Skip to main content

Committee urges Chancellor to re-instate distributional analysis

25 August 2016

Rt Hon. Andrew Tyrie MP, Chairman of the Treasury Committee, has written to Rt Hon. Philip Hammond MP, Chancellor of the Exchequer, to ask him to re-instate the distributional analysis used by his predecessor in his Budget and Autumn Statement reports prior to the 2015 Parliament.

Chair's comments

Commenting on the correspondence, Mr Tyrie said:

"Since the Summer Budget of July 2015, the Treasury has replaced its previously excellent budget distributional analysis series with a manifestly deficient substitute.

The new Prime Minister is committing her Government to making Britain a country that works ‘not for a privileged few, but for every one of us'. A high level of transparency about the effects of tax and welfare policy on households across the income distribution would seem to be a logical, perhaps essential starting point."

Background

The Coalition's first Budget in June 2010 published a "distributional analysis" showing the impact of changes to tax, welfare and public spending on households in different parts of the income distribution.

Although the methodology has changed slightly over time, the analysis always showed the impact of the Government policies since January 2010 on household incomes in each decile.

A distributional analysis showing the impact of the Government's fiscal policy on household incomes had by the end of the last Parliament become part of the standard Budget documentation.

Summer Budget of July 2015

However, since the Summer Budget of July 2015, a different kind of distributional analysis has been published. The presentation of the analysis has changed in three ways:

  1. The analysis now shows changes in the share of public expenditure received and taxes paid by households. In effect, it shows "an assessment of how public spending and taxes are distributed, rather than the amounts people receive."
  2. The analysis now compares the distribution of public spending and taxes in 2019-20, with a baseline of 2010-11.
  3. The analysis is split by quintile rather than decile

The first change, which apportions the benefit of public expenditure onto households, creates two problems:

  • First, attributing the benefits of Government expenditure streams such as those for example spent on prisons and courts is extremely subjective and therefore not easily verifiable or easy to subject to scrutiny.
  • The Treasury Committee took evidence from Paul Johnson, Director of the Institute for Fiscal Studies, who noted the methodological difficulties of apportioning public spending to different parts of the income distribution: "the problem is […] it is quite difficult to allocate spending on police, justice, defence, environment and so on to households.
  • Second, including the benefits received through the consumption of government services with actual cash payments received obscures the impact of a Budget on actual household incomes. Such transparent analysis is vital when scrutinising a Budget for both the Treasury Committee but also wider stakeholders.

The second change which altered the time scale over which the analysis is presented, has created two problems:

  • First, the analysis cannot specifically identify the impact of the policies contained in each fiscal event (e.g. the 2016 Budget) because it includes all policy measures from 2010-11 onwards.
  • Secondly, the analysis cannot be used to determine the cumulative impact of the Government's policies on households at present, because the comparator is set five years into the future, rather than the current financial year (2016-17).

Treasury's Budget 2016

Paragraph 1.7 of the Treasury's own Budget 2016 explanatory notes state their analysis cannot be used to interpret the impact of the 2016 Budget by income distribution, because their modelling assumptions change at each fiscal event:

HM Treasury conducts an ongoing programme of model development, and incorporates updated economic assumptions, as well as the analysis of new policy announcements, at each fiscal event. For that reason the charts in this publication are not directly comparable to the charts published at Autumn Statement and Spending Review 2015, and any difference between the two should not necessarily be interpreted as the impact of Budget 2016.

The third change, which aggregates income groups by quintile rather than decile, means that it is no longer possible to ascertain what the impacts of measures in each Budget or Autumn Statement are on both the very poorest and very richest households, an issue which is of interest to the Committee and wider stakeholders in the Budget debate.

History of Treasury Committee pressure on former Chancellor, Rt Hon. George Osborne MP

In evidence to the Treasury Select Committee on the Spending Review in 2010, the Chancellor said:

Mr Osborne: My predecessor as Chancellor had all this information available on the distributional impact of tax and benefit changes; they just chose not to publish it. But what has not been attempted by the IFS—and I think they may have mentioned this in their appearance before you—is to model the impact of public expenditure changes and the benefits in kind that come from those.

That is genuinely new work—cutting-edge work as the Permanent Secretary puts it, which it is in his world—and we genuinely welcome him, not just at this Committee but the broader academic community.

I don't want to speak for members of other political parties here but if we can get to a point where the political parties aren't arguing about the methodology, but instead are arguing about the public policy measures, I think we'd all be in a good place.

In his foreword to the distributional analysis published following the March 2015 Budget, the Chancellor wrote:

Today's publication represents the most comprehensive and robust assessment available of how the decisions we have made in this Parliament have affected families. We made a commitment that we would reduce the deficit in a fair way. The analysis shows that our commitment has been met (p.3).

On 17 December 2014, the Chancellor gave evidence to the Treasury Committee on the Autumn Statement 2014, expressing his commitment to publishing the distributional analysis:

Q359 Chair: Would you confirm that you are intending to stick with [the distributional analysis] and that you are intending to stick with the same type of analysis, with the same tools, so that we can get a good time series for the evidence you are producing?

[…]

Mr Osborne: Indeed. I want a continuous data series.

In evidence to the Treasury Select Committee on the Summer Budget on 21 July 2015, the Chancellor was asked why he had chosen to alter the construction of the distributional analysis:

Q259  Chair: Chancellor, just to clarify these earlier exchanges on the distributional analysis, you are saying you are going to provide us with the distributional analysis on the same basis as they were done in the previous Budget, and present them—

Mr Osborne: I have certainly seen distributional analysis in the preparation of the Budget, as you might expect, that was similar to that produced in previous Budgets, which was one of the reasons that, as you might expect, it was not readied for publication but I can get it, I can produce it ready for publication. But what I am not prepared to commit to is going on producing this analysis for future Budgets and future fiscal events. I think this is a much better approach. I also think once we have passed beyond 2015, it is incredibly difficult to make that analysis.

Q260 Chair: That is an issue that we can come back to. I just want clarification on what you are going to provide the Committee with, which is the distributional analysis on the basis set out in the previous Budget for this Budget's changes? Is that correct?

Mr Osborne: The quintile analysis we did is the one I have seen in the advice that was presented.

Q261 Chair: Just to be clear, what this Committee is asking of you is the distributional analysis that you did for the last Budget, applied to this Budget.

Mr Osborne: Yes.

Correspondence

Following the evidence session on 21 July, the Chancellor published a single chart showing the cumulative overall impact of public service spending, tax and welfare changes on households in 2017-18, as a percentage of 2010-11 net income (including households' benefits in kind from public services), by income quintile.

Unsatisfied with the completeness of the Chancellor's response to the Committee's request, the Chairman of the Treasury Committee wrote to the Chancellor on 10 August 2015 (PDF 368 KB), outlining the information required to satisfy the Committee's request.

On 9 September 2015, the Chancellor responded (PDF 346 KB), refusing the Committee's request and providing no additional information as to the distributional impact of the Summer Budget

On 28 September 2015, the Chairman of the Treasury Committee responded to the Chancellor (PDF 715 KB), reiterating the Committee's request.

On 22 October 2016, the Treasury Committee took evidence from Mr Osborne. During the session, the Chancellor was again asked when he will respond to the Committee's repeated requests for "the distributional analysis [he] did for the last Budget, applied to this Budget", something he consented to provide in July. He said he was "working on it". The Treasury Committee never received the analysis.

Further information

Image: PA