ࡱ> 7Root Entry` F:h UKnWordDocumenty v`8iCompObjQRD Cv%j'SummaryInformationy (;  9  !"#$%&'E>?@ABCDFGHIJKLMNOPQ5@dL@jn@<.h@<_+n ]G՜.+,0HPt|  Dell Computer CorporationG$  Dear Wendy, FMicrosoft Word Document MSWordDocWord.Document.69qOh+'0 , T ` l x Dear Wendy,bPreferred CustomerNormalPreferred Customer21Microsoft Word for Windows 9Root Entry` F:h UKnWordDocumenty v`iCompObjQRD Cv%j'SummaryInformationy (;  9(  !"#$%&'.)*+,-/0123456:;<=ܥhc $eYiVbbbbbbbcNc6cccc ccFh1cc(cccccc4d6d6d6d=sd fgwhXhCgbc-.ccccgcbbccccccbcbc4d UKnb,cJbbbbc4dcHcWendy Alexander MSP Dear Wendy, An Open Letter on GERS In your capacity as Convenor of the Finance Committee, you kindly invited us to attend its meeting on 16th January, which was to consider the topic of GERS - the Scottish Executives annual publicati GERS, there is an accounting adjustment to convert from Total Expenditure on Services (TES) to Total Managed Expenditure (TME). At the UK level the accounting adjustment in 2004-05 is 22.7 billion, of which 11.9 billion is non-market capital consumption of central government and local authorities: (NMCC is depreciation). As part of the GERS exercise, the Scottish Executive requests each November from Regional Accounts in ONS estimates of Scotlands NMCC. The ONS figures imply that 15.7% of NMCC is attributable to Scotland. When we asked ONS for a justification why this figure was so much higher than Scotlands 8.5% population share, it became clear that the reason was that ONS had allocated to Scotland a 32% share of a large general category within public administration and defence. ONS could offer no explanation for this 32% allocation to Scotland, writing as follows: Investigating the origins and rationale underlying these proportions has led to the delay in responding. Despite a concerted effort entailing trawling through historic documentation and databases, and contacting individuals involved in the production of the accounts at the relevant time (more than ten years ago in some instances) it has not been possible to arrive at a detailed explanation for the basis of the proportions used. (ONS email to M Cuthbert). Given this, there must be considerable doubt about Scotlands assumed 15.7% share of UK NMCC. If, in the absence of other evidence, an 8.5% population share was used to attribute this category to Scotland, this would reduce estimated NMCC for Scotland in GERS from 1.867 billion to 1.013 billion, that is, a reduction of around 850 million. Note, however, that since NMCC is a non-cash item which appears in both the revenue and expenditure sides in GERS, this change would have no effect on the overall gross GERS balance. It is, however, of great relevance to certain important presentational issues, which we discuss further below. c). Problems with the Allocation of the Residual Accounting Adjustment As noted at (b), an accounting adjustment is added to TES to arrive at TME. Once NMCC is subtracted, the residual accounting adjustment at UK level amounts to some 10.8 billion. In GERS, 9.56% of the residual is attributed to Scotland: the 9.56% factor used by the Scottish Executive is their estimate of Scotlands share of UK TES. We wished to probe the rationale of this attribution: but before we could do this we needed to understand what the residual category actually consisted of, and the size of the component parts. The Treasurys written response to a request for a breakdown of the UK accounting adjustment into its different components was as follows: Key differences between TES and TME are that TES does not include general government capital consumption, timing differences in data and some different data sources. These, and others, make up the accounting adjustments. The quantification of these differences are not something we hold centrally, and due to disproportionate costs cannot obtain. (Treasury email to M Cuthbert) Given that the accounting adjustment is 22.97 billion and represents 4.63% of UK TME, we find this response incredible. In the light of the Treasury statement, there must be significant doubt, both about the absolute size of the UK residual after deducting NMCC, and about the logic of applying a 9.56% attribution of it to Scotland in GERS. If the residual were attributed on a population basis, then this would reduce Scotlands gross GERS deficit by 114million. We are not, however, simply advocating that this change should be made: rather, what is required is a much more detailed analysis of what the UK residual actually comprises, upon which an informed judgement about choice of appropriate attribution factors for Scotland can be made. d). Errors in the Assessment of Gross Trading Surplus in GERS On the revenue side of GERS, one of the items is Gross Trading Surplus, Rents and Misc. This comprises the gross trading surpluses or losses of central government, local authority and public corporation trading activity, plus the gross operating surplus of the non-trading activities of central and local government, plus the rent income of central and local government. In November each year, the Scottish Executive requests a breakdown of this category from ONS. In clarifying exactly what figures ONS produced on this each year, we established that ONS have apparently been rerunning for the Scottish Executive an old analysis which ONS no longer publish as such: and this analysis double counts local authority housing revenue account surplus, amounting to almost 3.4 billion at the UK level. As ONS wrote, the series .. was in an old table that is no longer published. This series seems to have continued being calculated despite not being used, and hasnt kept pace with changes - it double counts ADAE. (ONS email to M Cuthbert). ADAE is local authority housing revenue account surplus. This double counting masked another error in the gross trading surplus etc. category used by the Scottish Executive. In 2004-05, the Treasury allocated the revenues of TV licences to the UK equivalent category. This was missed by the Scottish Executive. The amount involved for the UK was around 1.7 billion. The net effect of the above errors was that at the UK level, the gross trading surplus etc. category used in GERS was 22.668 billion, overstating the value from the Treasury budget statement of 21.6 billion. Correcting for these errors would increase the gross GERS deficit for Scotland by approximately 200 million. e. Questionable Attribution of UKAEA Expenditure to Scotland A point we noted in our earlier published critique of PESA, but which has not been addressed, is the attribution of United Kingdom Atomic Energy Authority expenditure, mainly on decommissioning. At present, this expenditure is regarded as identifiable in PESA, and 157.18 million is identified to Scotland, representing 44% of UK expenditure: (this primarily relates to decommissioning at Dounreay.) The size of this item will increase rapidly, as decommissioning expenditure increases in line with current plans. The identifiability of this expenditure to Scotland is clearly inconsistent with any reasonable interpretation of the who benefits principle of identifiability which is used in PESA. The UK atomic energy programme is a whole UK priority, like defence: and, like defence expenditure, should be treated as non-identifiable in PESA. Allocating UKAEA expenditure on a population basis would reduce Scotlands gross GERS deficit by 127 million. Similar issues also arise with other aspects of the atomic energy industry, although the effects are smaller: for example, support for British Energy, where 13.7% of current expenditure is allocated to Scotland; and nuclear liabilities management where again 13.7% of expenditure is allocated to Scotland. f. Questionable Allocation of Identifiable Expenditure by Whitehall Departments. One of the major benefits of the analysis of the detailed database underlying PESA, is that it enables scrutiny of the assumptions which Whitehall departments make, in attributing identifiable expenditure to Scotland and to the other countries/regions of the UK. It is incredible that such scrutiny only became possible, both for the public and non-Treasury departments, after we secured the original release of the data in 2005. The latest PESA database is particularly interesting, since the database now contains a time series of 5 years data. Scrutiny reveals several examples where Departments have used the same factors either over several years, or over several categories of expenditure - or indeed over both. It also reveals several examples where the factors used appear odd. To give only a few examples:- i). The six Research Councils: in 2003-04, the DTI attributed 13% of capital and of current expenditure of each of five research councils to Scotland. BBSRC, the sixth research council had 11% of both its capital and current expenditure attributed to Scotland in every year from 2001 to 2005. All this suggests that a rule of thumb is being used in attributing this expenditure rather than an attribution based on accounting for actuals: given that the total of expenditure involved amounted to almost 1.2 billion in 2004-05, a rule of thumb attribution could lead to serious inaccuracy. ii). The Department of Works and Pensions attributes a flat annual rate of 10.7% of its corporate services budget, the Child Support Agency, and the Appeals Service to Scotland. We clearly do not know what the correct attribution should be: but the application of a flat rate for each of five years and each of three very different services again suggests that a rule of thumb is being used. To give an indication of the potential inaccuracy resulting, given that the three categories of expenditure amounted to 2.25 billion in 2004-05, then changing the proportion to a population share would have a 49.5 million effect each year on the amount attributed to Scotland. DWP attributes 14.4% of its Health and Safety Executive expenditure to Scotland, a flat rate which has not altered in five years. Similar issues arise with the attribution of the administrative costs of measures to help unemployed people from welfare to work, and for the Job Centre Plus Agency, both of which are given a flat annual Scottish share of 13.2% (for total expenditure of 1.9 billion in 2004-05). Again, this is not an appropriate approach when total expenditure is so large. As regards the JobCentre Plus social provision, a flat 13.63% has been attributed to Scotland for three consecutive years. Again this suggests a rule of thumb is being used, in attributing a budget which was 1.25 billion for the UK. These are just a selection of examples. Similar examples can be found in many other Departments. The overall conclusion we draw is that Departments have put too little effort into the attribution of identifiable expenditure between countries/regions: and that the Treasury, who co-ordinate the exercise, have put too little effort into scrutinising the data. This also raises the further issue of why Scottish Executive civil servants did not insist at a senior level on obtaining access to the detail of the PESA data base before using it in GERS: and why it was left to outsiders (ourselves) to secure this access. Conclusions. We draw the following conclusions. 1) The above amply demonstrates that there are very large errors, and areas of uncertainty, in GERS. The effects identified at a), c), d), e) and f) above, not all of which point in the same direction, are large when compared with a gross deficit consistent with the latest GERS publication of about 6bn: (that is, the crude deficit from the published GERS of 6.5bn, assuming 90% of oil revenues are attributed to Scotland, further corrected for the already identified and acknowledged error in the handling of expenditure non-identifiable within England.) Further, the error identified above in the estimation of non market capital consumption, (see b) above), while it does not affect the gross deficit, affects the more important current deficit to the extent of a potential 850 million or so. Overall, the gross effect of the errors and uncertainties in the current GERS is too large for GERS to carry the weight it is being asked to carry in the current political debate. This conclusion contrasts sharply with opinions expressed by several of the invited panel at the Finance Committee meeting on 16th January. For example, Anton Muscatelli said that Given the parameters that the exercise has set itself, what we have before us probably represents the best that can be done. Peter Wood said, As far as the technical improvement of GERS is concernedit must be nearly as good as it will get, especially as regards the facts of the case on the expenditure side. We have heard about the methodological improvements that have been made on the revenue and tax-raising side and I doubt whether significant further progress will be made on that. Such opinions led Elaine Murray MSP to say the view I am gathering from the evidence that we are taking in the discussion is not that GERS is so inaccurate as to make it useless and a source of major errors, as has been described, but that it is useful. The evidence given above calls such opinions into question. 2) We must be clear, however, that the topics we have listed above do not represent all of the remaining errors and uncertainties in GERS. We are aware of further significant problems in those areas we have looked at in detail: but there are other areas we have not considered in detail which require investigation too. So what is required is a comprehensive review of GERS, rather than just attention to the above points. 3) Given the errors and unresolved issues with GERS, the quality of the data appears inconsistent with its publication under the national statistics logo. We also find it unacceptable that GERS was first designated as a national statistic after we had identified the error in the handling of expenditure which is non-identifiable within England: (we published the paper in which this error was first identified in November 2005): and that GERS has now been published for two years under the national statistics designation, with this error uncorrected. (You will recall that this error is estimated to add 440 million to the crude GERS deficit for Scotland in 2004-05). 4) The current difficulties with GERS point to serious problems with the way the production of GERS is handled within government. We are in touch with the relevant staff in the Scottish Executive, (whom we have informed about the issues identified above), and acknowledge their willingness to seek to improve GERS. But responsibility for important inputs to GERS rests with other departments. The Treasury is responsible for PESA, but has in the past been far too secretive, (for example, in not releasing the detailed PESA database to other departments, until we secured its public release under Freedom of Information): the Treasury has also failed to exercise adequate scrutiny of the key attributions of identifiable expenditure made by Whitehall departments, and which are fed into PESA. ONS produce other key inputs to GERS: but it is unacceptable for this to be done in a handle-turning way, producing an analysis which has come to double count significant items: (see (d) above). Improving GERS goes beyond questions of resources: it requires a much clearer focus of responsibility than exists at present. And it also requires senior civil servants in the Scottish Executive to adopt a more pro-active role as regards the Treasury, ONS and other departments: (for example, senior civil servants in the Executive should have been willing to correct the mistake in GERS arising from the handling of non-identifiable within England expenditure, without waiting until the Treasury gets round to correcting PESA: and they should be more willing than they apparently are to argue over questionable attributions like that of UKAEA.) 5) But improving GERS is not just about organisation, will, and resources. There are also key presentational issues which need to be addressed. For example, one important issue which has led to widespread confusion relates to presentational differences between GERS and the Treasurys Financial Statement and Budget Report, (FSBR). In GERS, gross revenues are compared with total managed expenditure, to give a gross balance. In the FSBR, the gross balance for the UK is split into two components, namely the current budget or balance, representing the difference between revenues, and current expenditure, (including depreciation.) net investment, representing the difference between gross capital expenditure and depreciation: this represents the creation of net new capital assets. The subtraction of these two items in the FSBR gives net public sector borrowing: it is this which is equivalent to the gross balance which GERS attempts to calculate for Scotland. The importance of the current balance is that this is the quantity which should average to zero over the cycle if the golden rule is to be respected. Unfortunately, much common which attempts to estimate public expenditure made on behalf of Scotland and revenues raised from Scotland. Your invitation, as you indicated, was in recognition of our diligence in improving its accuracy over the years. For reasons we communicated to you privately we were unwilling to accept your invitation. However, we think it appropriate to share our latest findings with you and the public by means of this Open Letter. Our last published analysis of the detailed database underlying the Treasurys Public Expenditure Statistical Analysis (PESA) had, as you know, identified a serious error in the handling of expenditure which is non-identifiable within England: the existence of this error is now fully accepted, (though not yet corrected in GERS or PESA), and has the effect of overstating the estimated deficit in GERS by 440 million. This January, we obtained from the Treasury the latest database underlying PESA, which now covers the years 2000-01 to 2004-05. We have, since then, done extensive work on specific aspects of this database, and the associated GERS calculations carried out by the Scottish Executive for the version of GERS they published in November 2006. The findings reported on below relate almost exclusively to our latest work. Specific findings arising from our current work include the following: a) Error in Handling Gross Trading Surpluses of Public Corporations in GERS On the revenue side in GERS, one of the items attributed to Scotland by the Scottish Executive is 8.2% of the gross trading surplus of UK public corporations. This will cause a problem if there is a public corporation operating in Scotland which has no counterpart in the rest of the UK. One important example where this happens is with respect to water, where Scottish Water is a public corporation, but where the water industry has been privatised down South. On the expenditure side in GERS, 527.4 million gross capital expenditure on water is attributed to Scotland for the year 2004-05 in the PESA database. On the revenue side in GERS, as we have seen, Scotland is allocated 8.2% of the gross trading surplus of UK public corporations. If the rest of the UK had had a publicly owned water industry, then this approach, while crude, would not have been orders of magnitude wrong: Scottish Waters actual capital expenditure would have been balanced against an 8.2% share of the UKs water industrys gross trading surplus. However, since the water industry in England and Wales is privatised, what will appear in the revenue side in respect of water in GERS is essentially just 8.2% of Scottish Waters gross trading surplus - rather than Scottish Waters full gross trading surplus. Given that Scottish Waters gross trading surplus is about 486.5m, (and after allowing for some further complications around interest payments), the effect of this error alone will be that GERS currently overstates Scotlands final deficit by over 320 million. b) Serious Problems in the Calculation of Non-Market Capital Consumption for Scotland In PESA, and inent on GERS compares the headline gross balance in GERS with the current balance in the FSBR. Under such an ill-informed comparison, the whole of the GERS gross balance is regarded, in popular parlance, as a black hole, or a union dividend: whereas a country operating with perfect prudence would expect to have a current balance averaging to zero, but a gross deficit, (normally met by borrowing), equal to the net creation of new capital assets. We conclude that an improved GERS should adopt the FSBR presentation: which is one reason why the error in the estimation of Scotlands NMCC, ( (b) above), is very important. Adopting the FSBR presentation, however, would have further implications. The FSBR also presents a data series showing the net worth of the public sector: that is, the excess of the worth of the physical assets owned by the state, over the financial debt of the public sector. This series is a key element in demonstrating that the borrowing of the government to finance its gross deficit, (and, as we have seen, a country operating with perfect prudence will normally have a gross deficit), is matched by an at least equal growth in the worth of the asset stock owned by the state. GERS should be extended to include a series on net worth comparable to that in the FSBR. Prudent management of the public sector cannot be demonstrated without this full presentation. Overall, we conclude that the following are required: a) that the quality of the basic data going into GERS is radically improved. b) that there is a much clearer rationale for deciding on what basis expenditure is attributed to Scotland in GERS. c) that there is a much clearer focus of responsibility for the production of GERS. d) that the presentation of GERS is extended, to show the current balance and net new investment, and also an FSBR type series on net worth. If these changes were made, GERS could play a key role in demonstrating whether or not the public sector and the Scottish economy were being adequately managed. However, even if, these improvements were made to GERS, there are limitations as to the uses to which GERS should be put. In this respect, we very much agree with the views expressed at your Finance Committee meeting by panel members and the senior civil servant, that GERS says little about how an independent Scotland would perform. Yours sincerely, Jim and Margaret Cuthbert. Edinburgh, 24th March 2007. PAGE 7 PAGE 2  .A 7 7"8@ w(DEO~".#/#F#&Q&F++88v=x==(>:>??@{Y}YYYYYYYYYYYYYYYYYYYYYYYYuPaP uDPUVh^="9= GwE%P ~"#&&R&B+C++.168z z z z z z  z z z z z z z z z z z z z z z z z z z z z z z z z z z z  z z z z  4'8889 9B<<@@BB#E$EIKKLMNNOQsRsUtUUUmVVNW?Y@YQYlYmYz z z z  z z  z z z z z z  z z z z z z Ez Ez z z  z z z z z z z z z z z z !  4 4"mYyYYYYYYYYYYYYYYz z xz z z z `% 4 ܥhc eYiVbbbbbbbzcNc,zczczcc czc