Finance Committee, 25 Oct 2005

Meeting date: Tuesday, October 25, 2005

Official Report 326KB pdf

---

Council Tax Abolition and Service Tax Introduction (Scotland) Bill: Financial Memorandum

Item 2 is consideration of the financial memorandum to the Council Tax Abolition and Service Tax Introduction (Scotland) Bill. I apologise to Tommy Sheridan and Gordon Morgan for keeping them waiting. They will appreciate that we were receiving some interesting evidence, part of which they sat through. Gordon Morgan is a researcher for the Scottish Socialist Party and I think that we all know who Tommy Sheridan is. After Tommy has made a short opening statement, we will proceed to questions from members.

I spent several hours working on my introductory remarks, but those have now been thrown asunder, because they started with "Good morning". It will now have to be "Good afternoon".

The financial memorandum, of which all members have a copy, is largely robust, but it includes some typographical and statistical errors. Those are referred to in the Scottish Executive submission that members have in front of them. The financial memorandum was based on six rates of taxation to be applied for the service tax, whereas the bill is based on five rates. In other words, a median rate of 8 per cent was removed from the bill, so the generated surplus is different from that which is suggested in the financial memorandum. The important point to make is that the Scottish Executive accepts that the surplus that is generated is still substantial—some £313 million.

I turn next to the provisions of the bill that deal with extinction of poll tax debts and council tax debts in excess of two years. The committee has received a supplementary paper that provides evidence to back up the argument that is made regarding the revenue that would be lost to the Scottish Executive, which is in the region of £100 million to £150 million. Gordon Morgan has worked hard to extract that information and will be only too happy to deal with detailed questions on the matter. It is an issue of 0.2 per cent here or there. We believe that the memorandum is robust in that respect.

My final point is that the majority of the costings that have informed the memorandum and the bill have been taken from the work of the Chartered Institute of Public Finance and Accountancy in England. Its report on improving the balance of funding is considered to be the most up-to-date and reliable paper for estimating the costs of removing council tax and replacing it with a form of local income tax.

We have tried at every turn to disaggregate figures for Scotland. I argue that we have not considered the lowest cost but have used a less than generous set of assumptions when estimating costings. I hope that individuals will bear it in mind that one of the biggest criticisms of local income tax that CIPFA makes consistently relates to the administrative set-up costs across the 400-odd local authorities in England and Wales. However, in paragraph 6.34 on page 35 of its paper, it states:

"A LIT would be a simpler and more realistic proposition if based on a universal structure of unitary authorities like that serving Wales and Scotland."

In other words, the administration and set-up costs would be much fairer if a local income tax were set up across the whole of Scotland, rather than in the way that is proposed by the Liberal Democrats in England and Wales, which would require each local authority to set up the tax.

I invite questions from members either to Gordon Morgan or to me.

Just before we move to questions, we should bear in mind a couple of points. First, the Finance Committee is not concerned with the policy approach; we are purely concerned with the details in the financial memoranda. Gordon Morgan's submission about council tax to the Local Government Committee appeared before us only today. It might be that some follow-up questions will arise from that submission, which we will pursue in correspondence. However, if members have had a chance to absorb all that information, they can ask questions today.

What is the basis for your estimate of the start-up costs as being between £3 million and £10 million?

A lot of the costs are in the CIPFA paper. Much of the work would be in the development of information technology systems, such as the Sage accounting package, in HM Revenue and Customs as well as in commercial firms and others that would be obliged to have a second set of accounts as part of a general tax programme for Britain.

From our informal talks with officials—we have not had an official response from the Inland Revenue—we know that much of the preparation work on the variable income tax for the Scottish Parliament has already been done. It says in the CIPFA paper that the largest costs that are associated with the work are for the determination of address. If there were to be a variable income tax in Scotland, whether someone works or lives in Scotland would be one of the key factors in determining whether they would be liable to pay the variable income tax rate. We understand that that work has already been prepared. The Inland Revenue changes represent the most substantial work. There would be an ability to include additional codes as there is in regard to budgets.

I speak with 30 years' experience in IT, in which I had responsibility for fairly large projects. The basis of the estimate includes some of the CIPFA figures, but it also takes cognisance of the work that we have already done. I do not believe that there would be substantial costs to the Inland Revenue in the development of the work.

Employers would simply have a set of codes, which would normally be in electronic form but would perhaps be in book form. That one set of codes would cover the whole of Scotland and replace the existing sets. There would be no administration costs for them. We are assuming that local authorities would enter into an agency agreement with the Inland Revenue. Local authorities would need to enter receipts from the Inland Revenue and the numbers in each local authority area would have to be determined. It is obvious that we need to go into the details. The CIPFA paper refers to approaches that are taken in other countries, such as Japan. If one takes an aggregation approach throughout the country, one can determine the tax on the basis of a person residing and working in one local authority at the beginning of the year and then make a general assumption on that.

We reached the estimate based on those considerations—it is in the ballpark of £2 million to £10 million, depending on the precise choice of scheme and detail that would be implemented.

I want to pursue you on a couple of those points. The devolution white paper in the late 1990s set the cost of introducing the income tax-varying power at £10 million. What you propose is more complex than what was proposed in that white paper because you suggest introducing six income bands and different tax rates. You also have a distribution factor between the local authorities—I presume that the revenues would need to be allocated on the basis of the tax collection or that at least some assignment would have to be made. The process that you have set out is more complex, but your top-end figure is the same as the figure that is given in the white paper for introducing a simpler scheme.

The key element is the postcode, because identifying where a person lives will tell us which local authority area they live in. Local authority disaggregation will be based on information that is available from the Inland Revenue scheme that I am assured is already in line to deal with the variable income tax rate. As a result, our proposal has close identifications with the local income tax scheme.

Although there would be six tax bands, there would be only one set of tax tables for the whole of Scotland. In some cases, separate entries might have to be made for national income tax and the service tax. However, the computer system will have to deal with only one set of tax tables, which will give two totals at the end. Although a little more programming will be involved, the scheme itself is still simple and unitary.

From our informal chats with Inland Revenue representatives, who will have a chance to get this information on the record when they speak formally at the Local Government and Transport Committee meeting on 8 November, we know that much of the set-up, recoding and sheer admin work that the costs highlighted in the Scotland Act 1998 referred to had already been carried out in the event that the Parliament chose to use its tax-varying powers. As a result, the £10 million cost in the 1998 act is not at the top end at all. The tax collectors we have spoken to have told us that, once implemented, the system will not be costly because they will provide largely the same service that they are already providing for income tax collection.

You have set out assumptions about the collection of tax either by local government or by the Inland Revenue. Why have you assumed that there will be no cost to employers if the service tax is collected by the latter?

Simply because, as Gordon Morgan and I have already stated, employers will receive a new set of tax table instructions. They will not have to carry out any extra calculations. All Scotland-domiciled employees, including those who live in Scotland but work in England, will have a separate code that will identify them as such. Inputting such information should not require any extra IT staff to be employed.

That view was certainly not shared by Confederation of British Industry representatives when they spoke to us three years ago about a 3 per cent increase in tax. We should perhaps cross-reference your comments with what they said.

You have assumed that if local authorities collect the tax, the cost to employers will be between £87.4 million and £91.8 million. How did you arrive at those figures?

As that question brings us back to the CIPFA paper, I ask Gordon Morgan to answer it.

The financial memorandum contains two tables, the first of which comes after paragraph 17 and the other after paragraph 19. The first table, which is entirely lifted from page 45 of the CIPFA paper—I have a copy if the committee wishes to see it—identifies the cost of collection by local authorities at £2,320 million to £2,460 million if we factor in the extra Inland Revenue cost of supplying information to them. There would be substantial cross-traffic to local authorities if they were to collect the revenue and set up collection mechanisms. They would, in essence, have to have duplicate sets of Inland Revenue-related information and there would be an enormous cross-check.

We have taken the figures and disaggregated them for Scotland, based on proportionality. We have assumed 8.46 per cent—this is in paragraph 18—for Inland Revenue administration and between 7.3 and 8.36 per cent for local authorities. We have taken the costs and applied them down to local authority level. The cost for local authority staff to collect the Scottish service tax comes to £67 to £70 million.

Although there is not an awful lot of movement between Scotland and England in regular work, people move frequently between local authority areas in Scotland. With such movement there would have to be multiple returns in relation to the coding structure, so authorities might be more interested if a person was in their area for part of the year and in another area for another part of the year. If there were variable rates, people might be encouraged to move their residence to another area. There would be a substantial additional burden on employers of form filling. We have assumed 8.2 per cent of the national figure if the local authorities collected the tax. The aggregate of Inland Revenue, local authority and employer costs comes to £161 million to £180 million, based on the CIPFA estimates.

You have made an eloquent argument about why it is not possible to collect the tax on a local authority basis. The costs and complexities would be so unmanageable that local authorities could not collect it. Is that a fair conclusion?

Gordon Morgan can reply too, but I just want to make the point that there is no hiding the fact that the proposal is related strongly to the suggestion that local authorities should contract to the Inland Revenue for the collection. We are not saying that that is an optional extra; there is no doubt that it is an important element in the success of the bill. If local authorities refused to contract, the costs would be large. However, with the £313 million extra revenue that would be generated, we would still generate more money than the council tax is generating. Less surplus would be generated, but we would not be moving into a deficit position in funding the bill.

I shall leave the issue of surplus to one side. I just want to be absolutely clear. Gordon Morgan talked about the costs and complexities associated with collection by local authorities. He was eloquent in setting out some of those complexities and I am sure that others could be identified. One always ends up with additional problems that were not envisaged when one introduces a new scheme of this kind. As Tommy Sheridan summarised, the practicalities of the proposal are such that the tax could be collected only through the Inland Revenue and not by local government acting on its own.

My only dispute is over your use of the word "only". It could be done by local authorities; it would just be a lot more costly.

The point that the convener is getting at is whether it is your preference to collect the tax through the Inland Revenue.

Absolutely.

That is fine.

What consequential assessment has been made of the impact on local authority employment of the collection of the tax through the Inland Revenue?

The policy memorandum states clearly that we envisage a policy of retraining and redeployment within local authorities, particularly given the issue of the surplus. I am sorry to return to it, but the issue is very important in this context. If we were to propose a measure that raised less money than the council tax raises at the moment and that involved the replacement of council tax finance staff, the question of how we could afford to redeploy those staff would be raised.

The key element in the £313 million surplus is that more than enough money will be made available for distribution to the local authorities for redeployment and retraining purposes.

Before you go any further, I should say that I am not pursuing the question of the surplus. I am interested in a quantification of the impact of the measure on local authority staff numbers. Obviously, one calculation for the financial memorandum is the cost of redeploying and retraining local authority staff. If we are to drive that calculation, we have to know how many people are involved.

Unfortunately, we have had no co-operation from the Convention of Scottish Local Authorities on a breakdown of each local authority's employment of finance staff. COSLA has refused to give us the information that would allow us to make a detailed breakdown of the number of individuals who are employed in posts that are, in the main, solely for council tax collection purposes. In fact, COSLA has also refused to give the information to the Scottish Parliament information centre, which is unfortunate.

Some finance departments would make the point that their staff are not employed solely in council tax collection but are also involved in housing benefit calculations and administration. The level of information that COSLA has provided makes it impossible to make detailed calculations. That is why I am giving generalised figures for the money that would be made available to the local authorities for redeployment and retraining.

But you accept that, as a result of the bill, a group of local authority employees would no longer carry out the functions that they undertake at present. At the moment, you are unable to define the number of individuals involved and the likely cost of retraining and redeploying them.

I will come in on one aspect of the question. We know from the Accounts Commission that the cost of collection is £68.5 million. We have indicated that that figure would be reduced to £12.7 million. The difference is therefore £56 million. Although some of the money may relate to IT systems, most of it will be staff costs and money would be available to redeploy and retrain them. The difficult question relates to the number of staff and the basis on which they are employed. If we adopt the figure of about 20,000, we come to the figure of about 2,000 staff across the piece who would be the subject of redeployment.

We have estimated the time for redeployment and have put the overlap at about three to four months. We are looking at an exercise that would have to take place across the whole of Scotland, with some lead-up time to the changeover. We estimate that three to four months would be needed for such a redeployment exercise.

You will see that the cost saving is established under paragraph 20 of the financial memorandum. The words "per annum" are missing, however. We state that there will be an estimated saving of £59.2 million, but it is expressed as if it were a one-off saving. Of course, it is a per annum cost—that is the level of saving we are talking about. You are right to state categorically that no council tax collection departments would exist. Those who worked solely on council tax collection would no longer work on that job. We would like them to be retrained and redeployed elsewhere.

It is accepted that at that point redeploying such individuals would in effect be a charge against the taxpayer, whether it related to the surplus from the service tax or to the general cost of introducing such a scheme.

Absolutely. The charge would be based either on the surplus or on the savings, which are not part of the surplus that would be generated. The £59 million of savings are not part of the £313 million surplus, so approximately an extra £60 million would be available to spend on retraining and redeploying people.

I will ask about billing arrangements for individual service tax payers. I see nothing in the documentation about a mechanism for costing that is attached to billing. How would people find out what they were due to pay?

We assume that the Inland Revenue as the collecting agent would have codes on payslips that told people how much their SST bill would be. Self-employed people who submit self-assessment forms would be subject to the same level of self-assessment as the Inland Revenue has at present. On that basis, the SST bill would be calculated.

Have you taken account of the cost of issuing SST bills and of the fact that more bills per household would have to be issued?

The proposed system is different from council tax. You talk about issuing a bill that is a source of information for someone about what they are liable to pay. Under the SST system, the tax would be deducted at source. That is a strength rather than a weakness. What the payment was for would be flagged up on an individual's payslip. A saving would be made, because the massive administration that is involved in informing people of their bills would not be needed. I am sure that you have constituents who inform you that they have received three, four or five different council tax bills, such is the complexity of the various rebate systems.

If we assumed for the sake of argument that people wished to be informed of their service tax liability, that could be done only through a bill to the individual, rather than through a recalculation of tax, which would be shown in the tax code, as you suggest.

I envisage that the information that the Inland Revenue supplies when informing people of their tax code would include a form of words that explained that a proportion of their tax liability related to the Scottish service tax. That would be included as extra explanatory notes in the normal process by which the Inland Revenue advises people of tax code changes when circumstances change.

The Inland Revenue system is not designed for that purpose. Its system would require to be redesigned to achieve what you suggest. I presume that people might feel that they had a right to know what the tax would be spent on. The advantage of councils doing the billing themselves is that they can tell people not only how much they must pay, but what the resource will be spent on.

Such information would be part of the process of legislative change in Scotland. Once the service tax system was introduced, everyone in Scotland would know that the section of their income that was paid under the Scottish service tax was ring fenced for local government jobs and services. If somebody had a complaint about the delivery of local government jobs and services in their area, that would still be directed to the local authority in relation to the misuse of the money that it received, much as happens now in relation to the small proportion of local authority revenue that local authorities collect and for which they set the council tax level. I do not envisage a major problem of individuals not being aware of what the money is for, if that is what you are suggesting.

The council tax is currently collected on a household basis. Effectively, people's water charges are collected through the same mechanism. That is a cost-effective mechanism, because it involves issuing one set of bills and offers significant savings. If water charges continue to be put out on a household basis—with the additional cost of that being borne either by local government or by the water authority—will not a significant proportion of the savings that you calculate be wiped out?

Not at all. I would have hoped that you and others would have welcomed the development that water and sewerage costs would be identified and billed individually. The current situation is unacceptable. Local authorities indicate that many of their council tax arrears are actually arrears for water and sewerage charges. There is a great deal of confusion. In Glasgow, people tell me that they cannot understand their arrears bill, because they do not pay council tax. I am sure that the same happens in your constituency. Many unemployed people do not realise that they are still liable for water and sewerage charges. It would be beneficial to Scotland for water and sewerage charges to be identified individually, rather than milled together with council tax, as currently happens.

That is a policy issue. I am interested in the financial implications. At present, a single set of bills covering water charges and council tax is issued. You are saying that significant savings are to be accrued from not issuing council tax bills. However, if a water bill has to be issued, some of the projected savings that you are indicating will fall elsewhere.

As you are aware, Scottish Water pays local authorities to carry out a billing exercise on its behalf. Under the service tax, Scottish Water will have to bill people separately. We do not know what the extra burden for Scottish Water would be. It would certainly not be great enough to tip the balance by adding significant costs.

The service tax looks like a clever variant of a local income tax. However, one issue that is always raised when the Scottish variable rate is debated is the administrative costs of using the power and whether that is worth while, given the revenue that it would generate. Is your proposal not a little broader even than that? The bill would affect not just Scottish taxpayers subject to the variable rate but those who are domiciled or own heritable property in Scotland. Would that not mean that, in addition to the Inland Revenue maintaining information on whether someone was a Scottish taxpayer, it would need to maintain a list of people who were domiciled but not resident in Scotland and of people who were neither domiciled nor resident in Scotland but owned heritable property here? Would the bill not introduce additional costs in that respect?

We do not think that it would. We think that most of the information to which you refer is already in the possession of the Inland Revenue. When contracting to collect the tax, the Inland Revenue would have to do so on a UK-wide basis, because some tax offices may deal with Scottish workers who are employed to work down south or in Wales but who live in Scotland. There is no question but that there would be a UK-wide Inland Revenue contract. An instance of the type to which you refer would involve a UK resident—let us say someone called Mr Fayed—who has a huge landed estate in Scotland and currently pays a very small amount of council tax for it. Under the new system, Mr Fayed would be liable for service tax based on his UK declared income. The bill would be substantial, but I am sure that he could afford to pay it. That information is largely available. There is little or nothing in the bill that would require extra information that is not already in the public ambit to be found.

Yes, but I am not sure whether the Inland Revenue routinely collects information on domicile, which I understand is a concept that is relevant only to inheritance tax, although I may be wrong. It is clear that there will be additional costs if information is not on the system.

Getting rid of the need to maintain the council tax register of properties would be one benefit of abolishing council tax and moving to a classic local income tax. I can see why you have proposed what you have proposed, but requiring to maintain a council tax register in effect and the other information would surely lead to duplication and additional costs.

Gordon Morgan wants to comment on that. However, I assure you that the Inland Revenue regularly collects information on domicile. Indeed, one of the biggest gripes of many people is that non-domiciled status is often given to multimillionaire friends of the Government who then avoid paying tax in this country.

I would like to make the same point. The taxpayer has an obligation to keep the Inland Revenue informed of their residence. Obviously, a lot of information comes from employers through the pay-as-you-earn system—which applies to around 71 per cent of people in the UK—so that information is with the Inland Revenue. The only question relates to the possibility of an additional burden on employers in having regularly to provide information under a scheme that is not unified at the Scottish level, with individual local authorities. However, we think that costs would be minimised under the proposed scheme. All the information that the Inland Revenue would require would be postcodes for where employees stay and information about where they work and whether they work in Scotland or England, which is information that employers already provide.

There would, as always, be hard cases. Somebody who works in the oil industry and is based in the north-east is a classic example. They would work in the Scottish economy, sometimes in the UK economy and increasingly outwith the UK. Perhaps they would be less of an issue in the administration of UK income tax, but would there not be additional costs if a service tax were also administered?

If the person were resident for more than 90 days in Scotland—they do not have to be consecutive days—they would be liable for the tax. The majority of people to whom you are referring would be here for a minimum of 90 days.

The paper that the committee received from Gordon Morgan this morning seemed to suggest that writing off debts would cover the gap between the figures in the Scottish Parliament information centre briefing and the figures that you have given. Will you tell us more about how the debt write-off would happen? You say that it seems uneconomic to collect debts after 10 years, although you admit in paragraph 9 of your paper that there is no justification for your figures.

I will briefly refer to the policy thrust behind the bill—Gordon Morgan can discuss the financial details. Our position is that pursuing debts that have been in existence in excess of a few years—never mind in excess of 10 years—is not only uneconomic, but with respect to the poll tax in particular, which was an unfair tax to begin with, immoral. The financial memorandum originally estimated a loss of revenue in the region of £100 million to £150 million and Gordon Morgan's subsequent work with the new Scottish local government statistics uprated the figures to between £140 million to £180 million. He has done most of the work on the matter.

Paragraph 9 of the paper states that

"information is not available to prove or disprove"

the contention that "vanishingly small amounts" of poll tax debt are being recovered. The reasons why the information is not available are that, first, local authorities have largely subcontracted the work to sheriff officers and, secondly, detailed figures have not been published recently on local authority collection rates for poll tax debt, in the way that we get such figures for council tax debt. The whole thing is lumped into one sum of money. Sheriff officer costs are not broken down by any local authority.

I read that in your paper. I seek clarification. Are you suggesting that debt should be written off after a certain number of years or when it reaches a certain percentage?

The bill relates to debts up to 1 April 2004, depending on when it comes into force. We are looking at roughly a two-year cut-off for debts. We want to make a fresh start and move on, instead of having a continual clawback that takes up a lot of resources and often relies on unreliable information. The Scottish Executive may think that three or four years would be better than two years, but once you get beyond four years it is uneconomic and unacceptable to continue to pursue such debts.

How will that policy affect any payments that are due currently or before your tax is in place? People will say that they will not pay it, because the debt will be written off in three or four—or even two—years.

I am not sure that it would be as simple as that. The council tax collection rate has improved. As you can see from Audit Scotland's figures, there are still gaps, but the rate has improved. Collection systems are very much in place; there are direct debits and information on where people work and on people's bank accounts. If people are given notice over the next two years that, if they do not pay their council tax bill the debt may be written off, it would be difficult for them simply to avoid paying it, given the level of information that is already available to local authorities and the current level of and methods for collection. We do not envisage that being the problem that it would have been, for instance, if we had said at the start of 1993 that by 1995 any debt would be written off. Collection systems were not in place then, the council tax had not bedded down, local authorities did not have information on where people worked and so on. We do not envisage that being a big problem.

Finally, looking forward, have you made any allowance in the service tax for any write-off two or three years down the line because, as with all taxes, there will be a degree of avoidance or non-payment?

Given the clear evidence of the success of income tax compared with property tax, the level of non-payment will be acutely less than it is in relation to council tax. We talk in the memorandum about 1 per cent either way affecting the surplus that would be generated, to which we keep referring—the convener quite rightly keeps telling me not to talk about it because it is a policy question. It is clear that income taxes throughout the world, let alone in the UK, have collection rates of up to 99 per cent, compared with the collection rate for property tax, which in Scotland sits at around 92 per cent, but which for some local authorities is as low as 88 per cent.

I am inclined to look at the potential unintended financial consequences. I worry about people migration. I worry about wealthy and skilled people who are mobile and about young couples who find that they need to be mobile. I also worry about the rise of the 90-day Scot—people who organise their lives so that they spend no more than 90 days a year in Scotland but who still manage to keep strong contacts with Scotland. Most of all, I worry about the impact of the tax on competitiveness and its cascading negative effect on Government finances. If investment decreases, if we lose skilled people and if business costs rise as a result of wage inflation, we could see an increased incidence of business failure and more businesses in Scotland being owned externally. Subsequently, staff could have much worse terms of employment. We could have fewer people in employment and more of them on lower rates of pay. That would pose a real cash-flow problem for Government.

I would love to answer that question, but it deals with a policy area. I will answer the question if the convener does not mind.

As the question was on policy—

I linked my question back to numbers. It started with numbers and it finished with numbers. My question is about the effect on Government revenue, which is a fundamental issue in any financial memorandum.

The question was not, strictly speaking, about the costings of the policy—

Please, please let me answer.

To be fair, Jim Mather's question would be more appropriate in a policy context. I do not wish to curtail him, but he should ask the question in the Local Government and Transport Committee rather than the Finance Committee.

Elaine Murray has a question.

I hope that my question will be counted as relating to the financial memorandum.

In the section of the memorandum headed "Costs on Other Bodies, Individuals and Businesses", paragraph 31 refers to a Scottish Parliament information centre estimate—which I have not seen—which suggests that 72 per cent of households would benefit from the SST. Given that nearly 20 per cent of households currently pay no council tax because they receive council tax benefit, do you calculate that a surplus would be generated by 8 per cent of the population?

No, the 72 per cent figure refers to those who are liable to pay council tax. As the SST would be a progressive tax, the first £10,000 would be tax free. Unfortunately, there are many citizens—47 per cent, according to the latest figures—who fall into the category of earning less than £10,000 per annum.

Progressive tax rates would then apply after the first £10,000. For those whose income is between £10,000 and £30,000 per annum, the rate would be 4.5 per cent. Therefore, even those who are on average incomes would stand to benefit from the SST by comparison with the average band D council tax charge. According to Inland Revenue figures, the largest number of taxpayers are located within that bracket: some 614,000 Scottish taxpayers earn less than £10,000 per annum; 580,000 earn between £10,000 and £15,000; 428,000 earn between £15,000 and £20,000; and 490,000 earn between £20,000 and £30,000. The bulk of Scottish taxpayers are in the £10,000 to £30,000 bracket.

So paragraph 31 should state that 72 per cent of council tax-paying households—not 72 per cent of all households—will benefit, given that many households pay no council tax.

Paragraph 29 contains a table showing those for whom the SST would, it is argued, be financially advantageous. However, 43 per cent of households in Scotland contain two or more adults, who may each have an income. For example, if a female firefighter lives with a male nurse, they would be £25 a week worse off than they are under the council tax—in Dumfries and Galloway, they would be £70 worse off—despite the fact that they are not on high salaries.

It is interesting that you use that statistic but, as you are probably aware, 47 per cent of households have only one earner. For the purposes of any research, it is acceptable to build in an assumption about one-earner households, given that they account for the largest percentage of households in Scotland.

You referred to the situation of a firefighter and a nurse who have a household income of more than £40,000 a year. You may or may not be aware of this, but only 10 per cent of households in Scotland have an income of more than £40,000 per annum, so that situation is very rare.

I think that the correct figure is that less than 10 per cent of people in Scotland have a personal income of more than £40,000.

No, the figure refers to households.

The problem is that you are comparing an individual tax with a household tax and making judgments based on your assumptions about households.

I assure the committee that substantial research was carried out on that. We took the composition of households in relation to the number of people in the household, the number of taxpayers in the household and the council tax band and local authority area that the house was in. We then built up that information into an overall table that analyses the composition of all households throughout Scotland in each of those categories.

Based on the correlation of various statistics, 77 per cent of all households in Scotland would be better off, or, if they paid no council tax, would be no worse off, than they are now.

There was a band in the middle where the figures were indeterminate. It was not certain whether the 7 or 8 per cent of households that had two incomes in the mid-£20,000 range would be worse off or better off, but 16 per cent of households would definitely pay more.

The research is not just based on single or double earners; we actually used the best available statistics and analysis to come to a determination of the number of households that would gain and those that would lose.

Did you include the fact that people do not necessarily live in a band D house? Some might live in houses in lower bands; if they were on a lower income, they would benefit less.

I hope that your experience of research down the years would allow you to accept that assumptions have to be built into any research. Given that band D is the average council tax band in Scotland, it is reasonable to make that assumption. It is quite unrealistic to think that someone on an annual income of more than £40,000 per year would live in a band D house.

I live in a band D house.

That is quite unusual for someone earning your level of income. I would expect you to live in a house in a higher band. When the domestic revaluation that COSLA is calling for is done, I am sure that your house will move into a higher band. However, that remains to be seen.

The household income figures are available in written answer S2W-19143 and show that less than 15 per cent of households have an income of more than £40,000. Therefore, 85 per cent of Scottish households do not fit into the scenario that you have just proposed. If I had built my research around a 15 per cent model, you would be asking me about the other 85 per cent.

Do you believe that that 15 per cent can generate a surplus compared to what is generated by the council tax?

No; that is a misunderstanding. You are mixing up the surplus that will be generated under the bill with those who will save as a result of the bill. This is an individual tax, so the tax base will be larger than that of the council tax, even with the built-in exclusion of the first £10,000 per annum. A large amount of money will be generated because of the amount of people who will pay. However, those who will pay the most will be people such as you and me who are on very good incomes. Those who will pay the least will be those on average and below-average incomes.

I am a bit confused about the comparison between paragraph 24 of the financial memorandum and paragraph 143 of the policy memorandum. Paragraph 24 of the financial memorandum estimates that the SST would in 2001-02 have raised £2.1 billion, compared to the council tax's £1.65 billion collection. The SST would therefore have generated a surplus of £505 million. In paragraph 143 of the policy memorandum, which looks at 2003-04 figures, CIPFA calculated that the total council tax to be collected by Scottish local authorities would be £1.83 billion compared to the SST's collection of £1.85 billion. Why are the figures so different? Why is there so much less of a gap between the figures in the policy memorandum, between which the difference is 0.79 per cent? The difference is something like 25 per cent in the financial memorandum.

Gordon Morgan will also respond to the question. There are two points to be made, the first of which is that we must be careful about the years that are compared. The last financial year for which the Inland Revenue has provided a detailed breakdown of income figures is 2002-03. On comparisons in respect of the generation of surplus, we must compare those 2002-03 figures with the council tax figures for 2002-03; we cannot compare apples with oranges by comparing figures from different financial years.

Secondly, the answer to the question was, I hope, partly answered in my initial comments about the level of surplus to be generated. The policy memorandum talks about a level of surplus being generated that would be based on six tax bands. When we drafted the bill, we included only five tax bands; we took out an 8 per cent tax band at the £20,000 to £30,000 income level, which reduces the level of surplus that would be generated. The Executive has corrected our figure, which is why we have used the £313 million figure.

The error relating to the 8 per cent tax band is in paragraph 24 of the financial memorandum. I can provide corrections for those figures based on the revised figures from the Executive. The calculations in paragraphs 24 to 27 were based on that error—where the document reads "£505 million" it should read "£313 million". We have notified the Local Government and Transport Committee of the Executive's change.

It would be useful to have the correct figures as you see them, because we need such significant discrepancies in calculations to be resolved so that there is clarity.

There is not a problem—the Executive has provided the correct figure, with which we have no problem. The key thing is—as I said earlier—that the Executive is not questioning whether a surplus will be generated, but what its size will be. The Executive suggests that the surplus will be £313 million, rather than what we had suggested, which was based on the omission of that 8 per cent tax band.

I am sorry if I am losing track of the various bits of paper. For clarification: in the policy memorandum at paragraph 143, line 5, the difference between the SST and council tax figures is £14.4 million. Why is the figure of £313 million ten times higher than the figure in the policy memorandum? I accept that I may have just got tied up in the wrong document, but there is a big gap between £14.4 million and £313 million.

I think that that figure from the policy memorandum refers to something entirely different. The policy memorandum says that, given certain assumptions, the amount of money taken in council tax would be £1.8 billion. I think that the £14.4 million figure might be something to do with different years, but I would need to come back to you on that. I prepared for questions on the financial memorandum, but I had not even noticed that the policy memorandum figure was different. I can come back to you on that after the meeting, if that is okay.

That is fine.

The financial memorandum indicates a yield of £1.649 billion from the council tax. However, the yield should include £285 million from rebate grants from the DWP. Therefore, the actual yield is £1.934 billion.

My understanding is that the £1.649 billion includes £285 million from council tax rebate.

That is not my understanding, which is that you need to add on the £285 million.

Again, convener, the Executive evidence that is before you makes it clear that the surplus that we have talked about—the yield—assumes the loss of council tax benefit. The £313 million surplus figure assumes the complete loss of council tax benefit. We do not accept that, of course; we would fight to retain all or part of it. If it was retained, the yield would be £313 million, plus £285 million.

But your figure was about £500 million in the first instance.

But that £500 million would be almost £800 million, if it were assumed that council tax benefit would be retained. The £505 million is net of the loss of council tax benefit. All the figures assume that we lose council tax benefit. Politically, I do not accept that assumption. However, in the context of facing attacks on the robustness of our surplus figures, we felt that it would be better to assume that we did lose all the council tax benefit.

The problem for the Finance Committee is that the figures involve very large sums of money. Before we can be confident about them, we need to be a bit clearer about which set of estimates is accurate. It is not helpful that there are different kinds of figures floating about, with discrepancies of £200 and £300 million between them. The committee must seek greater clarification of, and greater certainty about, the actual costs and the basis of comparison, so that everybody is clear, when the bill is debated in the chamber, about the basis of the difference between the systems.

Absolutely.

Elaine Murray raised the issue of the difference between a tax on households and a tax on individuals. It strikes me that the assumptions that you make to arrive at a figure of 73 per cent of people benefiting from a service tax system are based on a fundamental statistical error in that you have not made like-for-like comparisons. Whether you can calculate what proportion of households would benefit from replacing the council tax with a service tax is a difficult question to answer. I would have thought that it is certainly not one that you can answer with a precise figure, given the evidence that you have at this point.

I invite you to seek evidence from the Scottish Parliament information centre on that, convener. SPICe supplied much of our evidence and it was very helpful in answering our questions. It is important to bear it in mind that when we have percentages for household income across Scotland, a figure for the number of individuals across Scotland and a breakdown of the make-up of households across Scotland, it is reasonable to build a model that allows us to make comparisons. You might say, for example, that we should have used band A or E instead of using band D; but we thought that using band D would be best because it is the average.

Some people say that the assumption of a one-earner household is unrealistic and that we should use the assumption of a two-earner household. However, if we use the assumption of £35,000 income for a two-earner household, we go from a £35,000 level of tax, which is high under the service tax, to a £17,500 level of tax for each earner, which is much lower under the service tax. In other words, the assumptions that we have used have not overestimated or overegged our pudding—if anything they have undermined it. SPICe will give you figures that show that if we assume one-earner households, the level of saving is in the region of 72 to 73 per cent and that if we assume two-earner households the saving goes up to more than 80 per cent.

The point that I am making is that you are producing a figure based on a series of different assumptions.

Yes, but they are reasonable assumptions.

I am not sure that they are. It is difficult to make such assumptions on the basis of the household pattern. There are issues about benefit payments, which are not incorporated in the SPICe assumptions. We would need much more robust analysis either to justify or to undermine your figures. Given how the figure was arrived at, I am nervous about accepting that it is robust enough.

I have checked the answer to one of your earlier questions and I can now explain. In the consultation paper we refer to different years. The figure for council tax collection in 2001-02 is in paragraph 24 of the financial memorandum. That was the comparison year for the income tax received. You are correct that in 2004-05 the total figure moved up to nearer £1,900 million. It is confusing, because we do not say to which years all the figures relate, but I can provide a brief note on the figures and years so that you can identify exactly where the figures come from, particularly those in paragraphs 24 and 27, so that we are all singing from the same hymn sheet.

You said that 47 per cent of people in Scotland live in one-adult households.

The figure is 46.9 per cent, rounded up to 47 per cent.

If the calculations are made on the basis of who will benefit, should not the fact that all those people would be eligible for a 25 per cent council tax discount be taken into account?

It was taken into account in our calculations. We have not worked on the basis of simply assuming one-earner households; we have built into our model the fact that some households have more than one earner and that some have two or three earners—a very small proportion of households in Scotland have more than two earners, but some have. The figures are based on what we think are reasonable assumptions. The convener said that he does not think that the assumptions are reasonable. I would like to know what is unreasonable about them.

It would be useful to get a list of the assumptions that have been made, and not just those in relation to the calculation of household benefit; the list of the assumptions made about the differences in yield between service tax and council tax would be helpful. Having information on the core assumptions on which the calculation is based would be helpful so that it can be tested or at least inspected. They are the key financial elements, and we need to have absolute clarity on them. My point on the household issue is not that I have a better figure than you, but that, on the basis of the information that we have, I do not think that we can make a robust assumption about whether there is a benefit or not. It would be impossible to construct such an assumption—or it would be possible only after carrying out a substantial amount of extra research that was based on something other than aggregate figures.

On a point of clarification, am I correct to assume that, with regard to the comparative figures for the amount that different occupations would pay under SST, the intention is not that someone who is self-employed will be worse off under the proposed system?

The amount that those people will pay will be based on their income. Many people wrongly assume that the self-employed are paid an awful lot more than they are; indeed, evidence shows that they tend to earn a less-than-average income. As a result, under SST, they will pay less than they pay under the council tax.

So the intention is that a self-employed person who takes home £10,000 or £15,000 would pay the same as an employed person who earns the same amount.

Absolutely.

However, given the bill's current definition of "relevant income", someone who is self-employed might well be taxed on their turnover. That is very different from being taxed on take-home pay.

People will be taxed on their relevant income—in other words, on what HM Revenue and Customs considers their relevant income to be.

It would be nothing to do with what HM Revenue and Customs considers to be their relevant income. The bill says that "no account" will be

"taken of any … deductions applied in respect of that income in order to determine the individual's actual liability for income tax".

I would have thought that, for self-employed people, one deduction might well be the cost of providing their services.

But section 3 defines the term "relevant income".

Exactly; that is where I took my quotation from. Under section 3, income is essentially the income that is liable for income tax, except that no account is taken of allowances—which I understand—or deductions. However, for a self-employed person, one deduction might be the cost of the services that they provide. For example, a general practitioner who is self-employed would deduct the cost of employees such as receptionists or whatever against their income tax. Under the definition set out in section 3, the GP's income would be significantly higher than the income on which he or she is liable to be charged income tax. That might be a drafting rather than a policy matter. I acknowledge what you have said about the bill's policy intent but, from my reading, the bill does something very different from what you have suggested.

The bill was drafted by Thompsons Solicitors, which took advice from HM Revenue and Customs, particularly on the issue of self-employed people. When we looked at the breakdown of income for self-employed employees, we found that they are clearly not living the life of Riley.

Absolutely.

We understood that they would be taxed on the same basis by HM Revenue and Customs and that those types of allowances would be taken into consideration.

I suggest that you clarify that matter, because it is fundamental to the bill's impact, particularly on small businesses.

I think that that is a policy matter.

On behalf of the committee, I thank Tommy Sheridan and Gordon Morgan for attending the meeting. Under the current procedure, the committee will now agree a report, which will form part of the subject committee's consideration of the bill and its stage 1 report to Parliament.

What is your timescale for producing the report? We need to know the various deadlines for feeding relevant information into your deliberations.

I have been taking note of the clarifications that the committee requires.

Our timescale is fairly tight. We seek to agree our report on 8 November, which means that the draft will need to be completed by a week on Thursday. We will get in touch with you after the meeting.

The points on which you seek clarification relate mainly to the years of comparison and the nature of the assumptions. We can provide that information before the end of the week, if that is acceptable.

That would be very acceptable.