THE GOVERNMENT'S RESPONSE TO THE TREASURY
COMMITTEE'S EIGHTH REPORT, 1997-98: THE NEW FISCAL FRAMEWORK AND
THE COMPREHENSIVE SPENDING REVIEW
(a) We recommend that any subsequent substantial public spending
announcements be accompanied by a two-day debate (paragraph 4).
The Government agrees that there should be adequate parliamentary
time to discuss major public spending announcements. The length
of any debate would depend on the nature of the announcement.
This is a matter for the Business Managers to decide at the time
after consultation with the Opposition parties through the usual
(b) We recommend that the Treasury extends the innovation of
the pre-Budget Report (PBR) by issuing a consultative document
on spending plans prior to each rolling forward of departmental
expenditure plans. We further recommend that the Treasury should
seek to ensure that Parliamentary time is made available to debate
such a consultative document (paragraph 4).
The Comprehensive Spending Review was carried out with a zero-based
analysis of priorities and programmes across the whole of government.
When the Review was launched, the Government's overall priorities
were clear from its election manifesto; and Parliament and outside
commentators were free to come forward with suggestions and comments
on the details of the policies and programmes that should be pursued.
A number of departments consulted during their reviews and numerous
representations were received on specific issues. But the zero-based
nature of the Review precluded the Government from publishing
in advance a full draft set of spending plans for the years 1999-2000
to 2001-02 upon which comments could be invited. In future reviews
of spending, the Government will also be open to comments and
representations from Parliament and from other interested parties.
The next review of spending will coincide with the move to resource-based
accounting and budgeting, on which Parliament has been fully consulted
through hearings and memoranda.
(c) It is important for Parliament and the public to have advance
notice of when major economic policy statements are to be made.
We would, therefore, welcome a timetable from the Government setting
out the likely date of the PBR and the Budget and the date in
the summer of 2000 when the next all-encompassing public spending
statement is due (paragraph 6).
The Government agrees that it is important for Parliament to have
as much advance notice as possible of major economic policy statements.
The Pre-Budget Report will be published on 3 November but it is
too early to put a date to the next Budget or the next public
(d) We recommend that the Treasury should publish now, and
at the time of future Budget and Spending Statements, a set of
different possible outturns for key measures of government borrowing
based upon a range of scenarios for the growth of gross domestic
product. This will increase transparency and allow a better understanding
of the risks to official forecasts of the public finances from
different growth outturns (paragraph 11).
Documentation of Treasury forecasts of the economy and public
finances regularly includes information on average errors from
past forecasts and the sensitivity of the public finances projections
to the path of the economy. Table A4 of the March 1998 Financial
Statement and Budget Report showed average errors in the economic
forecasts and paragraph B15 gave average errors in the projections
of the budget deficit. Paragraph B16 also indicated the sensitivity
of the public finances projections to cyclical changes in GDP
and to changes in the average rate of GDP growth over a period
of years. While this sensitivity analysis was for the net cash
requirement, the results would not be significantly different
for net borrowing and the current budget surplus. The Financial
Statement and Budget Report (page 32) and the Economic
and Fiscal Strategy Report (page 46) also showed the sensitivity
of the projections to a more pessimistic assumption about the
level of GDP relative to trend. These sensitivity analyses are
based on the estimates of the effects of GDP on the public finances
described in Treasury Occasional Paper No. 4, September
(e) We recommend that the Treasury should set out clearly the
thinking behind the current level of the reserves, how the decision
on their magnitude has been taken, what part is to be played by
intra-departmental reserves, and what their size is expected to
be (paragraph 11).
The new expenditure planning and control system sets firm three-year
plans for all departmental expenditure limits, rather than requiring
annual revisions of plans. These DELs are based on the rigorous
analysis carried out in the Comprehensive Spending Review. Whereas
plans set in the Control Total in previous public expenditure
surveys were sometimes unrealistic, and needed to be topped up
in later surveys, the DEL plans will not be changed every year.
Departments are also considering whether to create unallocated
provisions to enable them to deal with unforeseen new spending
pressures. The DEL Reserve is therefore set not to fund future
injections of cash for departments in spending rounds, but to
deal with unforeseeable emergencies, for example, a problem such
as BSE. It is consequently appropriate to set the Reserve at a
prudent but lower level than the Reserve set within the Control
Total, which was kept higher in order to permit a policy of topping
up plans in later years. The DEL Reserve has been set slightly
higher in later years, but not by a large factor, as it will not
need to be called on in successive spending rounds. Hence it has
been set at £1.5/2.0/2.5 billion over the three years.
Within Annually Managed Expenditure, the AME Margin has been set
to provide some headroom within Total Managed Expenditure to help
deal with the inevitable variations of the demand-led programmes
in AME. The AME Margin has been set in a "wedge" shape
(£1/2/3 billion over the three years), to allow for the possible
knock-on impact of any estimating change to later years.
(f) We recommend that the Treasury quantify what is meant by
inflation varying "substantially" and state whether
if such a deviation occurs the Government does intend to make
additional payments to departments (paragraph 12).
The Treasury has agreed with Departments that it will not consider
reviewing the cash figures other than in the event of inflation
varying by more than 1.5 per cent cumulatively from the projections
in the Comprehensive Spending Review. There is, of course, no
presumption that any review would lead to any changes at all in
Departmental cash plans.
(g) While we understand and welcome the Government's emphasis
on the need for public sector capital investment where the returns
from this are high, we are concerned that the continuing large
differential between public and private sector wage increases,
in favour of the latter, could serve over time to undermine the
recruitment into key public services of high quality employees
The Government recognises that the key considerations in setting
public sector pay should include recruitment and retention requirements.
This will nevertheless need to be consistent with the need for
Departments to meet their CSR output targets for the delivery
of services within the three-year spending plans. The Government
believes that the CSR provides room for realistic settlements
on this basis.
The Government notes, in response to the point raised about large
differentials in wage increases, that the Average Earnings Index
and its components have been significantly revised this month.
The latest figures no longer support the case that there is a
large gap in growth rate between pubic and private sector earnings.
Also data from the New Earnings Survey shows that, during the
1990s, overall earnings growth in the public and private sectors
has been similar.
(h) There is an apparent inconsistency between the Monetary
Policy Committee position that unemployment must rise if the inflation
target is to be met and the Treasury's assumption that unemployment
will be held flat at its April level. We accept that by convention
Governments never publicly forecast a rise in unemployment, but
we believe that in the interests of transparency the Treasury
should now begin to publish its estimates for unemployment when
it publishes new spending plans (paragraph 14).
In line with the convention adopted by previous governments, the
Government does not publish forecasts for unemployment. However,
for the purposes of projecting social security expenditure, it
is necessary to make a planning assumption about the level of
unemployment. Since the July 1997 Budget, the Government has based
its public finances projections on the assumption that unemployment
remains flat at its most recent level. Accordingly, the CSR projections
assume that unemployment remains flat at its level in May 1998.
The flat assumption follows a long-standing convention used by
both Labour and Conservative Chancellors up until the 1996 Budget,
and has been audited and approved as reasonable by the National
Audit Office. Forecasts of unemployment are highly uncertainerror
margins are typically large and, over the medium term, the range
of outside forecasts for unemployment is very wide. As well as
cyclical effects, the flat assumption also discounts the effects
that fraud savings, the New Deals and other labour market policies
might have in reducing unemployment (see answer k).
(i) We recognise the benefits of publishing the current budget
balance in the interest of transparency, but expect the Public
Sector Borrowing Requirement to continue to be published alongside
the Public Sector Net Borrowing (paragraph 15).
The Code for Fiscal Stability states that the Financial Statement
and Budget Report will continue to include forecasts for the
net cash requirement (PSBR), as well as net borrowing and the
current budget balance.
(j) We seek clarification as to what "balancing the current
budget over the cycle" actually means. Furthermore we recommend
that the Chancellor make clear whether the projected increases
for Departmental spending, and in particular for health and education,
are guaranteed even in the face of a faster than expected rise
in social security spending, and if so how they would be funded
without breaching the Government's overall spending targets (paragraph
One of the Government's two fiscal rules is the golden rule: over
the economic cycle, the Government will borrow only to invest
and not to fund current spending. As explained in the November
1997 Pre-Budget Report and most recently in the Economic and Fiscal
Strategy Report, the golden rule will be met if the public sector
current budget avoids a deficit on average over the economic cycle.
For the purposes of the fiscal rules, 1997-98 is taken to be the
starting point of the economic cycle, as the economy is thought
to have been close to its sustainable level in that year. Table
4.7 of the EFSR set out projections for the average surplus on
current budget since 1997-98.
The Government has set its fiscal rules over the economic cycle
in recognition of the significant effect of cyclical fluctuations
on the public finances. This approach aims to enhance stability
by allowing borrowing to fluctuate due to cyclical changes in
output (ie by allowing the automatic stabilisers to operate).
Meeting the golden rule over the economic cycle would be quite
consistent with some borrowing to finance current spending during
a downswing, provided that this is at least offset by surpluses
on current budget at other stages of the cycle. However, the Government
is taking a prudent approach to meeting its fiscal rules and is
therefore planning for surpluses on current budget, as set out
in the EFSR.
The Government does not anticipate that it should be necessary
to revisit the Departmental Expenditure Limits set out in table
1 of the White Paper before the next review of spending plans
in 2000. It will keep Annually Managed Expenditure under regular
review alongside tax receipts, and ensure that the fiscal rules
(k) We wish the Treasury to be more explicit and give figures
for the extent of the saving on fraud, if any, that has already
been factored in to the social security spending projections.
On the issue of the impact of the New Deal on the social security
figures the Chief Secretary stated that "Yes, we have made
some fairly prudent assumptions as to what the reduction [in the
number unemployed] will be". Again, we wish the Treasury
to state what these assumptions are (paragraph 20).
Social security expenditure is projected forward based on recent
expenditure and caseload information. Expenditure projections
take full account of the anti-fraud programme of work. DSS also
estimates what is being saved from anti-fraud activities planned
for a given year separately from the projections of benefit expenditure.
DSS use the Weekly Benefit Savings approach, which assumes that
had the fraud not been discovered it would have continued for
another 32 weeks, a figure based on average durations of such
claims. This is the same methodology used in accounting for actual
fraud savings. It produces an approximate figure as by definition
it is impossible to know how long a fraud would have continued
had it not been discovered. DSS are currently reviewing their
methodology, in particular to reflect the greater emphasis on
prevention rather than detection of fraud as set out in the Green
Paper "Beating Fraud is Everyone's Business: securing the
The savings figures below are based on the programme of anti-fraud
work planned for each of the financial years shown:
The Government does not publish forecasts of unemployment, and
has based public expenditure projections on the basis of level
of unemployment in May 1998 (see answer h). A prudent approach
does not take credit for reductions in unemployment through the
New Deal until they actually arise.
(l) We will be writing to the Office for National Statistics
for confirmation that the Government's method of classifying Working
Families Tax Credit (WFTC) is in line with internationally accepted
accounting standards; and to increase transparency we recommend
in future presentations of the public finances that the cost of
the WFTC be set out separately (rather than within "accounting
and other adjustments"). We ask the Treasury to explain the
implications for costing WFTC if it is decided to raise the amounts
We note that the Committee has written to the Office for National
Statistics (ONS) concerning the classification of Working Families
Tax Credit (WFTC). For national accounts-based presentations of
the public finances in the FSBR, and subsequently in the EFSR
and CSR, the classification of WFTC has followed the treatment
in the national accounts at that time for similar income tax credits,
such as mortgage interest relief at source (MIRAS).
At the time the WFTC was first announced in the last Budget, the
national accounts followed the international conventions set down
in the 1979 version of the European System of Accounts (ESA79).
ONS adopted a new version of the European System of Accounts,
ESA95, for the UK national accounts in the 1998 Blue Book, published
in September. This involves changes in the treatment of income
tax credits, such as MIRAS. As WFTC will not begin until 1999,
the classification issue the ONS will need to address in respect
of WFTC is how it should score in national accounts under ESA95.
ONS have no operational need to decide the classification of WFTC
The full cost of WFTC was included in the income tax credits line
in Table B8 of the FSBR (Total Receipts). In order to improve
transparency, the Treasury will for future projections of the
public finances include an additional line in this table showing
the cost of WFTC separately from other income tax credits.
Future presentations of the national accounts public finances
aggregates will be in terms of ESA95. In these presentations,
the Treasury will follow whatever classification ONS decide on
for WFTC and other income tax credits. To the extent that part
or all of payments of such tax credits are classified in the national
accounts as expenditure, figures will be included in AME, in the
accounting adjustments line. A full breakdown of the accounting
adjustments is presented annually in Public Expenditure Statistical
Analyses, which is published in March or April along with Main
Estimates and departmental reports. The Treasury has provided
a breakdown of the accounting adjustments line in CSR Table A2
in a technical note for this Committee (27 July).
Because, when following national accounting conventions then in
force, only part of WFTC scored as public expenditure, disaggregating
the accounting adjustments did not show the full cost of WFTC.
In the FSBR, and subsequently in the EFSR and CSR, some 80 per
cent of WFTC payments have been scored in the national accounts
measures of public expenditure. This represents an estimate of
the proportion of total payments which will be in excess of the
individual recipient's total tax liability. Were the scope of
WFTC to be changed, then this proportion would need to be re-estimated.
A higher (lower) proportion scored as expenditure would, other
things equal, increase (decrease) AME and TME.
Office for National Statistics Response
The WFTC has not yet been introduced, so ONS has not yet had to
record it in national accounts since these accounts record outturn
only and do not address the future. The FSBR gives forecasts and
plans for the future and much of its contents are based on national
accounts. So Treasury had to decide how WFTC would be classified
under the system of national accounts in force at that time.
That system was specific to the UK. It had evolved over many years
along basic principles common to international systems of national
accounts. In UK national accounts, tax refunds paid to an individual
in excess of tax actually paid are treated as public expenditure.
This was consistent with IMF guidance on recording government
finances which says that such amounts should not be netted-off
tax receipts even if they are called tax refunds or tax credits.
The treatment is applied for example to Mortgage Interest Tax
Relief at Source (MIRAS). MIRAS is available to people who do
not pay income tax, as well as to tax payers. Inland Revenue estimates
the amount of MIRAS paid to individuals in excess of tax paid.
That amount is included in national accounts as public expenditure.
The rest is netted-off income tax receipts.
The treatment of WFTC in the FSBR follows the treatment of MIRAS.
The Government has estimated that, when considering taxpayers
individually, about 80 per cent of WFTC payments will be in excess
of tax actually paid. In the FSBR this amount was recorded as
public expenditure and the rest, that which is paid to individuals
with a positive net income tax bill, was netted-off income tax
In September 1998, ONS adopts the European System of Accounts
(ESA95) for UK national accounts. In preparing for this system
ONS reviewed the treatment of tax refunds. ONS concluded that
if the refund is not integral to the tax then it should be recorded
as public expenditure. This would apply for example if the assessment
of the amount of refund due were independent of the tax assessment.
In particular, if refunds can exceed the amount of tax actually
paid then it is likely that they arc not integral to the tax.
This approach is consistent with ESA95 which does not make any
references to splitting refunds between negative taxation and
expenditure. ESA95 is clear that the recording of taxes should
be restricted to regulations that require payments to government
rather than payments by government. Another disadvantage with
the old method was that accurate data might not be available to
split the payments between two categories.
In October ONS will be publishing its first annual national accounts
Blue Book under ESA95 rules and will adopt the new method
for tax refunds. For example all of MIRAS will be treated as government
expenditure. The new method will be described in publications
appearing at the same time as Blue Book.
It is likely that, under the revised rules, all of WFTC would
be treated as public expenditure in national accounts. The decision
will be made when the exact details of the scheme are known.
(m) We welcome the publication of separate spending figures
for capital and current spending, while recognising that there
are problems of definition. As the definition actually used is
central to the process, it is essential that it should be readily
available, by being published in the fiscal documents (paragraph
The Treasury has for a number of years included a section in the
Financial Statement and Budget Report that sets out the conventions
used in presenting the public finances, and would expect to include
similar information in future Budget documents. Paragraphs B.62
and B.63 of the March 1998 FSBR set out the main components of
capital spending, based on the distinction between current and
capital spending used in the national accounts. Further detail
of the precise conventions used in the national accounts is published
by the Office for National Statistics.
(n) Although we welcome the Government's clarity in setting
out its objective of keeping to the Golden Rule, we believe that
it may be possible in the future to refine and improve this necessarily
crude fiscal guideline. We recommend that the Government undertakes
further work on intergenerational and other issues which might
in time lead to refinement of the Golden Rule (paragraph 30).
The Government has acknowledged that approaches such as the golden
rule represent an approximate way of matching costs and benefits
of public spending across generations, and that some areas of
current spending can have important effects on future prosperity.
The Government recognises the need for further work on intergenerational
and long-term issues. The Treasury is currently collaborating
with the National Institute for Economic and Social Research and
the Bank of England to produce a set of generational accounts
for the UK. The Government is also developing long-term projections,
as required by the Code for Fiscal Stability, which will give
a further indication of the sustainability of current fiscal policy
(o) We recommend that data on the new basis should be made
available as far back as the fiscal year 1988-89 (paragraph 32).
Data will be made available during the course of the Autumn/Winter.
At the time of publication of the CSR, full data on the new basis
was only available for a period of nine years starting in 1993-94.
Nine years represents the normal length of time period for which
data are collected in the Treasury's main public expenditure database;
and is the standard timespan for tables in Public Expenditure
Statistical Analyses. Treasury staff will be extending the series
in CSR Tables A1 and A2 back to the late 1980s, using information
from old databases. Once that work is completed, these data will
be made available.
(p) We recommend that, for the sake of transparency, in future
the Government should refer to annual increases over the previous
year rather than a cumulative total (paragraph 33).
The Government believes that it is right to indicate clearly how
much new money was provided in the Comprehensive Spending Review
on top of the amounts made available for 1998-99. The full documentation
provided in the white paper "Modern Public Services for Britain"
made clear both the totals of new money, and the components of
those totals in each year. It has been common practice for Governments
of all persuasions to refer to figures across more than one year.
For example, the previous Government's last Red Book ("Financial
Statement and Budget Report 1997-98", November 1996) referred,
inter alia, to "over £3.5 billion of investment in national
roads over the next three years" and "£2.5 billion
over three years for the Housing Corporation" (figures which
refer not to increases over previous plans, but to totals including
money already allocated in previous Public Expenditure Surveys).
(q) We recommend that every year the Government publish sales
from the National Asset Register and a list of local authority
assets that have been sold, how much capital was raised and for
what the funds have been used. As a matter of principle it should
be relatively easy to be able to see how "dead" assets
are turned into "live" ones and whether asset sales
are being used to fund current expenditure and so reduce council
tax bills. We also recommend that the National Asset Register,
which we warmly welcome, should be updated at least every two
years and should be broken down into local geographical areas
such as local authority areas or postcode areas (paragraph 35).
In response to a recommendation from the Committee during the
last Parliament, the
Government has already started publishing aggregate figures both
for projected and actual asset sales. These were published both
in the March 1998 FSBR and in the June 1998 EFSR. In addition,
it has passed a breakdown of asset sales by department for the
years up to 1996-97 to the National Audit Office. In response
to the Committee's new recommendation, the Government agrees to
publish in future such a retrospective breakdown of sales by department.
It will also provide a commentary on some of the larger sales.
As regards publication of the National Asset Register, the Government
understands the Committee's reasons for recommending publication
every two years. The Government will, as part of the Resource
Accounting and Budgeting timetable and in the run up to the review
of public expenditure allocations in 2000, look into the practicality
and cost-efficiency of publishing future editions of the Register,
including a breakdown of the results into local geographical areas.
(r) We recommend that the figures on the Private Finance Initiative
(PFI) are updated twice a year, at the time of the Budget and
for the Pre-Budget Report. We would also welcome details of the
estimated payments each department is expected to pay under PFI
contracts over the next 25 years, and corresponding figures for
exposure to capital liabilities (paragraph 36).
Statistics showing outturn and anticipated investment in PFI projects
are collected and published twice a year; for the Budget Financial
Statement and Budget Report, and then around the end of September,
thereby providing Parliament and the public with regular six-monthly
reports showing the latest position. Information showing the estimated
annual payments departments are expecting to make over the next
30 years is collected at the same six-month intervals, and published
on an aggregated basis (in order to protect commercial confidentiality
in those cases where departments have only a small number of such
(s) We agree with the Procedure Committee that it is essential
that progress against the "trigger points" for Resource
Accounting and Budgeting is reported regularly to Parliament and
that "the decision as to whether the target date for implementation
is still viable must not be taken on the basis of information
available to the Treasury alone". We welcome the Treasury's
assent to this (paragraph 41).
The Government welcomes the Committee's agreement with the Procedure
Committee that progress against the "trigger points"
against which RAB implementation will be monitored and assessed
should be reported regularly to Parliament.
As noted in the Treasury's response to the Procedure Committee
report (HC 773), the Treasury and the National Audit Office will
monitor departments' performance carefully against the trigger
points and will report progress fully to Parliament. The Treasury's
Memorandum on RAB to the Parliamentary Committees in July 1998
contained a detailed assessment of departments' progress against
the first trigger point (Stage 1 approval).
This showed that 44 out of 49 departments had already secured
Stage 1 approval by the end of July, and that each of the remaining
five departments was well on course to do so within the deadline
for Stage 1 (end of December 1998), for the reasons given in the
(t) If the agreements have not already been published by the
time the Treasury responds to this report, we will expect the
Treasury response either to include the date or to give firm publication
dates for each agreement. We would expect the publication dates
to be before the start of the next financial year (paragraph 43).
The Government will publish the public service agreements for
departments later this year.
(u) We share the Government's view that "External scrutiny
plays a key part in ensuring that the providers of public services
are held accountable." To enable this external scrutiny to
take place, not least by select committees, it is vital that regular
assessments of performance against the public service agreementswhich
need not be in the form of Cabinet papersbe put in the
public domain. We recommend the Treasury give an undertaking to
publish this information and set out the mechanism by which this
will be done (paragraph 45).
The Government is giving further consideration to this proposal.
(v) We recommend that the Treasury set out which Minister is
ultimately responsible for the expenditure relating to each cross-departmental
initiative announced to date and in which vote the relevant expenditure
will appear (paragraph 46).
Each of the cross-departmental initiatives will have their own
PSA in which Ministers with lead responsibility will be identified.
HM Treasury, 26 October 1998
Report, Session 1997-98, The New Fiscal Framework and the Comprehensive
Spending Review, Volume 2, HC 960-II, p 71. Back
Report, Session 1996-97, The 1996 Budget, HC 129-I, paras
25-30; for Government Response see Second Special Report, Session
1996-97, HC 317, Appendix, paragraph 26. Back
p xiii. Back
1 approval involves the Treasury receiving trom departments: meaningful
illustrative Resource Accounts compiled in accordance with the
Resource Accounting Manual (RAM); a Departmental Resource Accounting
Manual, reflecting discussions on accounting policies with the
NAO, that deals with each department's own unique circumstances;
a letter from the departmental Principal Finance Officer confirming
that the department's accounting systems will be able to provide
the necessary resource accounting information; a copy of the NAO's
preliminary view on departments' accounting policies and systems;
and an updated project implementation plan. Back