Friday, 22 October 2010

Revealed: How CSR cuts will hit the poorest 15 times harder than the rich

Deputy Prime Minister Nick Clegg has laid down the challenge in the Guardian this morning. The deputy Prime Minister insists that any fair assessment of the CSR should not just look at tax and benefit changes, but look too at the changes on spending to the services people 'live by'. He said:

"It goes back to a culture of how you measure fairness that took root under Gordon Brown's time, where fairness was seen through one prism and one prism only which was the tax and benefits system. It is a complete nonsense to apply that measure, which is a slightly desiccated Treasury measure. People do not live only on the basis of the benefits they receive. They also depend on public services, such as childcare and social care. All of those things have been airbrushed out of the picture by the IFS."

Mr Clegg is not always wise in picking fights with the Institute for Fiscal Studies.

But let any such airbrushing end at once.

Let us immediately take up the deputy Prime Minister's challenge.

What happens if you undertake the fullest possible survey of the distributional impact of public spending changes - on the services that "people live by", as Clegg rightly says?

Is what the deputy PM thinks true? Are the spending changes much more progressive than the government's tax and benefit changes? Unfortunately for those hoping the government would be able to act on its "progressive austerity" commitments, he could not be more wrong about that.

A full analysis of the spending changes shows these are very sharply regressive. (Indeed, the specific changes which the government has announced have proved more regressive than was anticipated by allocating the overall scale of cuts equally to non-ringfenced departments).

The new post-CSR data is published by the TUC today. (The data can be downloaded here).

It shows that the poorest ten per cent of households will be hit 15 times harder than the richest ten per cent as a result of service cuts announced in the comprehensive spending review. The research been undertaken by Howard Reed of Landman Economics, experts in economic modelling, and Tim Horton, research director of the Fabian Society, to offer a post-CSR supplementary analysis of their 'Where the money goes' (PDF) report published by the TUC last month.

Using official figures to calculate how different groups benefit from different public services, the Reed and Horton analysis shows that the poorest ten per cent of households, with incomes below £10,200, will suffer reductions in spending on services equivalent to 29.5 per cent of their annual income on average, or £1,913 a year.

The second poorest group of households, with incomes between £10,200 and £12,900, will be hit hardest in cash terms – losing services worth £2,164 a year – and the TUC analysis confirms that the higher up the income scale people are, the less they lose from the cuts. The richest ten per cent will lose services worth just two per cent of their net income, the equivalent of £1,506 a year.

The analysis examines the impact of the CSR on different types of family and finds that lone parents will be hit the hardest, losing services worth 18.4 per cent of their income on average (£3,121 a year). Single pensioners are next, losing services worth 11.1 per cent of their income on average (£1,305 a year).

The analysis uses the same spending model behind the TUC report Where The Money Goes published last month on the eve of TUC Congress. This found that on average households benefit from £21,000 worth of services a year, and that those on low or modest incomes gain more than the better-off.

Where The Money Goes predicted that cuts of 25 per cent by 2012-13 (while ringfencing health expenditure and partially protecting education) would mean that the poorest ten per cent of households would lose around 20 per cent of their income.

But today’s analysis, using data from the CSR, shows that overall cuts to public spending (excluding benefits and tax credits) of £48 billion (in today’s prices) by 2014-15 will be even more regressive, partly because of deep cuts to services which are disproportionately used by the poorest households – such as social housing and social care.

The analysis examines the impact of the cuts on four typical families:

* A family with two school age children on modest earnings will suffer service cuts equivalent to 13.2 per cent of their income, or £2,631 a year.

* An affluent family with children at university will suffer service cuts equivalent to 19.4 per cent of their income, or £3,889 a year.

* A working lone parent with two children will suffer service cuts equivalent to 15.7 per cent of their income, or £3,132 a year.

* A pensioner couple will suffer service cuts equivalent to 16.2 per cent of their income, or £2,226 a year.

The Treasury's own distributional analysis of the CSR also shows that the spending changes are regressive. However, the Treasury analysis has chosen to model only about half of public spending - omitting home affairs, policing and most areas which are public goods like defence or environmental protection - while the TUC analysis seeks to model all the impact of all of the spending cuts in the CSR. The TUC reports finds that the smaller Treasury sample considerably underestimates the regressivity of the CSR by comparison with the model which seeks to count all of the spending.

To coincide with these findings, the TUC is launching a new ‘cuts calculator’, which allows people to work out how much they are likely to lose from the spending review. (The online cuts calculator).


Oranjepan said...

I find it interesting that your sources appear alligned politically, and this seems designed to make it hard to avoid the conclusion which you're so fiercely driving at.

As a methodology it is clearly unsound, irrespective of any facts presented, so it casts doubt on your analysis and any conclusions to be drawn.

You can get away with it when in government because the state is officially non-partisan, but not when in opposition.

So if you could provide some corroboration for these views which have been arrived at independently, rather than simplistic tit-for-tat ideological ripostes I'd be more likely to be convinced by your sums.

That is if you're more interested in political debate than a bun fight, of course.


Sunder Katwala said...


You may have to clarify what you are challenging about the findings. It will certainly be possible to find people who say 'distributional analysis is not important and doesn't capture what is fair' (those are value judgements and political arguments) but I do not think you will find any neutral or independent academic voice who could disagree with either of these factual analytical points:

(i) the overall pattern of overall public spending is progressive
(ii) the distributional pattern of the CSR spending cuts is regressive.

There is no significant conflict between the Howard Reed data and the Treasury data [distributional annexe] for the 50% of public services spending data which the Treasury include.

The Treasury published its first ever distribution of spending annexe in the CSR. We have been pushing them to do this. The IFS stated clearly in yesterday's presentation that the Treasury data shows a regressive pattern of spending changes.

If you read the Reed/Horton September report, you will see it is very clear about the modelling assumptions necessary. There are a series of detailed methodological debates - these in large part revolve around how to distribute (if at all) the value of public goods (like defence). The IFS discusses these here, in response to the first Reed/Horton report. But none of these issues could alter the issue of the progressive distribution of public service spending, see this July 2010 report.


The inclusion of more data does change the scale of lost services (the Treasury estimates losses of £7-11 a week for households based on the 50% of spending sample) and it does somewhat change the scale of regressivity.

The detailed Reed slides will be published today by the TUC. While the inclusion of areas of spending (on a flat-rate) for which there is no distributional household data shows a fuller picture of spending changes, it does not fundamentally alter the underlying distributional pattern, and I can't see any reason why this is in doubt.

You can also find Howard Reed's credentials as an economic modeller at the Landman Economics website. His research was covered in the CSR analysis of yesterday's FT and used to generate one of the graphics. The Economics Editor of the FT wrote that the most plausible defence of regressive spending changes is that all spending changes on this scale are bound to be regressive, given the progressive pattern of public spending. The government could choose to argue that - as some government supporters also advocate - but that is not the argument that Osborne or Clegg have used this week, which has been the factual claim that the spending changes are distributionally progressive.

There does not seem to be any evidence to support this, as is clear from the Treasury annexe.

Sunder Katwala said...
This comment has been removed by the author.
Oranjepan said...
This comment has been removed by the author.
Oranjepan said...

I'm agreeing with your point i), but arguing that ii) is unproven.

In particular how is the diproportionate benefit of public services to lower income groups which is indicated by your point i) reflected in either the impact to net income or net expenditure?

The answer is that it isn't.

And that is what Clegg is getting at - the issues you refer to can and do alter the issue of the progressive distribution of public service spending.

The main example given is defence spending. This is unhelpful as this tends to be more evenly distributed across income groups.

It is much more informative to look at education or health spending, which are in any case higher proportions of the total budget.

The fact is that lower-income groups are much more likely to see the benefits of spending in these areas as they are proportionately less likely to send their children to private fee-paying schools like Harrow or Eton, or have private health insurance and be registered with Harley St doctors.

And as the IFS discussion on the point which you give as reference states, even this is less clearcut: "the precise composition and manner of implementation of the package of impending spending cuts will crucially determine the extent to which they are progressive or regressive."

Indeed, the devil is in the detail.

So it's odd that the IFS has changed its tune in response to the CSR. And I think it's worth asking why.

Is it perhaps to do with the appointment of the IFS' Robert Chote to head up the Office of Budget Responsibility which occurred in September?

I've been looking for a comment from him on the CSR, but so far I've found none - can you help on this?