed “tax bandwidth” should say how likely that is.

    I like to distinguish between formal proof and proof to the point where if someone still doesn’t get it one is forced to hit them with the nearest heavy object for the good of humanity. I think assumptions 1) and 3) (with the 100% point caveat) are sufficiently proved for any useful case.

  28. Roue le Jour

    Oh, I see – if tax is maximised all production increase is taken in tax, so growth is zero. Really I suppose therefore you need NGDP targeting in order to set optimal taxation levels, rather than assuming that maximum=best.

  29. But we’ve shifted from arguing about the mathematical necessity of the existence of the Laffer maximum to arguing a moral point: growth is good.

    While I happily agree with that, even to the extent of saying that setting our economic parameters to maximise long-term growth is good, that is arguable. Well, it is certainly argued. Which, I’ll admit, isn’t quite the same thing. Most of us would also suggest that the minimum tax-rate required to support the necessary functions of government is a good thing, as is taking that tax in ways with the lowest deadweight costs (or the lowest impact on long-term growth). Again, not quite the same thing.

    And very much argued by those who believe in sin, green and overtly (even penaly) progressive taxation.

  30. Tim adds: Umm, yes….but growth will only disappear if all of the economy is taxed at that Laffer Maximum. And no one’s really arguing that at all.

    Surely there are different Laffer maxima for different taxes. And, in fact, if you have a progressive tax system, you will have Laffer maxima for each of your different tax rates. Which are unlikely to be the same.

    I suppose there must also be a Laffer maximum for the overall tax rate of a country. By the mathematical derivations above. But that would be a dreadful difficult graph to derive. And very dependent on exactly those components used to make it up?

    Tim adds: Find out empirically. Survey countries and tax as portion of GDP and run a regression on growth rates. You’d probably want to do three different studies. One of total tax take (inc. borrowing of course), one of State purchases of goods and services and one of simple redistribution.

  31. Jamie, it can’t be him.

    ‘I am Lawrence
    Please like my music
    It really is quite easy to play
    And I’m a good bloke.’

    Arnald is not a good bloke.

Leave a Reply

Name and email are required. Your email address will not be published.