I was asked yesterday if it was true there was such a thing as the Laffer effect - that if taxes in the UK are increased then tax revenues would fall.
I replied that of course there is, very crudely a Laffer effect - if taxes ar 0% or 100% on all income then there's no tax revenue. In between there are revenues, so in a sense there's a Laffer curve. But that's not the argument. The argument is whether or not we're at the point where cutting taxes would raise revenues. Kevin Drum raised this on Mather Jones a week or two back, and in reply to my questioner I offer some of his commentary as explanation as to why the argument that cutting taxes would help raise revenue now is wrong:
Okay, this isn't actually raw data. In fact, it's very, very cooked and calculated data. But just so you know, Peter Diamond and Emmanuel Saez have tried to calculate the tax rate on the rich that would maximize revenue to the government. Paul Krugman summarises:
In the first part of the paper, D&S analyze the optimal tax rate on top earners. And they argue that this should be the rate that maximizes the revenue collected from these top earners–full stop. Why? Because if you're trying to maximize any sort of aggregate welfare measure, it's clear that a marginal dollar of income makes very little difference to the welfare of the wealthy, as compared with the difference it makes to the welfare of the poor and middle class. So to a first approximation policy should soak the rich for the maximum amount–not out of envy or a desire to punish, but simply to raise as much money as possible for other purposes.
Now, this doesn't imply a 100% tax rate, because there are going to be behavioral responses–high earners will generate at least somewhat less taxable income in the face of a high tax rate, either by actually working less or by pushing their earnings
underground. Using parameters based on the literature, D&S suggest that the optimal tax rate on the highest earners is in the vicinity of 70%.
Actually, Krugman is being conservative here. If you assume a broad base and no deductions, Diamond and Saez peg the revenue maximizing rate for top earners at 76 percent. That's for federal income tax only. (See page 173 here.)
You can decide for yourself if you think top marginal rates should be that high. After all, revenue maximization isn't our only social goal. Roughly speaking, though, this is a calculation of the peak of the famous Laffer Curve. (For top earners, anyway.) Above 76 percent, you really can generate higher revenues by lowering tax rates. Below that, higher rates generate higher revenue, just like you'd think.
I but that: it accords with the world as I observe it.
So for all practical purposes the answer to the Laffer question of 'will we increase taxes by cutting rates?' is a simply and responding 'No'.
NB Can't remember who sent me the link - but thanks
Thanks for reading this post.
You can share this post on social media of your choice by clicking these icons:
You can subscribe to this blog's daily email here.
And if you would like to support this blog you can, here:
Buy me a coffee!
I don’t think that for a marginal tax rate of 100% there would be no tax revenue. People don’t only work for money.You could do that in a major war and patriotic people would still work hard. The Soviet Union kind of managed it. Not that that is any kind of recommendation – just pointing to a flaw in the underlying Laffer logic. At 100% marginal rate, you would still get something.
Also, the TJN blog pointed out another useful article in this broad area, also looking at Laffer and taxing the wealthy. http://taxjustice.blogspot.com/2011/12/barney-frank-farewell-in-2012.html
And for a good laff about Laffer, read this http://www.newstatesman.com/books/2007/11/supply-side-economic-tax-rich
@ Nicholas Shaxson
You note:-
“At 100% marginal rate, you would still get something.”
And imagine the laughter on the Isle of Man as devious, lawyers, accountants and bankers invent even more malicious tax dodging gizmo to ensure that the Treasury still “got nothing.”
PS: Do you have a “Fan Club”?
You’re correct. In a command economies they worked because it was unlawful to be unemployed.
Exactly. As I said, it’s not a recommendation – just a demonstration that the Laffer logic is flawed from the outset. The war is the better example, I feel.
You are right
What all this shows is Laffer has as much use as a bag of damp crisps.
If we can’t agree how to calculate the tax impact – but it’s obvious that we are not at the pioint where there is an impact we can safely ignore the claimed effect
Back to position A then
im a bit confused – i thought from previous posts you didnt agree the laffer curve existed, but now it supports an increase in tax to 70% you suddenly agree with it?
I’ve not said it does not exist in a crude form
I have said the analysis is irrelevant for all practical purposes – as this data show
Of course, that’s for the citizens of “the land of the free” where you are taxed wherever you live in the world. Would the figure be lower outside the US where you can just go and live elsewhere?
this high tax will hurt our “job creators” to borrow a term from fox news
I’m just thinking about how this could apply to the UK. The paper you source to is clear on Page 162 footnote that the applicable optimal top rate of tax includes Federal, Medicaid, Sales tax, etc [1].
So as I come to translate very loosely to the top tax rate in the UK, it would appear I should add up:
1) Income tax of 50% (to compare to the US federal income tax)
2) National Insurance of 11% (to compare to US Medicare)
3) VAT of say 15%? (to compare to US sales tax) [2]
That’s pretty darn close to the 76% number we started with, and I’ve deliberately not included the more contentious employers NI number.
And yes, the paper is absolutely clear it is referring to the “top rate” of tax, which is why we start with the 50% Income tax number.
[1] “The top tax rate Ï„ is 42.5 percent for ordinary labor income when combining the top federal individual tax rate of 35 percent, uncapped Medicare taxes of 2.9 percent, and an average combined state top income tax rate of 5.86 percent and average sales tax rate of 2.32 percent.
[2] I’m guessing here. Headline is more obviously, but we have exemptions. Not sure of the net impact. Happy to be corrected.
Employer NI is not included deliberately and at 50% tax rate top rate NI is 1%
So your logic is way out
I stand corrected on the employee 1% point, but given that the the US numbers include the Medicaid payments which are made by employers and are uncapped, our own equivalent is employers NI.
I am not commenting on where the incidence lies, simply creating a like-for-like comparison. If they include the employers contribution, then for comparison purposes we should to.
That would give us:
Income: 50%
VAT: c.15%?
Employee NI: 1%
Employer NI: c.12%?
Total: c.78%
Assuming of course no ‘tax externalities’ which the authors accept would reduce this optimal top rate.
Howard Reed actually has it right
Your analysis is miles out
You can’t add in VAT on cash not spent for a start
Actually employees NI is 2% above 42,475?
And at both my current and previous employer, I am part of a profit share scheme, where we pay the employers NI of 13.8% out of our profit share before we get it, so it is a tax on my income. So my maximum marginal rate could be over 60%.
Also, having gone through the research, the finding was that under the current taxable income base, the optimal tax rate was 54%. The figure of 76% was based on the elasticity of taxable income before deductions, which doesn’t make sense given people would look at their taxable income after deductions.
Another issue springs to mind, which would be the behaviour of high earners who have pricing power with their employers. A CEO netting $5m a year may simply push their gross salary up so that their net remains $5m, with the result that the cost of the increased tax passes on to the consumer and into inflation.
I think most people have no clue what their tax rate is, let alone their deductions
If they had accountants would be out of business
All evidence shows I’m right on this one
Actually in the current year (11/12) NI rates are 12% and 2%! So you would pay 12% on earnings up to about £42500pa and 2% on earnings above that figure.
Just being picky! 🙂
You’re right
And I was distracted!
But our taxes include services like pensions not in US tax, NIC is not Medicare and you’re just adding percentages, as already show, inappropriately
That’s not really what the paper says, though.
The relevant chunk says:
Citing 76% assumes that we can get rid of all tax deductibles and broaden the tax base. With tax deductibles, the federal best rate is 48%.
And we can’t simply pull that over to the UK, either. They posit that payroll taxes (medicare at 2.9%) and consumption taxes (sales tax at 2.32% average cited) are taken into account, which accounts for the delta between the 54 or 80 total rate and 48 or 76 federal marginal rate suggested.
We can have a look over here (albeit elasticity will be different). Assuming their methods and very similar elasticity, of every £100 marginal increase, £12.13 in employers NI, £2 in employees NI, £50 in top rate tax get deducted.
Of the remaining £35.87, the amount gone on VAT depends on how much is spent on VAT-able goods and services. This will range from none to all of it. If all, that’s 35.87 * (1- (1/1.2)) = £5.98.
So your residual marginal increase is £29.89 – £35.87. For an effective marginal tax rate (on their methodology) of between 64.13% and 70.11%.
That’s well above the 54% peak for a tax base which has tax-deductibles. To get 54%, you need a marginal headline top rate of between 30.67% and 39.9%
That is, if this data and methodology are seen as valid.
You have articulated and quantified far more rigourosly the same idea which I was stabbing at in my own comment (which I must have been writing at the same time).
I defer to your superior analysis and agree with your conclusion. But one thing – where does 54% come from? You’re trying to get back to the 48% surely?
Thanks. In the paper (directly linked here, the total marginal tax rate is what to aim for. Which is 80% if all deductibles are abolished; 54% if not.
This equates to 76% federal tax or 48% federal tax in the US setup.
From Howard Reed’s contribution below, he cites a 55% expenditure rate on VAT-ables, so that narrows down the specific number in the range I came up with (30.67%-39.9%) of 35.23%
And my point is a simple one: the same authors say rates of 80% are possible in Europe, and presumably aren’t lying and second that’s because of a different tax / cost of living structure
The first two paragraphs in the Europe article say it all, pre-tax income of the richest 1% has INCREASED at the same time as tax rate on the highest earners have come DOWN significantly. Surely that is all the evidence of the existence of the Laffer curve that we need…
Note they say also that these may be unrelated
There are many more factors than tax
Their increased ability to capture rents for their own gain (using the Marxist understanding of rent here) is much, much more significant
In other words – the ‘old boy’s club’ and the rise of the exploitation of the City explains this and has nothing to do with tax – except that the same lobbying that let them win the income let them reduce the tax
Your analysis is so simplistic it is ridiculous
Update: I understand a little more now. He’s not actually asking for 76% Federal Income tax per se; rather he’s advocating 80% in total, then assumes constant Sales and Medicaid tax rates, which gives him his 76%, but the 80% could come in any combination really.
But the upshot is he thinks the right number is 80%, and our UK comparison should be with that number, by adding Income, VAT, employee NI and employer NI. Which does still get us to around his number (or perhaps more depending on how you want to apportion employers NI).
Again, I refer to Howard Read
It’s amazing how many people on this blog seem to think that 12% employee NICs applies all the way up the income distribution… it’s 2% above the upper earnings limit, which is way below the 50% threshold. The relevant statistic would seem to be 50% +2% employee NI + whatever percentage of the remaining 48% of income is subject to VAT, which (according to HMRC statistics) is about 55 percent of expenditure. The VAT rate is 20/120= 16.7 percent approx. This gives 50 + 2 + (16.7*0.55) = 61 percent. Way below the 70 percent figure Paul Krugman suggests, let alone the 76% max figure in the Diamond and Saez paper.
If the Diamond/Saez figures carry over to the UK it suggests that the 75% top income tax rate introduced in 1973 under Ted Heath may have been about right for revenue maximisation (given much lower VAT then), whereas the 83% top rate which Labour introduced in 1975 may in fact have been too high to maximise revenue.
Thanks Howard
I was somewhat distracted by doing my grade 3 clarinet exam in the last couple of hours!
Your analysis is spot on
Not the employee NICs but the employer NICs. For which there is no cap. As Gary points out, the employer component of payroll taxation is included in their methodology.
I’ve assumed 2% employee NICs.
If 55% of expenditure is subject to VAT, then 35.23% marginal top rate equates to 54% of total marginal taxation (the first number – assuming the tax regime continues to have tax-deductible items) and 63.86% to the 80% (which assumes all tax-deductibles are abolished).
Howard, the 76% figure is based assumed elasticity of income before deductions, which doesn’t seem particlularly useful as people would consider the tax impact after deductions in making decisions. As Richard points out, most people don’t know the tax rate they pay, they only look at the net rate after dedcuations and this determines behaviour. Using the actual tax base comes up with the maximum tax rate of 54%.
Nonsense!
You can’t have it both ways!
Why would you include employers health contribution in the US but exclude it in UK? That’s just not consistent.
Because they have fundamentally different purposes
And pay for fundamentally different things
Not least because to compare like with like US employer’s contribution to private medical plans would have to be added to ensure consistency too as the systems are very different – and so are the taxes
Judgment is allowed
You, I think used it incorrectly
we may differ and I’ll leave it at that
The purpose in both instances is to pay for healthcare. Fundamentally the same purpose. Different providers, different ethos, different views on efficancy and efficiency. But both are simply payments for healthcare paid for by the employer.
Including employer healthcare payments in one country but excluding it in another isn’t judgement. Its just inconsistent.
Absolutely wrong
NIC does not pay for healthcare
It pays for pensions
Almost none goes to the NHS
So actually NIC should be out on this basis
I’ll tolerate ‘ee in but ‘er not at all
Howard, the Diamond/Saez research makes no distinction as to the nature or purpose of the tax; is simply seeks to establish the revenue maximising amount of tax, and so includes all taxes whether paid by employee or employer. That’s their judgement and so must be the basis for a UK comparison.
If you choose to override their analysis by omitting some UK taxes, it would be equally valid for me to claim that we should exclude Federal Taxes from the US number to reach a lower optimal top rate. That would be nonesense of course, but it simply illustrates the invalidity of accepting their judgement in one country, not accepting it in another and then claiming consistency/applicability.
In short, your 61% includes some but not all tax, and needs to be uplifted to include NI before you can compare to the US number (in a hypothetical non-deductable world).
But you assume all tax is ‘lost’ to the payer
That’s absurd: purpose does matter and changes behaviour
And like for like therefore need be compared
Sorry – but you’re making this up as you go along
I’m not making any assumptions at all. What I’m actually doing is looking at what Diamond/Saez did which was add up *all* taxes for whatever purpose, and total them. To be comparable we in the UK should do the same. Indeed when Diamond/Saez turn to Europe they do not suggest excluding particular taxes, so they are alrady telling us *not* to exclude taxes.
Or put another way, you are saying Diamond/Saez are right in US, but wrong in Europe.
They have themselves said the answer in Europe is 80% – link above
Now either they’re right about Europe or you’re wrong
Which should I believe?
My preferred option would be to say that Diamond/Saez are right on all counts: 80% for Europe in a hypothetical non-deductable world, and that 80% includes all taxes as per the US calculations.
In which case we are logic-bound to include all taxes; Income, VAT, employers and employees NI, all adjusted for the reality that we don’t in fact live in a non-deductable world. But the arithmetic is not as pretty as you might have first thought…
Is this the old “supply side” argument that, put simply, suggests that giving people tax cuts rather than imposing tax rises gives people more money in their pockets to spend on goods and services and therefore better for the economy?
Have the supply sider’s ever considered what happens when a government grants a cut in income tax? How does it make up the billions it loses from this cut in tax? It has to balance the books somehow. They either cut services or spending, raise taxes indirectly, sell off a couple of national assets or they add it to the deficit.
Either way, you are paying for that tax cut one way or another.
Note that here the same authors suggest EU rates of 80%
http://www.voxeu.org/index.php?q=node/7402
Those arguing otherwise please take note
I not believe employer’s NIC can be added in here
The equivalent in the UK is employee’s NIC – a point made explicitly by Gordon Brown
Gordon can be right about many things, but can’t have forseen the paper you refer to which wasn’t yet written when he said that.
In the US, employers pay for Medicaid. That payment is included in the paper you quote. Here in the UK, our equivalent is Employers NI. therefore in order to draw a like-for-like comparison, we need to include it in our UK calculations.
I’m not disagreeing with Gordon Brown, just saying that including it here makes it consistent with the methodology you referred to an implicitly endorsed.
And I don’t agree
The systems or not the same – and so one has to be careful and selective
You are not being so, at all
For the sake of correctness, as the research on the Laffer curve is based on the US rates of taxes, the following apply in 2011:-
Federal Income tax – 35% maximum
State Income Tax – 7.5% (average – ranges from 0% in Florida to a maximum of 11% in Oregon and Hawaii)
Social Security(FICA) – 4.2% employees, 6.2% employers on earnings up to $106,800 pa
Medicare – 1.45% employees, 1.45% employers; no upper cut off
GST(VAT) – varies but 4.5% on average x 55% = 2.48%
This is at variance with some of the figures quoted in the responses above, particularly in that the medicare contribution is not 2.9% for employers but split 50/50 employee/employer.
Isn’t this NIC thing a bit of a red herring? At the highest levels of income, you will structure your affairs to get dividends rather than remuneration and avoid NIC altogether on the top slice? Of course, there are even more sophisticated arrangements as well.
Frnakly, I think you can just ignore NIC altogether when calculating the top rate.
James
No they don’t – not in the City etc
They’re employees
I agree so in the smaller business – but remarkably few of them earn £150k plus….
If you think about it, the vast majority of people who earn over £150k actually earn £150-250k. Occupationally that means top level GPs, Medical Consultants, Dentists, Accountants, Lawyers and Management Consultants. That’s the bulk of the 1%. But few of them are living in the world of non-dom/offshore/income splitting/capital income. Some exceptions, but its predominantly PAYE and NI all round.
The esoteric stuff isn’t in the 1%, its in the 0.1%.
I sent you the Diamond-Saez and Krugman links, but I’m not bothered about ‘hat tips’, which is why I’m pseudonymous. Rather, I was thinking about the OBR being told to assess the 50p tax rate on higher earners. Before the D-S paper Krugman had written on August 7, 2010 in ‘Tax Cuts And Spending (Very Wonkish)’ about tax cuts on the very rich, and had used Milton Friedman’s ideas to explain why Friedman’s would think that it was the worst of all ideas. This is why the right-wing think tanks in the US use Lawrence Lindsey’s 1987 paper looking as tax cuts in the US 1982-4 for their justification, not Friedman – a paper that peer review said was inconclusive as a single variable over such a short time was too difficult to track effectively.
Now, if you know Friedman’s ‘stagflation’ paper of 1968 he says that, “experience suggests that the path of wisdom is to use monetary policy explicitly to offset other disturbances only when they offer a ‘clear and present danger’.” That ‘clear and present danger’ line is being used by Osborne and Cameron all the time in their parliamentary speeches, and this is where it comes from (page 14, THE ROLE OF MONETARY POLICY, in the AER March 1968). Friedman in this paper and elsewhere called the Great Depression the ‘Great Contraction’, and so the paper that was used by the Tories to give rising unemployment in the UK 1980-7 is still being used by Cameron-Osborne today. This is why they are backing the ECB to become a proper central bank, as you can’t solve a great contraction without an instrument through which liquidity can be funnelled. But if you assume in the UK that ours is a Great Contraction then you must explain why adding to the monetary base failed in Japan 1999 and US 2008 (see Krugman, ‘Friedman on Japan’ and ‘More On Friedman/Japan’). This is why Adam Posen’s first paper as a member of the BoE was about explicitly about ‘non-Monetarist’ QE.
So, in short, when the OBR reports, the anti-Lindsey side needs to be heard and to be loud, and the battle against ‘clear and present danger’ politics in which the Medium Term Financial Strategy of 1980-3 gets locked together with 2010-present is very urgent. Monetarism must be seen to have failed and that QE is not monetarism. Ditto low taxes on higher earners.