It has become fashionable in the last few days to describe VAT as a “regressive”, and by implication unfair, tax. This is usually followed by complaint about how hypocritical it is of the Liberal Democrats to agree to an increase in its rate.
But VAT is not, by the simplest definition, a regressive tax. A regressive tax is one where the rate of taxation decreases as the value of the thing being taxed increases. A progressive tax is the other way round. Income tax is progressive, because those on higher incomes pay a higher rate of tax on it. Council tax is regressive, being in general a larger fraction of the value of a small home than a large one. But VAT is charged at the same rate on everything to which it applies, and so, by this definition, is neither regressive nor progressive.
A more complex definition of whether a tax is regressive would be whether the poor end up paying a higher proportion of their incomes on it than the rich. By this definition, VAT actually makes a reasonable case for being progressive. By making “essentials” such as food and rent exempt, the taxable items can be thought of as “luxuries” bought more by those with higher incomes. But it doesn’t quite work like that, and many of the things that even the poorest need to buy are subject to VAT: clothing, fuel, cleaning products. And many of the things that the rich do with their money- buying property, for example- do not incur VAT. So in practice, it is possible that VAT is in fact a regressive tax.
The solution would be to look again at what items should exempt from VAT, and a good starting point would the Rowntree Foundation’s excellent report from a few years ago on what exactly is an “essential” in modern society. Making all of these things exempt from VAT would ensure that it is no longer a regressive tax. This would be expensive, but perhaps that is the price of fairness.
Follow @LibDemVoice on Twitter
Like us on Facebook
Subscribe to our feed
Sign-up for our daily email digest
My Parish may be small, but it is, apparently, mighty...
Latest PollBase is out, with all the British voting intention polls since 1943
Temporary traffic order - Menzieshill Road
The very lost tunnels of the Hereford and Gloucester Canal
Are some premature births caused by a average lower mitochondrial membrane potential in the baby?
John Hills
Tristan Ward
Peter Martin
55 Comments
Those on lower incomes spend more of their income, because they need to purchase the essentials. The rich have a lower propensity to consume; they save more, so the poor pay more VAT as a percentage of their income. It is a regressive tax, plain and simple.
Essentials like food and kid’s clothes are exempt anyway or do TVs and games consoles count as essentials now like that family featured on BBC website who live entirely on benefits and spent two thousand pounds on Christmas presents for two very small children?
Essentials in the sense that they are part of a good life, yes. You don’t have to be on benefit to be poor. You can only buy so many TVs, so the rich do not spend as much as the poor as a % of their income. So it hits the poor hardest.
Some confusion here over exempt and zero-rated.
I don’t think that the EU allows states to cut VAT that is already 5% or over to below 5%.
Tony Greaves
I think people commenting on kids clothes being exempt should look at what constitutes as ‘kids clothes’ first . A good example of this is childrens shoes, up to size 3 is VAT exempt. How many under 16 year olds wear a size 3 or less? The rules to what is exempted from VAT are very complex, and to make things worse when shopping there seems to be clear defination of what you don’t pay VAT on.
That should have been * be clear no defination/guide* :/
“A more complex definition of whether a tax is regressive would be whether the poor end up paying a higher proportion of their incomes on it than the rich. By this definition, VAT actually makes a reasonable case for being progressive. … So in practice, it is possible that VAT is in fact a regressive tax.”
Quite astonishing.
Malcolm Wood could have saved himself this half-baked, arm-waving speculation by simply looking at the IFS calculations on the effect of the VAT rise on different income groups:
http://www.ifs.org.uk/budgets/budgetjune2010/browne.pdf
The results are exactly as would be expected. The group that loses most as a percentage of income (2.25%) is the 10% with the lowest incomes. Second worst hit is the 10% with the next lowest incomes. Third worst hit is the 10% after that. And – surprise, surprise! – the 10% with the highest incomes will loses least as a percentage of income (less than 1%).
Quite clearly the VAT rise is regressive in the sense that “the poor end up paying a higher proportion of their incomes on it than the rich”.
Quote: “Essentials like food and kid’s clothes are exempt anyway or do TVs and games consoles count as essentials now like that family featured on BBC website who live entirely on benefits and spent two thousand pounds on Christmas presents for two very small children?”
Oh dear, I appear to have stumbled upon the Daily Mail website by mistake – anecdotes in place of analysis. Silly me.
Can someone answer this for me regarding the VAT rise?
As the cost of shipping items will increase will shops pass that cost onto the consumer?
I want to have Anthony’s babies.
Considering your Party was predicting doom-laden Tory rises when it was in virtuous opposition or genuinely expecting triple figure seat numbers, I don’t think we have to look too far for the fashion-setters.
Oooo, look! I wonder if we’ll have a repeat of this utterly graceless article which left a US Congressional candidate appalled.
Yeah, Peter, clearly Andrea is a better class of person as she deserves her computer.
Oh dear. VAT is the boring tax that no one wants to understand!
Given that food, kids clothing and reading materials are zero rated, and utilities are only taxed at 5%, I can’t see how it is regressive. If anything, it’s an attack on the consumer, un-ecological lifestyle that we all want to enjoy, getting in to heaps of debt on hire-purchase if necessary.
“Oh dear. VAT is the boring tax that no one wants to understand!
Given that food, kids clothing and reading materials are zero rated, and utilities are only taxed at 5%, I can’t see how it is regressive.”
Well, if you can be bothered to read the comments above, you’ll see that it clearly _is_ regressive, whether you can understand it or not!
@Rob
“A regressive tax is a tax imposed in such a manner that the tax rate decreases as the amount subject to taxation increases.”
Surely, given that VAT is a tax on expenditure, NOT on income, it is progressive in relation to expenditure, and regressive in relation to income? Thus given that the tax is applied to expenditure, your definition would describe it as progressive – the amount subject to taxation will tend to include more luxuries?
There didn’t seem to be much analysis about the other regressive taxation implemented in the budget, such as the increase in cigarette, fuel and alcohol duty, all of which impact more on the poor than the rich.
There doesn’t seem a great deal of point expending a lot more time debating whether VAT is regressive. While the effect can be reduced by exemptions (and there is much to be said for JRF arguments about changing the structure of exemptions so they are more clearly focused), a flat sales tax is a textbook example of a regressive tax, for all the reasons previous posters have identified.
The other element to this story is that combining a sales tax increase with freezes in benefits and/or upgrading key benefits in line with CPI not RPI means the squeeze will get progressively worse over time for many households.
Add in the swingeing cuts to public services and any attempt to claim the overall picture is anywhere near progressive feels slightly surreal. (And surely it is the overall picture that is much more important than individual policy changes.) It might be that fiscal retrenchment is unavoidable, but pretending that “we’re all in it together” and that the poor will not be hit disproportionately is pointless because it is so lacking in credibility.
Anthony Aloysius St said: “The results are exactly as would be expected. The group that loses most as a percentage of income (2.25%) is the 10% with the lowest incomes … Quite clearly the VAT rise is regressive”
You’re right to quote the IFS. They have justifiably established a reputation for independence and objectivity.
But you’re misunderstanding the IFS position. Have a look at this article: http://www.ifs.org.uk/budgets/gb2009/09chap10.pdf
In that article, the IFS couldn’t be clearer when they say: “Myth 2: VAT is a regressive form of taxation”.
You quote the statistic that the poorest pay a higher proportion of income on VAT.
The IFS point out that this statistic is misleading, because of those on low incomes who have are living off savings – the retired living off savings, and those on very temporary unemployment.
To measure how regressive VAT is, they say we should look at VAT as a proportion of expenditure. If you do, they provide a chart (Figure 10.2) which shows VAT is essentially neither regressive nor progressive.
Obviously, it’s more complicated than that. Have a read of the article if you want more details of their argument.
George
“you’re misunderstanding the IFS position”
Eh? I didn’t say anything about “the IFS position”. I simply quoted the IFS data, which show that those on lower incomes will lose a larger percentage of their incomes through the VAT increase than those on higher incomes. (And as you can see above, Malcolm Wood’s article is clearly written with the assumption that income is the appropriate way of classifying people when considering this question.)
Of course I’m aware of the argument that it’s more appropriate to classify people by expenditure, which is also mentioned in the presentation I cited. But on the whole I think low income is a more appropriate measure of poverty than low expenditure.
However, in the context of discussions about the fairness of this budget the point is academic, because the overall effect of the tax and benefit changes will be that both those on low incomes and those with low expenditure will be among the hardest hit.
For that matter, though, the argument about VAT and the other tax and benefit changes is fairly academic, bearing in mind the punitive effects that general cuts in public spending are likely to have on the poor. As pointed out on another thread, a study has concluded that the overall effect will be equivalent to more than 20% of the income of the poorest families.
Anthony Aloysius St “on the whole I think low income is a more appropriate measure of poverty than low expenditure”
The IFS take a different view, as do I.
Take an extreme example: Someone has a £40,000 job, loses it, but as they have substantial savings and fully expect to be employed within a couple of months, they carry on with their previous spending. They clearly aren’t “poor”, but they’d be in the bottom decile of income. According to the IFS, there are enough people like this to significantly distort the statistics that suggest that VAT is regressive. Which is why it’s better to measure expenditure if we want to work out if VAT is regressive or not.
Of course, my comments here are a simplification of the IFS argument. If anyone wants to debate this, it’d be great if you could read the IFS report [ http://www.ifs.org.uk/budgets/gb2009/09chap10.pdf ], and respond to that, rather than my, inevitably, superficial summary.
I admire the IFS. They are no creature of the right or the left. Their recent report, which you refer to, argues that big spending cuts will hurt the poor a lot more than the specific changes announced in the budget. This makes hard reading for LibDems like myself.
But that’s an issue for another thread. With regard to VAT, much though it goes against received wisdom, Malcolm Wood’s main point is correct: while it isn’t progressive, it isn’t regressive either.
It also depends how you define poor. I know when I was poor I did not have a tv, hoover, any electrical goods, maybe a cheap sound system. All I had to buy was food and a gas bottle for the cooker every 3 mounths. I look back now and think why was that the best time of my life? It was simple, nothing much to worry about. I had brought a caravan of tha dids for £20 (The first home of my own). It was easier to sell it to one of us than scrap it for the alluminium themselves. Everything elsewent on my bike, booze and other substances. Very little went on VAT. Moral: When you don’t have much money you find a way, you work more with the people you live around making the community really really close. In that kind of group you don’t need money for fun, you have the people around you, time of my life. Never be frightened of being rextremly poor. It is good for your mental health.
George
Obviously my comments were made in response to the article by Malcolm Wood above, which explicitly refers to income, not expenditure, as the appropriate measure of wealth/poverty, and which says “A more complex definition of whether a tax is regressive would be whether the poor end up paying a higher proportion of their incomes on it than the rich”. What I was pointing out is that in those terms the data indicate quite clearly that VAT _is_ regressive (as, in fact, Wood acknowledges at the end of his piece).
Whether it’s regressive by other definitions is a diffierent question. I am not convinced by the alternative method of grouping households by expenditure. Unless I am misinterpreting things, in the lowest expenditure decile, on average income is something approaching three times expenditure. In the second decile, income is about twice expenditure. In the third decile, 1.5 times expenditure. So I don’t believe this grouping is really isolating poverty at all. These groups must contain substantial number of people with medium or even quite large incomes who don’t spend very much. But maybe I am misinterpreting things somehow. In a way, I hope I am.
In any case, as I’ve said, this is largely an academic argument, because the IFS figures indicate that whichever method you use, the lowest decile is among the hardest hit overall by the tax and benefit changes in the budget.
“I am not convinced by the alternative method of grouping households by expenditure.”
Nor am I convinced by the comparison between the groups of gains and losses as a percentage of expenditure rather than income. If you take away 10% of the income of a poor man who has to spend the whole of his income to get by, that is very different from taking away 1% of the income of a rich man who is saving 90% of his income.
@Anthony, thanks for the IFS link. That would indeed have saved me some handwaving. Since reading a bit more, and talking to more people, I’ve been pretty much convinced by the argument that VAT is, in practice, regressive relative to income. And the point of my article was mainly to propose an approach to solving that problem, though the comments so far haven’t really addressed that. An interesting debate, though.
That’s on page 9. On the very next page of the document you cited, it explains why this is the incorrect thing to measure, and on the page after that, it rearranges the data based on expenditure rather than income, and concludes that VAT is progressive. Their statements on this subject are clear and unmistakable.
Now, are we supposed to assume that you stopped reading the document in the middle? Or that you found the words on page 10 too hard to understand? Or is it more likely that you understood this very simple document, and decided to lie about it because it didn’t support your position?
Andrew
“Now, are we supposed to assume that you stopped reading the document in the middle? Or that you found the words on page 10 too hard to understand? Or is it more likely that you understood this very simple document, and decided to lie about it because it didn’t support your position?”
????
I’ve just explained precisely why I don’t accept the the basis on which the IFS concluded that VAT was progressive!
Had you not read my comments? If you hadn’t, you might have had the courtesy to do so before attacking me. If you had, then it’s pretty clear that _I’m_ not the one being dishonest here.
If you have any interest in a sensible discussion, why don’t you address the points I made, rather than dancing around like a small child, yelling “Liar, liar”?
“That’s on page 9. On the very next page of the document you cited, it explains why this is the incorrect thing to measure, and on the page after that, it rearranges the data based on expenditure rather than income, and concludes that VAT is progressive. Their statements on this subject are clear and unmistakable.”
And just to correct _your_ misrepresentation of what that document says.
It says, first, that it is true that VAT is regressive “if we measure household incomes at a point in time and express losses from VAT rise as a percentage of net income”. Which, at the risk of repeating myself, is the definition used in the part of Malcolm Wood’s article that I quoted and to which I was responding.
Nowhere does it say that is “the incorrect thing to measure”, as you claim. It says that expressing losses as a percentage of expenditure “gives a very different answer” (which is no more than a statement of the obvious), and it says that total expenditure is _arguably_ a better guide to lifetime living standards than income. “Arguably” obviously means the point is arguable!
And, finally, nowhere does that document describe VAT as “progressive”, as you claim.
Malcolm
“@Anthony, thanks for the IFS link. That would indeed have saved me some handwaving.”
I’m sorry, my comments were too harsh. It’s just that I’ve seen similar handwaving time and time again in these discussions of the budget – usually deployed with the intention of defending the rise in VAT. That’s a bit frustrating when the relevant data are readily available.
Page 11.
Page 10.
Interested readers are strongly encouraged to read the document rather than the troll’s comments.
As I said, Andrew Suffield’s statements about that document are highly misleading.
I have already quoted exactly what is said on page 10 – that expenditure is _arguably_ “a better guide to lifetime living standards”. Saying something is “arguable” is obviously not the same as saying it is true and that alternative viewpoints are incorrect.
And as for page 11, which is claimed to contain the IFS’s “conclusion”, that “VAT is progressive”. Not only does it not say that, it doesn’t say anything at all. It’s simply a chart without any accompanying text.
Of course, as George has pointed out, another IFS study – which I hadn’t seen when I commented above – does reach such conclusions (not that I agree they are soundly based, for the reasons mentioned above). But the document I quoted is rather more even-handed, which suggests to me that the IFS doesn’t have the kind of official line on this that George was suggesting. But in any case, I hope none of us is going to accept anyone’s _interpretation_ without having a good look at the data for ourselves and thinking things through thoroughly.
The great irony of this, of course, is that if we do adopt the alternative method of grouping households by expenditure – which, for the reasons I stated above, I think is at least as problematical as grouping them by income – then on the IFS’s figures the poorest 10% of households are actually the hardest hit by the tax and benefit changes in the budget in 2012-13, and the second hardest hit in 2014-15.
The conclusion about the budget is much the same, whichever method is used. And the really sad thing is that these tax and benefit changes of a few percent are going to be tiny compared with the consequences of the general cuts in public spending. No matter how abusive and dishonest people get, I don’t believe they are going to be able to defend these policies.
I seem to remember getting criticised the other day for saying that VAT was midly progresive, can’t quiet remember by who but a certain name feels familiar. But reading the article it certainly appears to say , in these circumstances that it is mildly progrresive. I was going by my own experience at the time. Now I have it from the IFS, thanks. Srange how some people only like these documents when it fits in with what they want.
@ Anthony. We have to defend the cuts in public spending, because they have to be done due to the mess labour left.
Richard
“But reading the article it certainly appears to say , in these circumstances that it is mildly progrresive.”
The other IFS article – the one cited by George – does say that. But in order to reach that conclusion, the authors:
(1) Group households according to their expenditure, not their income, and
(2) Express the cost of the VAT rise as a percentage of household expenditure, not household income.
I’ve already explained why I don’t think this procedure makes sense.
For one thing, the “poorest” group, according to that method, only spends about a third of its income, on average. Now honestly, does that really sound like poor people to you? Have you met many people you’d describe as poor, who can afford to save two-thirds of what they earn? The problem is that this group of the “poorest” obviously contains a large proportion of people with quite high incomes but very low expenditure.
For another thing, if you express everything as a percentage of expenditure, not as a percentage of income, that makes many measures look progressive, that would be viewed as very regressive in terms of percentages of income. As I said above, can you really view the following as equivalent:
(i) a cut of 10% in the income of a poor man who has to spend everything he earns and
(ii) a cut of 1% in the income of a rich man who saves 90% of what he earns?
Expressing things as a percentage of expenditure, rather than income, has that effect, and frankly I think it’s nonsensical.
@ anthony .I don’t agree with you that it is regresive bu tI definitly agree with you that the definition of poor is way out.
Richard
I think the problem is that there are difficulties with both methods – grouping households by income and grouping them by expenditure. Either way, you get people in the “poorest” categories who don’t really belong there, because for some reason there is a big discrepancy between their income and their expenditure.
It seems to me that – based on the IFS data – grouping by expenditure involves bigger divergences between income and expenditure in the lower groups, and is therefore more suspect.
But as I keep saying, whichever method you use, the poorest are going to be among the worst affected by these tax and benefit changes, taken as a whole. And on top of that there are going to be the enormous spending cuts, which are virtually guaranteed to hit the poorest hardest.
You can say “We have to defend the cuts in public spending, because they have to be done due to the mess labour left”, but it’s within the government’s power to determine where the burden will fall. As things stand, the burden will be falling disproportionately on the poorest. That’s something I don’t think Liberal Democrats should be defending.
I like the old one : if you spend 19s 6d ino every pound you are a rich man if spend £1 6d you are a poor man. I suppose a lot of you dn’t remember pounds, shillings and pence but I am sure you get the drift.
Poor to me is: You don’t have a washing machine you use your hands. You don’t have a fridge you walk to the shops most days, not needing transport. You don’t have a television you have friends instead who normally don”t have tvs either. You wear thermals and things in the cold weather , hence small fuel bills. Living like that there is hardly any VAT involved. I have lived like that for years at one time. Actualy that is quite a good life style and you get really close to your friends. So I suppose when you have less than that you are poor.
“You wear thermals and things in the cold weather , hence small fuel bills.”
Luxury! When I were a lad we used to wrap poisonous snakes round our thighs to try and keep warm, and when they bit us, Eric Pickles would dance about on our graves singing “Allelujah!”
And you try and tell the young people of today that …
If you actually read the article, the suggestion being made is that far more things are made exempt and therefore the proportion of income would be changed. There is a policy proposal here to exempt more things to ensure that it is progressive.
(Also, Malcolm is accepting a variety of definitions of progressive etc… and some of the critics on here seem to only be taking the version that best enables them to build a straw person argument.)
Henry
If you actually read the comments, you’ll see that – as I’ve already explained several times – what I was pointing out is that according to the second definition in the article, which the author wasn’t sure would make VAT regressive or not, the data clearly show that it _is_ regressive.
Obviously people are incredibly defensive about this issue. I suppose that’s not surprising.
Henry,
You’re quite right. Our discussion has missed the point of the main article.
While we can’t zero rate new items, we could 5% rate new items. (EU rules don’t allow us to reduce VAT below 5% for an item that is already VAT-rated) There is a debate to be had there, but I don’t think there’s any appetite for further complicated changes in the rules of VAT.
The IFS (aren’t we quoting from them a lot?), actually suggest the opposite, that it would be more progressive to remove zero-rating altogether, and use some of the money raised to increase benefits. They point out that this is how many European countries sustain high rates of VAT, that it would raise significant extra revenue, and that it would simplify the tax system. I’m not convinced, because for political reasons I suspect we’d see the ending of zero-rating, but there wouldn’t be the corresponding rise in benefits. But it’s an interesting point, and maybe in the current crisis, it’ll be considered.
Hi Anthony Aloysius St,
You asked on another thread for me to respond to your comments on the IFS methodology for analysing VAT, so here goes…
Any methodology is going to be an approximation.
You point out that, if you define the poor as those with low expenditure, that’ll include people with a larger income who are spending very little.
The IFS argue that, if you define the poor as those with low income, that’ll include people with low income, but substantial savings, who are spending a lot.
In my opinion, there aren’t many people with high incomes but expenditures in the bottom decile. I know some, who when faced with a substantial salary increase, aspired to keep their spending down. Unfortunately, despite the best of intentions, it didn’t happen.
The IFS say “many households with low incomes currently are not those we would normally think of as ‘poor’, e.g. pensioners living on savings, temporarily not working, those with volatile earnings”.
That sounds to me like a lot of people, and that it would significantly skew the statistics.
I’ve not done research on this, but the IFS have, and what they say seems extremely plausible. So, lacking any serious evidence to the contrary, I believe the IFS, and that the widely held belief that VAT is regressive is an urban myth.
This is a genuine change of mind. Not long ago, I’d have thought the opposite, but when faced with evidence that I was wrong, I’ve changed my mind.
George
“You point out that, if you define the poor as those with low expenditure, that’ll include people with a larger income who are spending very little.”
What I actually pointed out was that on average the income of households in that bottom decile appears to be _three_ times bigger than their expenditure. If you think about it, that doesn’t just mean that it includes some people with incomes larger than their expenditure. It means it includes _many_ people with _much_ larger incomes.
For example, one way in which the average income could be three times the expenditure would be if half the households were genuinely poor, spending all their income, and the other half had incomes _five_ times bigger than their expenditure. I would suggest that by any sensible definition, someone who can get by on only 20% of his income is pretty well off. And _half_ the households in that bottom decile would fall into that category, in the example I gave.
As I pointed out above, there are also large factors between average income and average expenditure in the second and third deciles of households grouped by expenditure. And I would note that these discrepancies are larger than in the grouping by income.
My impression of you from your comments above was that you were a fair-minded person who had studied the evidence in some depth and was basing his comments on that independent study (“The IFS take a different view, as do I”). I have to say that that impression has been rather shaken by your response that you haven’t studied the evidence yourself, but that “[you] believe the IFS”.
I’d ask you again – please consider the implications of those aspects of the data that I’ve quoted. Doesn’t the large number of evidently wealthy households in those lower deciles grouped by expenditure mean that the methodology suffers from flaws at least as serious as the more conventional grouping by income? Please reassure me that you _are_ being fair-minded and not just casting about for any evidence, no matter how suspect, to support the party line.
@Anthony Aloysius St,
Apologies, I think having two different IFS documents has had us talking at cross-purposes.
I’ve been basing my comments on the IFS report: http://www.ifs.org.uk/budgets/gb2009/09chap10.pdf. It has detailed text accompanying the charts of data, and I find it much more straightforward to understand than the pdf you’ve been referring to.
http://www.ifs.org.uk/budgets/budgetjune2010/browne.pdf looks like a power point presentation rushed out to respond to the budget. It lacks detailed explanatory text, and the graphical data on page 11 makes no sense to me. I’m sure it would have made a lot of sense if I’d heard the accompanying presentation, but without that, I’d be wasting everyone’s time with what might be a misreading of the data. And, if the document is a rushed job, it’s always possible the author made a mistake.
The 09Chapter10 pdf is a proper document, with detailed text and references. The graph that summarises the situation for me is on page 5, and that’s the one that’s driven my comments in this thread. It shows the data both by expenditure and by income deciles, and the conclusions is basically the same. The point the report makes is that the graph on page 4 is misleading.
George
The data I was referring to are in http://www.ifs.org.uk/budgets/budgetjune2010/browne.pdf . That is indeed a power-point presentation, but I would certainly hope that in itself is no reason to assume the data are incorrect. (As far as I can see, the other document you refer to doesn’t contain equivalent data.)
In case you were unable to work out what I was referring to, please look at page 11, where the impact of VAT is expressed in two different ways – as a percentage of income, and as a percentage of expenditure. Unless I am completely misinterpreting them, these data imply that in some sense average income is three times average expenditure for the bottom decile, grouped by expenditure.
Do you see some other interpretation? Or do you have some reason to believe the data presented by the IFS are incorrect?
If not, and if the average income really is three times the average expenditure in the bottom decile, is it really possible to view that group of households as “poor”, in any meaningful sense? And to conclude on that basis that VAT isn’t regressive? In fact, isn’t it evident that the methodology of grouping by expenditure rather than income is deeply flawed?
I have to say that I’m disappointed, though not altogether surprised, not to have received any response.
Sorry Anthony, but I don’t think I have much new to say.
I don’t understand the table on page 11 in the Browne.pdf powerpoint. The accompanying text, what there is of it, indicates it contradicts figure 10.2 in the 09chap10.pdf document, which makes me think that either I’ve completely misunderstood what it’s saying, or the author, in the rush to produce his post-budget Powerpoint slides, mislabelled it.
And the idea that the 10% of people who have the lowest standard of living have up to three times as much income as they are spending seems preposterous. Rich people may have very high saving rates, I don’t think people living on the breadline do.
And I can’t get too worked up about it, ‘cos it’s a rushed Powerpoint presentation, which was probably never intended to be analysed in this way.
09chap10.pdf is written as a serious document, I understand it, and it fits in with my personal observation of society around me.
“I don’t understand the table on page 11 in the Browne.pdf powerpoint. The accompanying text, what there is of it, indicates it contradicts figure 10.2 in the 09chap10.pdf document, which makes me think that either I’ve completely misunderstood what it’s saying, or the author, in the rush to produce his post-budget Powerpoint slides, mislabelled it.”
Have only just seen this. What exactly makes you think the table in the Browne document contradicts Figure 10.2 in the Crossley one?
If there is an apparently contradiction – which I can’t see – wouldn’t it be better to try to understand what’s going on, rather than arbitrarily assuming one of the documents is wrong?
@Anthony
Little point in us arguing about this further. I’ve sent the following email to James Browne. It may be summer holidays, so I’m not expecting an immediate reply. But hopefully, we’d know who is right before long.
The letter I sent is as follows:
Dear Mr Browne,
I’ve been engaged in vigorous debate about how regressive VAT is, and your slide show http://www.ifs.org.uk/budgets/budgetjune2010/browne.pdf has often been referred to.
I have been particularly puzzled by the table on page 11, and I think there might be an error. Would it be possible to let me know if this is the case?
The graph appears to suggest the following:
First decile – the 10% who spend the least.
They will lose about 0.9% of their expenditure, but only 0.3% of their income. In other words, they have an income about three times their expenditure. This implies that there are millions who are spending very little, have substantial incomes, but choose to live on the breadline.
Final decile:
The 10% who spend the most have an income of only about 66% of their expenditure. So, those spending the most are saving the least.
In other words, the rich are saving far less than the poor.
This appears to contradict the IFS document,
http://www.ifs.org.uk/budgets/gb2009/09chap10.pdf , which says “those with high current incomes tend to have high saving”. And your own slide show, which says, “those with low incomes tend to have high expenditures relative to their incomes”.
Is it possible that the figure in your document on page 11 is mislabelled. And that the key should show the purple bar charts to be as a proportion of income, and the blue bar charts to be as a proportion of expenditure?
Best regards,
George Kendall
“This appears to contradict the IFS document,
http://www.ifs.org.uk/budgets/gb2009/09chap10.pdf , which says “those with high current incomes tend to have high saving”. And your own slide show, which says, “those with low incomes tend to have high expenditures relative to their incomes”.”
But the chart on p. 11 relates to households grouped by expenditure, and the comments you’ve quoted relate to the characteristics of different income groups. Surely the whole point of the IFS analysis is that these two classifications are quite different.
“Is it possible that the figure in your document on page 11 is mislabelled. And that the key should show the purple bar charts to be as a proportion of income, and the blue bar charts to be as a proportion of expenditure?”
????
You think the lowest decile might be spending three times more than its income?
Anyway, if you compare those purple bars to the dark green bars in Figure 10.2 in the other document, the relative heights look similar, so I doubt they’ve been incorrectly labelled. (One shows the effect of the rise in VAT and the other the total amount paid in VAT, so there could be some differences relating to zero- and lower-rated spending, but surely not the huge differences that are apparent between the light blue and dark green bars.)
“there could be some differences relating to zero- and lower-rated spending”
Sorry – obviously there could be some small differences relating to lower-rated spending, but zero-rated goods wouldn’t figure at all.
@You think the lowest decile might be spending three times more than its income?
The lowest decile in income would include self-employed people, who are earning nothing at the time the data is collected, but at other times earn huge salaries.
I agree it’s unlikely. But more likely than the opposite. And not flatly contradicted by other parts of both documents.
@If you compare those purple bars to the dark green bars in Figure 10.2 in the other document, the relative heights look similar
I was trying to think of a simple explanation, but I think you are right. If so, I hope he comes up with an explanation why the poorest have income about three times their expenditure. It’d be really annoying if I just get a “there is no mislabelling” reply, without an explanation.
And I hope I haven’t made a fool of myself by missing something blindingly obvious!
“The lowest decile in income would include self-employed people, who are earning nothing at the time the data is collected, but at other times earn huge salaries.”
But we’re talking about the lowest decile in expenditure, aren’t we? Isn’t grouping by expenditure meant to eliminate that kind of anomaly? It seems to me that it actually introduces a different anomaly, which judging by the disparities between income and expenditure, is even bigger.
“If so, I hope he comes up with an explanation why the poorest have income about three times their expenditure. It’d be really annoying if I just get a “there is no mislabelling” reply, without an explanation.”
Well, I would still question whether grouping by expenditure is a good way of identifying “the poorest”.
One possible explanation could be that there are quite large numbers of elderly people in the lower expenditure deciles, who may have quite large investment incomes, but not very large expenditures. For example, they may have paid off their mortgages, finished spending money on their children, they may not go on a lot of foreign holidays, may not spend a lot on entertainment/socialising, may not travel around a lot, and so on.
@Anthony Aloysius St
James Browne of IFS gave a very useful reply, see below.
It’s revealing of how these studies are conducted, and the problems researchers have producing data.
The simple answer seems to be that we’ve been over-analysing the graph. The data it was based on was imperfect, and the IFS had to do a lot of work to make it useful. Their work didn’t iron out all the anomalies. We saw an anomaly, which implied that the poor are earning three times their income. They aren’t.
Dear George,
Thanks for your message. The graph is labelled correctly, but I don’t think that your interpretation of this is correct. The data we use for this analysis relies on a household survey in which people are asked to keep a diary of all their spending during a two week period. This fails to pick up a large proportion of actual expenditure, with the result that we have to scale up the losses to households from a VAT rise, and the total expenditure of households. Obtaining the data in this way also means that we are measuring expenditure rather than consumption, so someone who happens to buy a large durable good in the week when they are surevyed will appear in a high expenditure decile, even if in practice their consumption is much lower. Therefore, those with high current expenditures are likely to have expenditure greater than their income at the snapshot when they are observed. The fact that those with low expenditures appear to be saving is likely to be due to measurement error and/or because they happen to have very low expenditure in a particular week in a way which might not reflect consumption (so they may be running down supplies of food they have in their home in that particular week, for example).
Either way, the conclusion is that increasing VAT is mildly progressive (so it is not as progressive as increasing income tax or National Insurance, for example). This is because, although the goods subject to the standard VAT rate form a larger proportion of total budgets for the lifetime rich, there is no tax free threshold or allowance in the same way as there is for income taxes.
Best wishes,
James Browne
George
Thanks for posting that. It’s evidently not worth wasting any more time discussing these IFS figures. What they seem to be saying is that the apparent factor of 3 between income and expenditure in that group is due to some kind of methodological error.
How they feel justified in drawing any conclusions at all when their data contain such huge errors is beyond me.
@Anthony Aloysius St
This is a response to your comment in https://www.libdemvoice.org/opinion-genuine-progressives-should-suggest-cuts-20352.html
“How can you give credence to the report, when the expenditure data it is based on are so seriously flawed?”
I think we’d both be surprised about the amount of statistical work that researchers have to do on raw data, in order to draw meaningful conclusions. That the raw data is messy doesn’t mean the conclusions are worthless.
In a report like “The IFS Green Budget: January 2009”
http://www.ifs.org.uk/publications/4417 they don’t provide us with the raw data. Ultimately, there is a degree of trust. Trust in the IFS, trust in the ESRC who part-funded it, and in the wider academic community who will have critiqued it.
Ultimately, what I find compelling is that a set of serious researchers, who are subject to a great deal of scrutiny, have concluded that, based on the studies they have done, that bottom decile of expenditure is much more typical of the poor than the bottom decile of income. This is even more compelling when the reason they give makes so much sense. I can’t see how anyone could argue that rich people living off savings are poor.
And then it’s interesting to read their conclusions that, despite the rich spending more on private health care and on mortgages, they still end up spending a higher proportion on VAT payable goods. When that’s based on examining the results of real research of the real spending behaviour of real people, I find it quite compelling.
I’m sure you’ll say, based on the tiny amount we know about James Browne’s work, that you don’t trust it. For myself, I do. Though I’ll be interested if someone can provide a link to another set of serious academic research which gives new insights on the issue.
@Anthony Aloysius St
I’m not going to continue this discussion. I don’t think there sufficient data to make any more meaningful comments. If you want, you could send an email to James Browne to ask for more clarification. But I’m not going to do that. I think his reply was extraordinarily helpful and honest, and I don’t want to take any more of his time.
And I’m not going to talk about it in the other thread. The discussion on this thread has been interesting for us, but extremely boring for anyone else. If we repeated it on the other thread, we’d probably kill discussion about anything else. Let’s just agree to disagree.
George
See my response on the thread you are trying to run away from.