Supplementary memorandum by the Clerk Assistant on behalf of the House of Commons Service (EST 08A)

Introduction

1. This supplementary memorandum is submitted for the Procedure Committee’s inquiry into scrutiny of the Government’s Supply Estimates in response to the letter to me from the Chair of the Committee of 20 July. It provides answers to the eight specific questions in that letter. The House of Commons Service remains ready to provide any further assistance to the Committee during the inquiry.

The rationale for fixing in Standing Orders the latest dates for agreeing Estimates resolutions, and the rationale for the two dates chosen

2. The original rationale for fixing dates by which Estimates motions must be approved if they are to be agreed to without debate (other than those previously subject to debate on Estimates days) lies in two imperatives, relating to the needs of business management in the House of Commons (and indirectly the House of Lords) and the need to ensure certainty and continuity in the funding of the State.

3. The date by which the outstanding motions to approve the Main Estimates must be approved if they are to be taken without debate is 5 August. As noted in paragraph 16 of my initial memorandum, the Votes on Account approved before the start of the financial year are intended to provide funding for the new financial year until the passage of the Summer Appropriation Act. The totals in the Votes on Account are usually set at 45% of the amount already voted for the corresponding Departments and services in the preceding financial year. This is usually sufficient to cover departmental expenditure on continuing services until the Appropriation Act is passed, and is posited on the assumption that the relevant Appropriation Act will be passed before the Summer recess.

4. The specific date of 5 August was first set in a Sessional order in 1896 proposed by Arthur Balfour as Leader of the House of Commons, which subsequently became a Standing Order. In the late 1880s and early 1890s the passage of the Appropriation Act had often not taken place until late August or even early September with both Houses continuously sitting until the Act had been passed. With no formal powers for the Chair to constrain debate on individual Estimates motions other than acceptance of the closure on each separate motion (which were over 200 in number) and on any amendments to each motion, debate could be extended almost indefinitely to frustrate progress on legislation or to enable further debate on policy and administration. The requirement to put all outstanding questions relating to the Main Estimates on a day no later than 5 August, followed by proceedings on the Appropriation Bill which usually took about a week, had the effect of ensuring that, from 1896, the Summer recess almost invariably started before 12 August. (It was remarked upon at the time that this enabled Balfour and other politicians so inclined to be on the grouse moors for the ‘Glorious Twelfth’.) In recent decades, the House has tended to rise for the Summer recess in the second half of July, so that the deadline of 5 August has not been directly engaged.

5. Although the rationale for the House to agree all motions relating to the Supplementary Estimates, Votes on Account and numbers in the armed forces and to give statutory effect to the authorisation of funds[1] before 31 March has never been in doubt (as discussed in paragraphs 11 and 12), the Balfour reforms did not establish a date on which the questions on all such outstanding motions could be put without debate. In the inter-war period, according to Winston Churchill, debates on Supplementary Estimates could consume five or six days in the early Spring and were “the most worthless of any that I have known in my career”.[2] A provision for all outstanding questions relating to the Supplementary Estimates and Votes on Account to be put without debate on a particular day was introduced in 1947 to provide certainty for the House and to avoid undue pressure on the House’s time in February and March. The date was originally set at 31 March, but was modified in 1968 to 25 March[3] and then (in 1973) to the current date of 18 March.[4]

6. The original rationale for both the August and March deadlines was to provide the Government with a certain date by which debate on the Main and Supplementary Estimates respectively could be concluded. The changes in 1982 providing that no debate can ordinarily take place on Estimates motions other than those on Estimates days mean that this original rationale has been largely superseded. Because the roll-up motions are no longer debateable, the Government can provide the necessary certainty about the timing of approval through its control of the scheduling of the business of the House. The deadlines now operate as a restraint on the Government’s timetabling of business for the benefit of the House, in that Estimates motions other than those on Estimates days can only be taken without debate if they comply with the deadlines set out in Standing Order No. 55 (and the other requirements of that Standing Order). However, there also remains a wider rationale for the inclusion of these deadlines insofar as they reflect an acknowledgement of the necessity to approve Estimates on a timetable that provides a reasonable assurance about the continuity of State funding.

The consequences should the House fail to pass motions to approve Estimates by these deadlines and whether in practice the dates could be postponed in a particular year by order of the House

7. The consequences of the House failing to pass the motions by the deadlines depend on the circumstances giving rise to that failure.

8. No procedural issues would arise if the motions to approve the Supplementary Estimates, the Votes on Account and the limits on defence numbers were not approved on or before 18 March, provided that it were a minor extension which did not jeopardise the passage of the Spring Appropriation Act by 31 March. There has been a past instance where the House has agreed a business motion to substitute a later date (prior to 31 March) for the 18 March deadline for the purposes of the operation of the Standing Order in a particular year.[5] In the event of the House not agreeing to such a business motion, the relevant Estimates motion or motions could still be taken on a date after 18 March in the course of that month, but would be debateable. It would be for the Treasury to advise on any impact of such delay beyond 18 March on the matters relating to the release of funds to Departments (this is considered further in paragraphs 11 and 12).

9. In the case of the 5 August deadline, the procedural position would be the same as with the 18 March deadline, namely that a business motion could be agreed to applying the Standing Order in a particular year substituting a later date in August for 5 August or the motions could be taken after 5 August without a requirement on the Chair to put the questions forthwith. The Treasury is best placed to advise on the risks to the continuity of the financing of the State in the event of the Summer Appropriation Act not being passed by early August.

10. The consequences would be of a very different character if there were a significant political prospect of either the Spring Appropriation Bill not being passed before 31 March or the Summer Appropriation Bill not being passed before the Summer recess.

11. In respect of the financial year about to end, the Spring Appropriation Act sets the final limits in relation to which departmental resource accounts are audited. Accounting Officers are expected to ensure spending is within the limits for the year ending on 31 March by that Act. In the event that the Act was not passed by 31 March, this would risk creating uncertainty and would imply an element of retrospection. Departments may not draw down cash in excess of the voted net cash requirement (arising from the existing Estimate) until the Supplementary Estimate has been voted. If approval to a supplementary net cash requirement were not granted until very late in the financial year or even after 31 March, any additional cash in excess of the voted limit in a previous Estimate required by a department would need to be drawn via an advance from the Contingencies Fund before the end of the financial year, requiring notification to Parliament in advance. The Treasury would be best placed to expand on this issue.

12. In respect of the coming financial year, the Spring Appropriation Act authorises the Votes on Account which enables the funding of departmental expenditure pending the approval of the Main Estimates and the passage of the Summer Appropriation Act. Although the House has sometimes sat on a Saturday to ensure approval of the relevant motions before 31 March, we can trace no instance where the 31 March deadline has not been met. The Votes on Account are contingent in nature, and do not prejudge final decisions on the quantum or distribution of expenditure. It is thus not evident that any political or constitutional purpose could be served by delaying passage of the motions giving rise to the Spring Appropriation Act or of that Act. The Treasury is best placed to advise on the consequences for the continuity of the funding of the State in the event that the Votes on Account were not given legislative effect by 31 March.

13. In the case of the Summer Appropriation Act, if an additional round of Supplementary Votes on Account were not approved, it is possible to imagine a “fiscal cliff” being reached in the late Summer or early Autumn when neither the Votes on Account nor any redeployment from other sources of funding available to the State which are limited in scale or scope (namely, the Contingencies Fund, the National Loans Fund, the National Insurance Fund and Consolidated Fund standing services) would be available. Again, the Treasury might advise on when. I would suggest, however, that this theoretical possibility is posited on a conflict between the political wishes of the Executive and those of a majority in the House of Commons which is it hard to see developing in the modern UK political context.[6]

The average period the Liaison Committee has to make a decision on subjects for debate on (a) the Main Estimates and (b) Supplementary Estimates etc

Main Estimates

14. There has been an average of 24 sitting days between the publication of the Main Estimates and the date of decision by the Liaison Committee over the last four years. The dates of publication of the Main Estimates, decision and debate in that period are set out Table 1 in Annex A.

15. There is no provision in Standing Orders preventing Estimates motions being taken at any time after the Main Estimates have been published, although at least two sitting days’ notice must be given of the motions. In practice, the time available for making decisions on subjects for Estimates day debates following Main Estimates is currently a consequence of:

• the commitment by the Government in 2009 to publish its Main Estimates within 5 weeks of the Budget Statement: this means publication of Main Estimates is usually in late April; and
• the established practice whereby Main Estimates are not debated and voted upon until late June or early July.

16. There thus can be up to 11 weeks between publication of Main Estimates and Estimates day debates, meaning there is up to 9 weeks between publication and a decision on debates. This normally equates to only around 20 to 30 sitting days between publication and decisions on debate because of the Whitsun recess and (since 2012) prorogation falling in the period.

17. At present, work on considering possible topics for debate does not necessarily need to await the publication of the Main Estimates themselves, although final decisions are not normally made until after publication.

18. It should be noted that the Estimates day in late June/early July does not generally take place in a General Election year, both because the Main Estimates are published later in such years (on 2 July in 2015) and because the Liaison Committee is not established in time. The delay to publication and the time lag prior to the establishment of Committees are likely to constrain opportunities for an Estimates day in the Summer in General Election years even if new arrangements were made for allocation of debates.

Supplementary Estimates

19. There is a much smaller gap between the publication of the Supplementary Estimates and the date of decision by the Liaison Committee. In two of the last three years, the decision has been taken on the day of publication. On the other occasion, the decision was taken three siting days later. The dates of publication of the Supplementary Estimates, decision and debate in that period are set out Table 2 in Annex A.

20. The Government seeks to publish Supplementary Estimates sufficiently late in the financial year to enable as many known factors, including up-to-date forecasts and decisions on Reserve claims, to be taken into account; and to tie in with the timetable for the Statement of Excesses, produced following the statutory deadline for the previous years’ accounts (31 January). In recent years, the Supplementary Estimates have been published in the second week in February (the week before the February half-term).

21. The deadline for the holding of the second of the two Spring Estimates days is 18 March, but both days have been scheduled in the first week in March in recent years. Standing Orders effectively require a minimum of 14 calendar days between the publication of the Supplementary Estimates and the second of the two Spring Estimates days, but this need not be a sitting fortnight. However, decisions have been reached on the allocation of debates almost immediately after the publication of the Supplementary Estimates to enable a fortnight’s notice to be given of the subjects for debate on Estimates days.

22. At present, work on considering possible topics for debate has to be undertaken before the publication of the Supplementary Estimates themselves, although final decisions are not normally made until immediately after publication. As a result, it is extremely unlikely that a proposal from a select committee or a decision of the Liaison Committee will arise directly from scrutiny of the Supplementary Estimates or any in-year departmental budget change reflected in those documents, unless that change was foreshadowed in information previously provided to a select committee. In short, the current timetable vitiates the prospects for linkage between the content of the Supplementary Estimates and the selection of the subjects for debate.

An indication of the number of weeks required to complete the process of “reasonable examination” of Estimates by the House, from laying to approval, as set out in paragraph 37 of the House Service memorandum

23. My initial memorandum set out six stages for “reasonable examination” of Estimates. The table on the next page gives the requested indication of what could be viewed as a “bare minimum” process and a more comprehensive process for the first four stages. Neither timetable allows for the conduct of an ordinary select committee inquiry, which might be considered to be a desirable element of an optimal process. I have included the final two stages referred to in that memorandum for completeness, but have assumed that the processes described there would be undertaken in parallel with the fourth stage.

Table: Number of weeks required for each stage of “reasonable examination”

24. Taking account of recesses,[7] the current timetable for the publication of the Main Estimates and subsequent debate aligns with the period required for the more comprehensive process. The current timetable for the publication of the Supplementary Estimates and subsequent debate falls well short of what is required for even the basic process. Further information about how the figures in this Table have been arrived at, and the assumptions made, are provided in Annex B.

The location in the Main Estimates and Supplementary Estimates of the figures for the Government’s proposed funding for the Scottish Parliament, Welsh Assembly and Northern Ireland Assembly

25. These are contained in the Estimates as follows

For the Scottish Government/Parliament:-

Scotland Office and the Office of the Advocate General Estimate

Part I: Non–Budget expenditure.  This is the total sum the Westminster Parliament votes for the Scottish Parliament. 

Part II: Non-Budget spending.  This gives a breakdown of the above amount into two elements:

1. Line E: Grant payable to the Scottish Consolidated Fund: This represents the cash sum payable in respect of the Barnett consequentials for Scotland, with a deduction for devolved taxes.  This amount is sometimes known as block grant.
2. Line F: payover of Scottish Rate of Income Tax to Scottish Consolidated Fund.  This is a new line from 2016-17 which represents the amount of income tax collected by HMRC on behalf of Scotland and paid over to the Scottish Government.

For the Welsh Government/Assembly

Wales Office

Part I: Non–Budget expenditure.  This is the total sum the Westminster Parliament votes for the National Assembly of Wales.

Part II: Non-Budget spending.  This repeats the amount shown in Part I, but showing it against:

1. Line C: Grant payable to the Welsh Consolidated Fund. This represents the cash sum payable in respect of the Barnett consequentials for Wales, with a deduction for any devolved taxes.  This amount is sometimes known as block grant.

For the Northern Ireland Executive/Assembly

Northern Ireland Office

Part I: Non–Budget expenditure.  This is the total sum the Westminster Parliament votes for the Northern Ireland Assembly. 

Part II: Non-Budget spending.  This repeats the amount shown in Part I, but showing it against:

1. Line C: Grant payable to the Northern Ireland Consolidated Fund. This represents the cash sum payable in respect of the Barnett consequentials for Northern Ireland, with a deduction for any devolved taxes. This amount is sometimes known as block grant.

26. The amounts of grant approved by the Westminster Parliament for each of the devolved institutions are cash amounts.  In each case the sums voted are included within the overall Net Cash Requirement sought by the Scotland, Wales and Northern Ireland Offices.  In fact they make up the vast bulk of each of the cash requirements within those Estimates. 

27. The devolved administrations may not draw down into their respective Consolidated Funds in a given year more cash than that voted by the Westminster Parliament.  In addition, HM Treasury operates administrative rules which require the respective devolved governments to manage spending on an accruals basis within Resource and Capital Departmental Limits (DELs) determined through the Barnett formula.  The cash amount shown and sought in the Estimate is the sum of the DELs set through Barnett, plus any Annually Managed Expenditure agreed by the Treasury, plus or minus any adjustments for items which require more or less cash in a given year than accrued spending (e.g. depreciation), less deductions in respect of devolved taxes.  This calculation is detailed in the accompanying Estimates memorandum.

28. The Estimates memoranda for the Scotland, Wales and Northern Ireland Offices’ Estimates also set out details of any additional Barnett consequential amounts or deductions made since the original Spending Review, as a result of fiscal events (Budgets and Autumn Statements).  The amounts are itemised to show the UK spending decisions which have generated the adjustment.

How Members could require the Government to account for the decisions it has made in determining how expenditure decisions in England have affected the calculation of funding for the devolved institutions

29. Ways in which this could be achieved might include:
1. Greater transparency over links between UK spending decisions and funding for devolved administrations;
2. Greater transparency over the calculations of the sums involved;
3. Better and more consistent explanation of the reasons changes to UK spending are being proposed;
4. Requiring UK Ministers to be scrutinised over their spending decisions, where they affect devolved administrations’ spending power.

30. At present, the Estimates memoranda for the Scotland Office, Wales Office and the Northern Ireland Office set out details of the sums allocated to devolved administrations via the Barnett formula for the current year as a result of fiscal events since the relevant Spending Review.  The memoranda do not
1. give a clear explanation of which UK level Estimates include the funding proposals (from either the Spending Review or subsequent fiscal events) which have led to the additional (or reduced) funding for the devolved administrations, or
2. show the specific arithmetic behind the amounts awarded (although the factors used are set out in the Statement of Funding Policy published at the time of a Spending Review).

31. Members or committees could therefore request further details of the relationship between UK level funding changes and the Barnett consequentials. If these details were included in the relevant UK department’s Estimate or estimates memoranda, rather than just those for Scotland Office, Wales Office and Northern Ireland Office, it would make the link between UK spending decisions and the Barnett consequentials more transparent. 

32. The table below is an example of how such information could be presented.

Main Estimate for Department for Culture, Media and Sport (illustrative, not actual)

Included within this Estimate are amounts allocated for spending by the UK government which have led to changes to funding for the devolved Parliaments.  Details are set out below:

All in £ million

33. A similar table could be included in UK departments’ supplementary estimates, which would show only amounts awarded at Budgets and Autumn Statements for the relevant year.

34. For further transparency, the UK government could also be asked to show the calculations underlying each amount, i.e. the comparability and population percentages used to arrive at Barnett consequentials in each case.  Because this could involve a variety of differing population percentages and comparability factors, and hence be quite lengthy, this might be better placed in the memoranda for the Scotland, Wales and Northern Ireland Estimates.

35. Existing information as to why the UK Government has decided to increase or decrease spending at a UK level can be patchy.  Sometimes Spending Reviews, Budgets, Autumn Statements and Estimates memoranda give explanations of what the Government is aiming to achieve by increasing or reducing expenditure in a certain area, or for a specific project or programme, but this is not always the case.  The Treasury’s current Estimates manual includes an Annex, written by the House of Commons Scrutiny Unit, which includes the requirement to “give details of any new programmes, as well as any that are being expanded or scaled back by 10% or more.  The memorandum must explain the underlying reasoning behind such spending decisions and cannot simply refer to the mechanism through which the decision was made or who took the decision”.  It is for committees and the Scrutiny Unit to enforce this requirement, and the requirement itself could be made more specific to the needs in relation to Barnett consequentials, i.e. requiring explanation whenever a Barnett consequential is generated.

36. Building on this, it might be possible for UK Government Ministers to be answerable to the Scottish, Welsh and Northern Ireland Affairs Committees for decisions the UK Government has taken which affect levels of funding for those administrations, perhaps through hearings at the time of the Spending Review.

How early before a designated Estimates day would it be feasible to publish motions for the approval of each estimate, in order to allow for amendments to be tabled

37. The text of motions for the approval of individual Estimates (as well as the roll-up motions) are currently prepared by the Clerk of Supply in the Public Bill Office, and then checked by HM Treasury, before being passed to the Government for tabling. The content of the motions is derived from the content of the Main and Supplementary Estimates and so the motions could be prepared and checked within two working days of the publication of the Estimates or Supplementary Estimates concerned. Under the current timetables mentioned earlier, this would be around 30 sitting days before the debate in the case of the Main Estimates and around 10 sitting days before the earliest debate in the case of the Supplementary Estimates.

38. It should be noted that, in compliance with the rules relating to the financial initiative of the Crown described in paragraph 1 of Annex A to my initial memorandum, these motions can only be tabled by a Minister of the Crown. The timing of the tabling of the motions is therefore ultimately a matter for the Government; the motions are usually tabled at present between 1 and 4 sitting days after the House has agreed to the proposals from the Liaison Committee on the allocation of Estimates days. If the Procedure Committee were minded to propose a change of practice including early publication of motions to allow for amendments to be tabled in order to serve as an indicator of interest in possible Estimates to be debated, such a proposal would therefore need to be directed to the Government in view of its responsibility for tabling the motions. It would also probably be sensible to confine such early publication to the Estimates for the main Ministerial departments, unless there were clear evidence of a wish for a debate or amendment relating to other Estimates.

Annex A: Tables

Table 1: Dates of publication, decision and debate: Main Estimates

Table 2: Dates of publication, decision and debate: Supplementary Estimates

Annex B: Calculation of number of weeks required for each stage of “reasonable examination”

(a)              Review of each of the Estimates and their content, in order to understand what is proposed

1. Any review of the Estimates is likely to take a certain amount of time.  This is simply to absorb the mass of information provided, and to understand what is proposed and why. A typical set of Estimates will run to over 600 pages, and contain up to 60 individual Estimates. 
1. Of these around 20 relate to major government departments (such as the Home Office, each generally led by members of the Cabinet);
2. A further 16 or so relate to smaller government departments or equivalent bodies, such as the Food Standards Agency;
3. A further 5 relate to regulators such as Ofgem and the Office of Rail and Road;
4. There are a further 8 which relate to occupational pension schemes;
5. The remaining estimates relate to the parliamentary and independent bodies, including the House of Commons and the National Audit Office.

Each of these is accompanied by an Estimates memorandum, possibly 20 pages in length in each case.

2. Furthermore, to make analysis meaningful, it is generally necessary to look at these documents in conjunction with others, such as past spending patterns, past announcements, policy and legislative changes, aspirations and targets, Committee inquiries and reports, NAO reports etc.

3. A high-level, but not detailed, examination of the documents for the major departments would be likely to take the current Scrutiny Unit finance team of five at least a working week to digest and summarise headlines in written form.  This might do little more than identify key items of interest and the headline numbers.

4. A more comprehensive analysis of major departments including visual graphic presentations for the major departments, would be likely to take around three working weeks in total with current resources.

5. The estimate for this latter process of three weeks assumes that examination would be confined to the major departments mentioned at paragraph 1(a) of this Annex. Any comparable examination of the Estimates mentioned at paragraphs 1(b) to (d) would extend this period by at least one week.

(b)              Consideration by Committees and/or officials as to what might warrant debate;

6. Having carried out the analysis, there needs to be a process for committees to consider what might warrant debate. This could take account of analysis undertaken above. At present consideration of topics for debate is essentially a separate exercise which is not dependent on analysis of the Estimates, but on Committee reports, so can be undertaken, as indeed currently happens with Supplementary Estimates, before the Estimates are actually published. If instead the Estimates themselves formed the real basis of the debate, I have presumed that consideration by the House or Committees would have to await analysis (stage (a) above) first.

7. Following this analysis:

• Officials (e.g. the Scrutiny Unit) could suggest to Select Committee(s) possible topics for debate, perhaps based on specified criteria, and the analysis above; and/or
• Select Committees could be invited to make suggestions for topics for debate; and/or
• Wider circulation of Scrutiny Unit analysis to Members of the House as a whole generally could enable Members to consider whether to suggest a topic for debate.

8. The bare minimum required for these steps would be two weeks, but a more comprehensive process would require three weeks, particularly if papers were to be circulated for consideration by select committees or Members more generally.

(c)              Decisions to be taken by the House or its committees on what to debate

9. Following receipt of bids, the Liaison Committee (or other committee given such a responsibility) would need to schedule and hold a meeting to agree which topics should be selected for debate. I understand that this decision has on occasions been taken without a meeting by the Liaison Committee, but members of that Committee still need due time to consider a proposal. If the role were to be assigned to the Backbench Business Committee, that Committee already meets on a weekly basis to reach decisions on debates.

10. I have therefore estimated that the process could be undertaken in about a working week, although a fortnight would allow for a more comprehensive process, with more consultation and more opportunity to hear representations on competing bids and assess the support for and merits of each bid.

(d)              Advance warning to be given to Members of topics/estimates selected for debate to enable them to plan whether to attend and contribute

11. There is a minimum notice period for an Estimates day under current procedure of just under a week necessitated by the need for the House to agree the proposal from the Liaison Committee and the need to give notice of the motions then tabled. Given the accompanying need for reasonable notice to be given by way of a Business Statement, I have equated this to one working week as the bare minimum. It is clearly highly desirable for the announcement to be in the preceding Business Statement, covering the week after next, to allow for Members wishing to participate in a debate more time to arrange their diaries and prepare adequately, and to facilitate public awareness of an engagement with the debate. This therefore equates to two weeks for the more comprehensive process. An even longer period of notice would clearly be of benefit to Members wishing to participate and the wider public.

(e)              Preparation of any briefing, whether to Members generally or to Committees, in support of any debate

12. At present, topics selected for debate are almost invariably derived from select committee reports. Select committee chairs and members are likely to be well-sighted on the topic, and they, along with Members of the House more generally, can read the Report(s) and evidence, and the Government response(s) to learn more. Committee staff may prepare a brief on the topic in question. The Library does not normally prepare a debate pack.

13. If estimates or topics for debate were chosen other than on the basis of work by select committees, the Library or the Scrutiny Unit might be required to prepare briefing or a debate pack. I have assumed this work would be undertaken in parallel with stage (d), although the quality of this material provided would necessarily vary depending on the time available.

(f)              Time to resolve any issues arising from scrutiny, debate or amendment of the estimates (particularly if the process is amended)

14. This stage in my original memorandum alluded to any additional requirements from the House Service or Members of the House relating to a new process. This might relate to the provision of information to the House (for example on consequences for spending in devolved areas or relating to amendments). Again, in timetabling terms, this does not affect the overall calculation because any such activities would necessarily take place between the debate being announced and the conclusion of the debate itself.

September 2016

1