To cut income tax – a progressive tax – at the same time as increasing VAT – a regressive tax – is always going to hit the poorest hardest.
Our Guest writers are Howard Reed, of Landman Economics and Tim Horton, Research Director of the Fabian Society
A few months ago we published an analysis of why the Liberal Democrat election manifesto commitment to spend £17 billion raising the income tax personal allowance was not a particularly progressive policy. Contrary to how it has been presented (as something “especially to help those on low incomes”), the measure does nothing to help the millions of families in the UK whose incomes aren’t high enough to pay income tax, and most of the giveaway goes to people further up the income distribution.
That Lib Dem manifesto commitment has now become the coalition government’s aspiration, and George Osborne started to move towards it today with a £1,000 increase in the income tax allowance.
The government is lowering the point at which people start to pay higher rate income tax so that taxpayers currently on the higher rate will not benefit from the policy. But, even so, as green line in the graph below shows, the distributional impact favours middle-income households rather than poorer ones (that is, middle-income households gain a lot more as a proportion of their overall income than poorer ones).
While the income tax cut is estimated to cost about £3.7 billion by the fiscal year 2012-13, it is overshadowed by an increase in VAT from 17.5% to 20%, scheduled for January 2011, which will raise around £13 billion.
VAT is a regressive tax – those on lower incomes pay a larger proportion of their income in tax than those on higher incomes. This is in large part because household expenditure takes up a larger proportion of income for lower-income households than for higher-income ones.
Incredibly, some of the briefing around the budget today suggested that the coalition saw the income tax cut as some kind of compensation for the VAT rise. But, as the graph above shows, that isn’t the case at all.
The blue line shows the combined impact of the income tax cut and the VAT rise using one standard method of calculating the distributional impact of VAT. (Basically, the distributional impacts of VAT are harder to estimate because this depends on the relationship between income and expenditure for each household. This technique uses spending figures and income figures from the UK Expenditure and Food Survey (EFS) and work out the distributional effects of the VAT increase as if it were a reduction in disposable income using EFS income deciles.)
The graph shows that the poorest deciles lose far more from the VAT rise than they will gain from the income tax cut. The graph also shows the impact is extremely regressive across the income distribution, with low-income groups losing out far more proportionately than middle- and higher-income groups.
And some groups will be especially badly hit by the VAT increase – especially those that don’t earn enough to pay income tax: pensioners, the unemployed and parents in low-paid part-time work.
The IFS has suggested that VAT is not as regressive as this, and instead that the impact is proportional to incomes. (This would mean that the impact of VAT is flat across the incomes deciles.) So the red line in the graph above shows the combined impact of the income tax cut and VAT rise using the IFS’ approach. As can be seen, the result is less regressive than before, but still regressive nonetheless. The reality is probably somewhere in between the two approaches.
Whichever one you use, however, it still undermines the coalition’s claim to be creating a fairer tax system. To cut income tax – a progressive tax – at the same time as increasing VAT – a regressive tax – is always going to hit the poorest hardest.
Elsewhere, the budget announced a further range of changes to the tax credits and benefits systems. Some of these, such as higher child payments for the Child Tax Credit, and increases in the Pension Credit, are progressive. But they are not generous enough to completely overturn the regressiveness of the tax package as a whole.
One further aspect of the budget will further exacerbate this regressiveneess over time: the announcement that benefits are to be indexed by the Consumer Price Index (CPI) rather than the Retail Prices Index (CPI). This is regressive in the short run, because the CPI is currently around 1.7 percentage points lower than the RPI. This amounts to a real-terms benefit cut in future years by stealth. The fiscal projections in the budget assume that the government will save almost £6 billion by 2014-15 through uprating by CPI instead of RPI. If this is an accurate estimate, the regressive effects of the uprating change will continue over the whole of this parliament.
Far from being ‘unavoidable’, all these policy changes are discretionary choices. A government committed to fairness could have done so much better.
43 Responses to “Osborne’s tax changes are regressive (not “progressive”)”
Anna PS
Does the IFS claim that VAT expenditure is proportional to income? If so, please could you link to this claim?
This IFS report states that the bottom decile of households by income spends roughly 19% of net household income on VAT, while the top decile of households spends roughly half that, at just over 9%: http://www.ifs.org.uk/budgets/gb2009/09chap10.pdf (PDF – see Figure 1, in ‘Myth 2: VAT is a regressive form of taxation’).
So in fact, VAT is not proportional to household *income* at all – there is a notable spike in VAT for the people with the very lowest incomes.
The IFS goes on to claim that VAT as a proportion of household *expenditure* is a better method for measuring the impact of VAT, and indeed VAT paid as a proportion of expenditure is fairly constant across income deciles, or even slightly progressive (Figure 2 in the above).
So, in this report at least, the IFS does not claim that VAT is proportional to household income, but rather to household expenditure.
[As an aside – Why does the IFS believe that expenditure is a more reasonable measure for calculating the impact of VAT than income? Well, this report states that many people in the bottom decile of income are only there temporarily – e.g. students or those between jobs. Their spending (and thus VAT expenditure) may be relatively high compared to their income, since they do not need to reduce it during their short spell of poverty. While this is clearly true of some people in the bottom decile, it seems unlikely to be true of all of them – the IFS does not provide any analysis, so it’s hard to evaluate. Meanwhile, ONS analysis suggests VAT is regressive even as a percentage of expenditure, as the IFS acknowledges here: http://www.ifs.org.uk/publications/4813 ]
So, calculating by income deciles, the true impact of the VAT rise is always closer to the blue graph than the red graph. A truly bleak picture.
This is clearly a complex topic, so have I missed something?
One more thing – the IFS report warns that the Family Expenditure Survey ‘significantly under-records expenditure on VATable goods’ (p. 197), by a factor of around 1.4. Do you use a similar multiplying factor for the UK Expenditure and Food Survey in your blue graph?
Politics Summary: Wednesday, June 23rd | Left Foot Forward
[…] of the state. He is seeking to arrange a once in a generation change in the terms of trade.” Left Foot Forward showed yesterday that the tax changes – when taken together – were […]
Daily Cuts Briefing – Wednesday 23rd June « A Thousand Cuts
[…] Left Foot Forward on why the Budget is regressive, not progressive […]
Howard Reed
Anna – many thanks for your comments. In terms of the IFS analysis, I think the point they are trying to make is that the UK Expenditure and Food Survey (which is what they use to calculate the distributional impact of VAT changes) doesn’t measure household income very well as it’s based on a snapshot weekly income measure rather than (e.g.) an annual measure. There are a lot of households at the bottom of the EFS income distribution who seem to be spending a lot more than they are receiving in income. There are several possible explanations for this. They could be consuming out of savings, or expenditure could be “lumpy”, or the income measure could be volatile. IFS suggest using expenditure deciles to look at VAT changes because they are closer related to longer run income measures. On expenditure deciles, the impact of VAT is (roughly) proportional. So the idea of the red graph was to show how much difference it makes to the results if we assume a proportional impact of VAT. As it turns out, the tax package as a whole is still regressive even if we do that. And the true impact is likely to be somewhere in between red and blue as you say.
That’s a really good point on ONS – they do suggest that VAT is regressive even based on expenditure, yes. So no – I don’t think you’ve missed anything.
We do correct for the under-recording of VAT expenditure in EFS, yes. The figures here multiply up the distributional effects across deciles so that the overall yield from the VAT increase comes in at £13bn.
Cliff James
RT @leftfootfwd: Budget 2010: regressive (NOT “progressive”) whichever way you look at it: http://bit.ly/9vCnFX