Our third item of business today is evidence from HM Revenue and Customs officials on the Scottish rate of income tax. We are joined once again by Edward Troup and Sarah Walker. I welcome our witnesses to the meeting and invite Mr Troup to make a short opening statement.
Thank you, convener. It is good to be here again. I am happy to be able to update the committee on the work on the Scottish rate of income tax. The good news is that, as you will have seen from the memorandum that we have submitted, we are on track to implement the Scottish rate on time and within—indeed, a little bit below—the original budget.
The main activities at present are communications, identification of Scottish taxpayers and ensuring that employers have a good level of awareness of their obligations and what the Scottish rate will involve. We have had good communications with employers, payroll software companies, pension providers and professional bodies. We have already put out specifications and guidance for software developers and we have put a lot of information about the Scottish rate online. This week, we published guidance for Scottish taxpayers on www.gov.uk, the UK Government website.
Later in the year—we have not quite fixed the date, but it will probably be in December—we will write to all the taxpayers whom we believe are Scottish residents based on their postcode. That is slightly over 2.5 million people. That will give those individuals an opportunity to correct their address details if, for instance, they have recently moved from Scotland and are elsewhere or have recently moved to Scotland and become aware of the need to identify themselves.
We are doing and will continue to do a series of checks of the addresses that we hold on our systems against a range of databases to give us as high a level of confidence as possible that we have correctly identified those people who are likely to be Scottish taxpayers. Internally, our first information technology release, which is linked to that identification of Scottish taxpayers on our system, has gone smoothly and we have no reported problems with it. The next internal IT round, which will implement the pay-as-you-earn changes on our systems, is on schedule.
As I said, the estimate of the costs of implementation remains a set-up cost of £30 million to £35 million against an original estimate of £40 million to £45 million.
That is the progress on the Scottish rate. Of course, we are very conscious that the Smith commission proposals are coming down the track with the Scotland Bill, and we are staying engaged with that. We expect to be able to implement the proposed changes according to the anticipated timetable for Smith.
Thank you very much for that opening statement. I will start with a few opening questions and then bring in my colleagues around the table.
Paragraph 4 of the HMRC update states:
“HMRC has been carrying out a comprehensive programme of communications activity, such as publishing written material on GOV.UK, featuring detailed articles about the SRIT implications”.
You touched on that in your opening statement. How many folk look at www.gov.uk? Do many taxpayers look at it?
It is heavily used. I do not have any figures for overall www.gov.uk usage. It is not so much an active communication method as a resource for individuals who become aware of or have questions about a particular issue and want more information. It is not the primary source for proactively telling people about their obligations but it is a source that people can go to as awareness of the Scottish rate becomes more widespread.
We believe that www.gov.uk is the natural place that people go to now, and most of the search engines on topics related to Government business will take you to www.gov.uk as the top hit. It is an important core source of information.
In paragraph 8 of the HMRC update, you state that
“in around 85% of cases HMRC held an address for a taxpayer that could be matched against an address held elsewhere.”
You go on to detail how you could identify other Scottish taxpayers, and you say that
“about 98% of the taxpayers for whom HMRC holds Scottish addresses are correctly identified as likely to be Scottish taxpayers.”
In paragraph 9, you talk about your plans to write in December to those taxpayers for whom you hold a Scottish address.
What is being done to raise awareness in the rest of the United Kingdom? Clearly, there may be an issue with people who have addresses on both sides of the border and with those who may have an address on the other side of the border.
Is any work being done to remind customers of the importance of notifying HMRC when they move house from south of the border, for example from England to Scotland or from Wales to Scotland? Is anything being done on that side?
There is no obligation to notify HMRC of your address or of a change of address. I think that I discussed that point with the committee the last time that I was here. We are relying on our own and other data sources to identify when individuals become resident.
Typically with an employee, the employer and the payroll system—and hence the PAYE records—will record a change of address, as indeed happens now if someone moves into the UK from elsewhere. There is no obligation on them to tell us their address in the UK, but a payroll system will pick up the fact that they have a UK address and they will become part of the PAYE system. The same system will apply in the case of the Scottish rate of income tax.
We are not—for value-for-money reasons—proposing to do a marketing campaign in the rest of the UK to tell people that if they move to Scotland, they will need to notify us. We estimate that something like 2.3 per cent of the Scottish population turns over in any year—either through people moving into Scotland or through people moving out of Scotland to the rest of the UK or elsewhere—which is a relatively small number against the totality of the Scottish population.
Sarah Walker is responsible for the operational work and may have something else to say about the UK addresses issue.
Yes. Because collection of the Scottish rate will depend on where someone’s main residence is, the important thing will be to pick up people who move into Scotland and become Scottish residents.
Later this year, we are looking to do some targeted publicity, for example around the Zoopla website, which people who are moving house tend to look at. There will be other targeted advertising to try to pick up such people. As well as using the sources of information that Edward Troup talked about, we are contacting estate agents and solicitors, so that they can remind people who are moving to Scotland that they need to ensure that we know about their change of residence.
As far as the rest of the UK is concerned, we need to contact and inform employers throughout the UK, because an employer anywhere in the UK might encounter someone with a Scottish tax code. For example, a person might have recently moved out of Scotland. Our communications with employers and payroll operators are UK wide, but communications with taxpayers themselves will focus on people in Scotland.
I am pleased that you are working proactively with estate agents and so on. I did not think that you would contact every person in the UK, but I suppose that when people get their tax returns they could be told, “Scotland has a new tax system. If you are moving to Scotland, please notify us”—blah, blah, blah. Is that a helpful suggestion?
First, remember that the majority of people do not fill in tax returns, because their tax affairs are dealt with under PAYE. The self-assessment return will have a box that asks whether someone is a Scottish taxpayer, so that will be an implicit reminder. If someone has moved or is about to move to Scotland they will be prompted, because the question about Scottish taxpayers will lead them—online or through the guidance—to information about whether they should answer yes or no to that question. There will be sufficient prompts in the SA return—if people fill it in properly, as most people do—to capture those people.
The initial challenge is to ensure that the starting database is as accurate as possible.
You say in paragraph 12 of your update:
“HMRC has also worked with the Scottish Government on the possibility of being able to access address information from NHS Scotland records.”
Given what you have said about other sources of information, what would that add?
The background to that is that we have our own database records, which indicate individuals’ addresses, but we cannot be certain that they are accurate. You quoted the point about our now having got to 98 per cent confidence. As with any large data sets, the best way to improve ours is to compare it with other data sets. We have been using commercially available data sets, the electoral register and the published address and postcode information from the Royal Mail, which are helping us to improve the accuracy of our data set.
If we had access to the NHS Scotland data set, that would be another data set against which we could check our records, and doing that would no doubt further improve the accuracy of our records. However, that is not an essential part of our checking; no individual data set is essential, but each one contributes to greater accuracy and confidence in the addresses and names that we have.
You take a belt-and-braces approach.
Well, yes. With very large data sets—we are talking about 2.5 million people—we are never going to be 100 per cent accurate. That is just the reality. Apart from anything else, people come and go, as I said, and there are taxpayers who—one way or another—manage to stay off our radar. It is a matter of always looking for additional ways to improve the accuracy of our knowledge and information about the taxpayer population.
In paragraph 17, you say:
“If the Scottish rates diverge from the rates which apply elsewhere in the UK, there will be an incentive for taxpayers to claim that they live on one side of the border, when they live on the other.”
What differential would start to have an impact? Are we talking about 1, 2, 3, 4 or 5 per cent?
I think that human nature is such that any amount of tax saving creates the inclination in individuals to ask themselves whether there is some way in which they could avoid paying. That is the nature of tax raising. Our experience is that the higher the differential rate and the greater the ease with which it is possible to make the changes to exploit the differential, the more likely it is that those things will contribute to the number of people who will make those changes.
There have been deliberate changes of rate. For example, we cut the duty on unleaded fuel to encourage people to change, and it was easy to change—people just had to pick up a different nozzle at the pump. People did not avoid tax, but they saved tax by making a choice. However, moving your family lock, stock and barrel from Scotland to England or vice versa is a slightly more life-changing event. It will take quite a strong differential to encourage people to move entirely for tax reasons, but if you are buying a house near the border it may be one of the factors—along with stamp duty rates and everything else—that will determine which side of the border you want to live.
11:30
You have no x/y-type graph to show where the majority of people would change their behaviour. We have just been to the Basque Country and we got some quite interesting information from the finance minister there. As far as the Basques are concerned, there does not seem to be any impact at all from the differentials in taxation between that part of Spain and the rest of Iberia, even though there are significant tax differentials. It is possible to raise tax to such a level that revenues fall to such an extent that it becomes detrimental to the overall tax take in the economy. I just wonder where that level is.
Our experience with the wider tax system has been that those individuals who change their behaviour for tax reasons tend to be those who have most to gain, and they are the better off, so it is the levels of tax that apply to the wealthy that are most likely to result in a behavioural effect. It is a matter of fact that such people often have more than one home anyway, and it is a lot easier for them to say, “I’m going to spend seven months in my London home and five months in my Edinburgh home, rather than vice versa,” than it is for someone who has one home and all their friends and connections somewhere in Glasgow—it would be slightly more disruptive for them to have to move south of the border.
If I had to conjecture, I would say that, for the overwhelming majority of the population, the sort of differential in tax rates that is likely to come about—as suggested by past experience—will have an almost negligible effect. However, to the extent that there are significant differentials for the better off, that is where we would expect to see a change in behaviour, and we would probably be looking to ask compliance questions about that.
I have just one final question before I open up to the committee. I know that everyone wants to ask about this, so I thought that I should get in first. It is about the traffic-light system—I can see that other members are thinking, “Aw, I wanted to ask that.” Can you talk us through some of the issues, particularly with the ambers and reds?
The red, amber and green traffic-light system is a well-used tool in risk management. However, in simplifying things in order to present hot spots, it risks oversimplifying the situation, and risk management is an extremely complex business. It is complex because, before you even start, you need to think of all risks in two dimensions, considering the likely impact of the risk arising—for example, HMRC systems failing completely would have a massive impact—and the probability of that risk arising, which might actually be very low. On the other hand, the probability of not having 100 per cent of names accurate on a register is pretty high, because it is almost certain that we will not have 100 per cent accuracy, but the impact of that is relatively low. Before we even start, we need to look at risks against those two axes. Then there is the element of time. These are tools that should be used practically to help inform where we put our attention to keep the project on track.
The last time I was at the committee, the red-rated risk was about the identification of Scottish taxpayers, which we have just discussed. It was not that it was not going to happen, but it was a relatively high-impact part of the project, and at that point we had not taken the steps that we have since taken to ensure that we can have a high level of confidence about identification. It was right that that risk was red last time. I am pleased that it is not red this time.
The one that is red at the moment—and that is partly because we have very little control over it—is the notification letter, which we will send to taxpayers, or those who we think are taxpayers, in December. We are concerned that we do not have certainty about when we are going to send that letter. In part that is because we are not yet clear about the timing of the Scottish budget. The timing of the letter relative to the timing of the Scottish budget impacts on the letter’s content. We are forecasting that that risk will come down, because we expect you to announce the date of your budget—if you have any news on that, I would be delighted to hear it. Once we know the date of your budget, we will be able to determine when we can send the letter and whether we will be able to say what rate has been fixed for the Scottish rate. That is the risk that we are concerned about at the moment. Although we have very little control over when you set the date of the budget and, hence the timing of the letter, we are confident that we will be able to deliver on that.
Perhaps I should pause there, because there are quite a lot of ambers. We might want to pick up some of the amber risks at the same time.
I will not go into any of the other indicators, because colleagues might wish to do so. On the last one that you mentioned, as you will know, the spending review is on 25 November and I think that our budget is likely to come out in the first week of January. I do not know whether you are still intending to send out letters in December. Some of them might get lost in the Christmas rush of mail and people might not take as much notice of them—I do not know whether they will or not. I certainly think that the first week of January is the most likely time for our budget to come out.
I will leave it at that to allow other colleagues to ask their questions. John Mason will be followed by Jackie Baillie.
You say in paragraph 3 of your submission that employers will already be familiar with the Scottish variable rate. However, we have received evidence that employers were, in fact, not very familiar with it and that they had assumed that there would be no changes; perhaps they are assuming again that there will be no changes this time. Is that a consideration? Are employers really that aware of it?
I will let Sarah Walker add more detail, but I should first point out that paragraph 3 does not say that employers were familiar with the Scottish variable rate. It says that, because the Scottish variable rate was on the statute book, the functionality needed for it had already been built into most payroll systems.
We recently held a conference for payroll providers and found a high level of awareness of the forthcoming Scottish rate. The combination of functionality in payroll systems for the Scottish variable rate, which effectively transposes into the Scottish rate, and what appears to be a good level of awareness among payroll operators and software providers gives us a level of confidence, but I will ask Sarah Walker to say a bit more about our engagement with payroll providers and software people.
The vast majority of employers will use some sort of proprietary accounting system such as Sage to run their payroll, and those systems will have had the functionality for the Scottish variable rate built in. That will allow them to distinguish a code with an S, which will be a Scottish code, and to operate a different tax calculation when they get one of those codes. That is basically all that we are asking people to do now.
At the recent conference of payroll operators to which Edward Troup referred, we held a session on the Scottish rate, and when the person running the session asked for a show of hands on who was aware that it was coming, practically everybody put their hand up. We are confident that the payroll industry is ready for this. Part of the new information that we have put on the internet this week is guidance for employers on how it works. A lot of it is about reassuring them that we are not asking them to do anything complicated; we are not asking them to form a judgment based on their employee’s address as to whether they are a Scottish taxpayer. This is something that they are very well used to. They receive a particular sort of tax code, feed it into their payroll system and the right result ensues. That is the reassurance that we are giving employers.
In paragraph 5 of your submission, you talk about guidance for “customers”, but elsewhere you talk about “taxpayers”. Are “customers” different from “taxpayers”?
No. Customers are taxpayers and taxpayers are our customers.
I prefer “taxpayers”.
We talked about the fact that you have 98 per cent certainty with regard to the addresses that you hold and that, in 85 per cent of cases, HMRC holds an address for a taxpayer. We have also touched on the fact that the higher-risk people are probably those who are better off and who might have more than one property. Are you targeting them and trying to pin down where they are resident?
I am not aware that we are targeting them. People on the higher rate tend to fill in self-assessment returns—in fact, they will almost universally do that—so we are likely to have more information on them than we will for what I will call an average PAYE payer, for whom our only contact is likely to be through his or her employer.
Sarah, do you want to say anything about higher-rate taxpayers?
We have a high net worth unit, which looks after some of the very wealthy taxpayers and is very much involved in this work. As it will have a customer-manager relationship with those taxpayers, it should know quite a lot about them, including their up-to-date addresses. It is looking at whether we ought to do more to pick up any risks that relate to those people, particularly if, as we discussed earlier, the rate is different. After the end of the year, self-assessment taxpayers—people on higher incomes tend to self-assess—will be asked directly whether they lived in Scotland for most of the year. A positive return will be required from those people.
How much trust is required? Paragraph 9 says that
“it will be important to confirm which of these addresses held is their main place of residence”,
but paragraph 13 says that
“taxpayers are already able to amend their address using GOV.UK”,
which suggests that it is quite easy to change one’s address.
It is quite easy to change one’s address, just as it is quite easy to fill in a tax return and put false figures on it. However, that is not a good thing to do, because our compliance work is quite effective. We resource according to risk.
We are not that worried about people giving false information, because the better-off are aware of the consequences. The concern is about ensuring that we have in our net everybody who ought to be there. People are more likely to be tripped up by a lack of awareness than a deliberate misstatement of their position.
I am sure that there will be people who have homes in London and Edinburgh. In fact, I was talking to a director of a Scottish company who, although Scottish, lives in London and spends only a month or two up here. She was rather upset to hear that she will not be a Scottish taxpayer, because she wants to be; I do not know whether that is because she thought that there would be a cut in the rate. In any case, there are clearly people in that category. There will also be people who fill in self-assessment returns and have to answer the question whether, taking into account certain factors, they are Scottish taxpayers. They are likely to be the better-off, who are generally more compliant.
It is also worth saying that we expect to repeat the comparison of our address database with external data sets in the future. We will be able to check whether someone is giving us a wrong address by matching it with the addresses that they have given other people—for example, the address where they are registered to vote and those sorts of things. [Interruption.]
We are trying to work out where that noise is coming from.
It is not me.
In paragraph 11 of your submission, you say:
“The scan has produced a list of individuals for whom HMRC holds an address elsewhere in the UK”.
How long is that list? How many people are on it?
There are roughly 20,000 people on it.
That is a fair number of people.
It is, but it is less than 0.1 per cent of the total number on the list—I am trying to do the arithmetic in my head. Perhaps it is 1 per cent.
Paragraph 20 of your submission says:
“If the SRIT is set at 10%, the running costs are estimated to be £2m-£2.5m.”
However, if it is not set at 10 per cent, the running costs will be £5.5 million to £6 million. I was quite stunned by the suggestion that the costs would be two or three times more, even if we changed the rate by only 1 per cent.
11:45
Sarah Walker can give you more details, but the work that she and I have described—identifying the taxpayers, keeping the taxpayer base up to date and making sure that employers are operating the payroll—in a sense provides the baseline or steady state. Until the SRIT changes from 10 per cent, nothing more is needed, because there will be no difference in payments or collections. Effectively, the codes will be exactly the same for Scottish and non-Scottish taxpayers and everything will continue smoothly.
As soon as the rate changes, however, there will need to be a coding exercise to apply different codes to Scottish and non-Scottish taxpayers. Because of the differential payments, we will need to apply compliance to ensure that we collect the correct amount in the different cases. There will also be queries and contact from taxpayers whose tax bill has changed and who will be getting on the phone to ask HMRC what is going on. There might well be issues for employers, too, who will also have to contact us. As soon as the rate changes in either direction, there will be an increase in activity within the system and in the consequential activity of contact with us. All of that will take the level of activity, which is relatively low and steady at the moment, up to a significantly different level.
I ask Sarah Walker to amplify on what actually will be different if the rate changes.
Edward Troup has covered the main aspects. The market research that we recently carried out on attitudes to the Scottish rate makes it clear that people will be much more likely to phone us up or raise queries about their income tax if the rate is different than if the rate is the same and the amount of tax that they are paying is no different.
If the UK raised its rate by 1 per cent from 20 to 21 per cent, that would go through very smoothly, because it would require payroll providers just to slot in a different figure. Is that not right? Does the fact that we will be doing this through the code rather than through the rate create more of a problem?
I do not think so, although that, too, will raise questions.
A change in the UK rate would have a similar extra administrative cost, because we would have to change people’s PAYE codes, and they would be more likely to phone us up to ask what was going on. There would be a similar sort of cost if there were a UK-wide change in the basic rate.
But would somebody’s code change if the tax rate were changed?
It could, because some items would be, if you like, coded out. Not everybody’s code would change, but where the tax code was adjusted to recover a precise amount of tax, the code would have to be different depending on the marginal rate that is assumed.
Okay.
Typically, it would happen in the case of benefits in kind that we were notified of at the end of the year. That would give rise to a tax charge for the previous year that would need to be recovered through the tax coding for the following year, and that tax coding would have to be adjusted to reflect the rate at which we would be collecting the amount of the benefit-in-kind charge, which would be the rate for the second or third year. That is where the actual coding change would depend on the status of the taxpayer if there were a different rate in Scotland. The PAYE system is very effective at helping us avoid issuing self-assessment returns to 30 million people, but that makes it quite complicated internally, precisely because it collects the right amount through the coding.
Thank you.
I want to return to the highest risk that you have identified, but, first of all, I have to say that I love colours on a chart—they make things wonderfully simple.
The convener was right—you are unlikely to be issuing the notification letter in December. However, I fear that you will not be doing so until you have certainty over the SRIT, and the budget is not actually going to be passed until February. Given that, do you think that there is a significant risk of pressure on your service in the form of a huge number of inquiries over a short period of time, say, from February 2016 until implementation? I absolutely agree with Ms Walker that employers are well aware of what is happening and I know that your communications about payroll have been very good, but the fact is that employers are champing at the bit to tell their employees, and they just do not know what to tell them. Could there be a stage before, in which those people could be used to put out key messages about the Scottish rate of income tax? You could follow that up in February with a letter explaining the rate and so on.
We will not necessarily leave the letters until after the Scottish budget, although we do want to know when the Scottish budget is going to be. We might well feel that the right thing to do is to issue the letters in December, saying that there will be a Scottish budget and that it might set a different rate. It will be a matter of writing to people, saying that we think that they are Scottish taxpayers, and the letter will go out of its way to say that, if they are, they do not need to do anything. Basically, it will say, “Please don’t ring us up.” That said, there might be people who, for some reason, think that they are not Scottish taxpayers. We might write to them at a Scottish address, but that might not be their main residence. The letter might have been forwarded to them because they have just moved, or the address might be a holiday home and they just use it for convenience or whatever.
We could still—and might well—send the letters in December. We feel that if the Scottish budget were to take place in December—although you are more or less telling me that that will not happen—there would be a greater level of awareness if we sent the letters around the same time, and there might be more of an impact than there would be if the budget did not happen then. As you have rightly pointed out, if we leave things until February, it will be very close to the implementation date. For us, it will also run into the self-assessment peak, which is around 31 January, and that will put significant pressure on our call centres.
Now that the Parliament has been reconvened, we hope that we will get certainty over the date soon, and we will then make a decision quickly about sending the letters. My expectation remains that we will send them out in early December, but we have not finally agreed that yet.
That is helpful. We, too, are a bit in the dark about exactly when the budget will be. We reckon that the budget process will start in early December, but it will not finish—and a rate not absolutely confirmed—until February. You can still exercise your judgment about maximising the impact on that basis.
I want to ask about advertising. We might know about the Scottish rate of income tax but, when I talk to people in my community, they have no clue what I am talking about. In view of that, are you taking a wider approach than simply sending a letter? How much money are you spending on that activity, and where are you spending it?
I will let Sarah Walker fill in some of the details, but first of all I must make the overarching point that our responsibility is to ensure that the tax is administered effectively. We are obviously happy to support any wider messaging about the tax, but it is not our role to raise awareness per se. We need to ensure that we have the correct and most accurate set of information possible in order to administer the tax properly, which means that, as far as value for money and direct effectiveness are concerned, HMRC’s—and hence your—paying for some billboard campaign up and down the length of Scotland is probably not the right thing to do. After all, we believe that we are already at about 98 per cent accuracy.
As a result, writing to the people who we think are Scottish taxpayers through other soft means—for example, through promoted articles, through conferences or by going through the professional community, the estate agent community and some of the more interesting professions in terms of mobility such as offshore workers, people in the fishing industry and students—will be more effective for us than a wide advertising campaign. At some stage, however—once we have sent the letters and done what we can through soft communications—we will want to have some wider campaigns, and Sarah Walker will fill in the details, to the extent that we have them, about what we propose to do on those campaigns.
In the original budget that we shared with the Scottish Government, we had provisionally allowed for quite a substantial advertising campaign, because we thought that we might need it if we did not have confidence in our address data. Now that we have better confidence in that, we do not think that a widespread advertising campaign would represent value for money. That is not because of any constraints on the budget—it is the Scottish Government’s budget—but simply because we have to take responsibility for what represents good value for money.
We expect there to be a lot of interest in the Scottish media and the press, particularly around the time of the Scottish budget, on the Scottish rate and the arguments around how it should be set, and we expect to try to build on that interest by, for example, offering information to the personal finance pages in newspapers to explain how it works. We will also offer employers material to put in their newsletters. We will use all sorts of what might be called non-paid-for community channels to put out that information.
As Edward Troup has said, we will look at targeted paid-for advertising later, as and when we think it is necessary, to ensure that we reach any audiences that we think might have been missed. However, I repeat that the letters are our main means of communication.
I confess that I had heard that there were constraints on the budget. What I am hearing from you now is that there is no intention to advertise—unless you need to do so—because you think that you are covered. If I am sitting at home watching TV, there will be no television adverts, and there will be nothing on the radio. There will be nothing of that nature to inform the wider public about this.
At the moment, we do not plan to do that. As Sarah Walker said, we got the message at a meeting with some colleagues of ours that we had said that there were constraints on our budget. That is absolutely not right. What we said was that we wanted whatever we did to be driven by value for money and the likely effectiveness of the actions. At the moment, television advertising and billboards do not look like good value for money in chasing down what we think is going to be a very small proportion of the population—people for whom we do not have accurate details.
Can I press you on how much you originally planned to spend and how much you are spending now?
We originally quoted £4.2 million, but that was not all for paid-for advertising. It was for the overall campaign, including the cost of sending the letters. I do not think that we have revised that. We do not have a revised budget and we have not shared one with you or the Scottish Government. We hope that the cost will be significantly less than £4.2 million, but our written communications—our letters—come with quite a significant cost.
Sarah, do we have a breakdown of where we are on the £4.2 million?
We do not have a specific figure for paid-for advertising, because we are still trying to work out the best targeting for that.
Could you supply the committee with the revised figure?
We can tell you where we expect to be. We think that the impact of the letters will be affected by the timing of your budget, although that is not to say that, if you have a January budget, they will not be as effective. There might be a second bite of the cherry. If we send the letters in December without an indication of the rate and then there is publicity around your budget, that will be an opportunity to have a push through the professional and other trade journals and the wider non-paid-for media such as newspaper articles. At that point, we will be able to judge how well we are doing and hence what we want to spend on further paid-for advertising.
We can give you a range for roughly where we think we might be, but I am not sure how helpful that will be. We have shared with the Scottish Government an outline of the principles of the communications campaign, which you might have seen.
It would be helpful if the committee could see that. Any information that you can provide on finance would be helpful, because value for money concerns us, too.
I move on to my final question. Again, I might not have understood the first table in your update very well. IT projects are notoriously difficult to implement, so I share the pleasure at what you are telling us about the initial implementation, but I also have a slight concern. At the bottom of page 7, you mention a
“risk that the Project does not manage its relationship and dependency with the Digital Programme.”
What is that risk and what have you put in place to mitigate it?
Sorry—I am just trying to track through to what you are looking at.
It is in the helpful table at the bottom of page 7. It is the last item in your risk register.
12:00
Oh, right. Yes. Do you want to talk about that, Sarah?
Yes. We have a major programme at the moment of implementing digital services for both personal taxpayers and business taxpayers. That is a big preoccupation for the department. We need to make sure that what we are doing for the Scottish rate, which is slightly more of a niche project, fits in with the overall plan and makes the best use of the new digital services that we are putting in place. For instance, we are putting in place personal tax accounts for taxpayers so that they will be able to, in a secure environment, log on to their own records online and update them. Clearly, that will be a really useful thing for us to use to help people to update their addresses and to see how much of their tax is being paid to the Scottish Government and how much is for UK services. We want to make use of those digital services and we need to make sure that we are well plugged into the work that is going on elsewhere in the department to produce those new services for people.
How confident are you that we will not be disadvantaged in terms of accessing those new services?
It is going really well, which is reflected in the fact that the risk is coming down.
I will pick up a point that Sarah made. There are obviously always risks in any large system, but the digital services will be a positive opportunity because we, not the taxpayer, will pay for the running of personal tax accounts. It will not be a cost for Scottish taxpayers and it will be an opportunity that will allow them to have more access. At the moment, for example, people can change their address online but cannot see their personal tax affairs, because they do not have a personal digital account. When we get those accounts operating, that will help the operation of the Scottish rate.
Thank you both very much.
I want to continue with the theme of communication, but in terms of businesses rather than individual taxpayers. The taxation issue is certainly not the talk of the steamie in any of the areas that I represent, but you believe that there is a really good understanding of it among payroll managers, who use Sage and other programmes. I am concerned about the many people who are responsible for PAYE systems in Scotland—particularly in the Highlands and Islands—who are very small employers with three or four staff and are likely still to be operating a manual wage system. Sage will not be known to them. What kind of communication about the issue can we expect from business organisations? The Federation of Small Businesses, for example, has a large membership across Scotland, and there are similar trade organisations here. Is it better that they deliver the message in newsletters or other ways? Will there be a series of workshops with them in order to spread the word?
Can I step back a bit? If there is no change to the Scottish rate in the first year, in a sense we will have longer to continue to engage with those employers, because they will not need to do anything different. Their tables will change, because they will change with the personal allowances et cetera this year. In a sense, it will not be a huge issue for them. If there is a change in the Scottish rate and the budget, whenever it is, sets a rate different from 10p, we can anticipate a huge amount of interest from individuals and employers, because they will read that the tax rate has changed and they will want to know what has happened. That is why there will be costs for us, because they will ring us up. At that point, I think that all employers will become extremely aware of the need to do something.
We obviously have to plan for both those scenarios, but particularly for the change-of-rate scenario. Because we have engaged well with the payroll software providers, I think that we can be confident that, if there is a change of rate, employers who use software will be well taken care of. They will be reassured by Sage or whatever programme they use that any change has been taken account of in their software and that it will all be fine.
Those who are still doing PAYE manually—I do not have a figure for them—will get communications from us, because if they are doing it manually they will be in communication with us. They will get communications that will update their tables to reflect where they are. I suspect—although I may be wrong—that for a small employer in the Highlands and Islands who does PAYE manually there would be no question about the residence of their employees. I imagine that those employees would all be Scottish resident and relatively local, and such an employer would not face the complexity of having different codes for different employees. I suspect that, because manual payroll operators are likely to have all Scottish employees, they will have a simple set of tables for those employees.
As I have said, from April, all employers across the whole UK will, potentially, have to be able to operate a Scottish S code. We communicate with all employers through the “Employer Bulletin”, which goes out regularly and has contained articles about employers’ needing to be ready for the Scottish rate. We are also speaking to organisations such as the Federation of Small Businesses, and we are talking to the accountancy profession about it because a lot of small businesses will consult their accountant about their PAYE system and how to operate PAYE. We are doing everything that we can. We have put on the www.gov.uk website more material that is specifically targeted at employers, and they can use that for reference. We are using all the communication channels that we have to ensure that all employers are ready for the change.
Maybe I am missing something. If the rate changes in the budget, a lot of people will be interested in that, particularly if they are going to pay a higher rate of tax. However, the Scottish rate of income tax is just that—it is already a separate tax, whether it is 10p, 11p or 13p. People will obviously have an emotional interest in how much tax they pay, but I cannot quite see that whether the tax is 10p or 11p will make a massive difference to its collection in practice. John Mason noted your final point, on page 6 of your submission, about the dramatic increase that there would be in the cost of collecting the tax if it was set at above 10p. However, surely the Scottish rate of income tax must be identified in your offices.
That is the point. If the rate is not changed from 10p, the amount of tax that is collected will be entirely unchanged. The basic 20p rate of UK tax will be replaced by a 10p rate of UK tax and a 10p rate of Scottish tax, but the aggregate amount will remain exactly the same, the employees will see exactly the same deduction—all things being equal—and the same amount of tax will be paid. The employers will, in effect, do the same thing. The work will have to be done by us, because we will have returns for Scottish taxpayers that are designated by the S code and we will, through our systems, have to aggregate out how much tax at the Scottish rate has been collected. For the first two years, because of the transitional arrangements, even that will not impact on the amount of money that gets adjusted through the block grant and paid across; nevertheless, we will do those calculations. We will do all that work on the basis of the S codes. For the employer, if the rate does not change from 10p, nothing at all will change in what they have to do beyond their having to ensure that they have the S and not-S designations on their systems.
However, if the rate changed, a different amount would be collected. If the rate went up to 11p or down to 9p, 21p or 19p in the pound would be collected, which would represent a change in the amount of tax being collected and a change in the deduction that the employee would see. We would still have to do the same work behind the scenes, but that change from the 10p rate, rather than the introduction of the 10p rate, would cause—I will not say “challenges”—interest, activity and questions among employees and businesses.
I am not sure that I understand the difference. I am imagining tax collection involving the Inland Revenue delivering bags of money to Scotland—it is impractical; like a kind of cartoon sketch. Whether the sum is 10p or 11p does not seem to make a difference. I still cannot determine the costs that you are relating to the collection of that or the differential from the employer’s point of view. If an electronic system is set up to collect an S tax or an E tax—for want of better expressions—then, presumably, it will work in exactly the same way.
It will do. Leaving aside the manual operators and focusing on the electronic side, which concerns the bulk of people, the majority of the additional cost will concern the fact that people will see their tax bills changing and will have questions about why that is happening. To be honest, the system might not operate correctly in every case—for example, because people have been wrongly designated as Scottish taxpayers due to a house move or whatever. That will create activity, and that is where the additional work and the additional burden arise.
With regard to the example of where the rate does not change, it does not matter—except to us, because we need to know—whether someone is correctly or incorrectly designated as a Scottish taxpayer, because their tax burden remains exactly the same.
Would we have the detail of that difference when the time came? I mean the increase in the cost—presumably we are paying for that.
Yes. We would continue to provide a breakdown of costs.
I have a couple of questions. The first concerns the differential that the deputy convener and Jean Urquhart touched on. Were those additional costs to materialise, would they be absorbed within HMRC? I have not seen any sort of recharging system mentioned in any of the papers, so I assume that that would not happen.
Sorry, but which additional costs are you referring to?
You have said that, if the SRIT was set at a level other than 10 per cent, the costs might be between £2 million to £2.5 million and £5.5 million to £6 million. Will those costs be absorbed within HMRC?
No. Those costs will be re-charged to you. We absorb all the fixed costs of running HMRC. The costs that are recharged under the agreement on costs are costs that are specifically identifiable and are identified. They are audited and recharged to you.
And those additional costs remain the same irrespective of the rate above 10 per cent that is applied. There is no sliding scale.
The numbers are estimates, so the cost might not be exactly that amount. The additional cost arises from the change, and whether the rate is 12 per cent, 15 per cent, 9 per cent or 6 per cent, it will not make a material difference to the additional costs. The costs will be what they are. The numbers are not a quote, and they are not tax linked. They are our estimate of the cost of the additional activity that will be specifically attributable to servicing Scottish taxpayers and employers as a result of the Scottish rate change.
My next question is about the risk register. It has been a long time since you last appeared before us, so forgive me if my recollection is incorrect, but I seem to recall that I questioned where the numbers sat in terms of probability and impact and I asked for those details to be given to the committee. Would you be willing to do that? When we look at the information in the current format, we see that certain things show up as amber or red, which sets alarm bells ringing. However, when we look at the probability or the impact, there is less concern.
I am sorry if we failed to follow up on something that I said that I would do last time. I will make sure that we do that this time.
I could be recalling incorrectly, but I seem to remember asking for that information.
The red, amber and green colours, which relate to the numbers, are based on a grid that concerns probability and impact. There is a standard convention that determines which of the corners of the grid we mark red, amber and green. We can certainly share that grid with you so that you can see how the colouring relates to the probability and impact scores that the numbers have. However, as I said at the outset, giving one of three colours to figures as complex as these risks oversimplifying the position quite dramatically.
12:15
Sure—I appreciate that.
Annex B shows the previous year’s risk matrix, and some of the indicators have changed marginally. The obvious change is the risk that has gone from red to amber, but that is only because of a scoring difference of 1. There has not been much change in the intervening period, yet in the current matrix most of the risk scores are forecast to drop quite materially. If there has not been much change since the previous period, what leads you to assume that those forecasts will materialise?
Sarah Walker can add some details, but I have a list of the risks that have changed since the previous matrix was produced. Five risks that were on the previous matrix are not on the current list. In other words, the confidence that we expressed last year that the risks were coming down and that risks would be resolved has been fulfilled, and we have taken five risks off the risk register. There are one or two new risks. Sarah might be able to call them out to me.
The document is dynamic in the sense that we have been moving broadly—or rather, quite firmly—in the direction of improving the management of the risks and landing the project.
Some of the changes might relate simply to the stage that the project is at. We are moving to a point at which some of the risks are likely to either crystallise or not.
In recent months, we have improved our preparedness for implementation, but we have perhaps not been able to form a view as to how likely some of the risks are to reduce. We are now moving into the implementation phase, and things will start to change pretty quickly. Some risks have changed between the previous matrix and the new one, and some new risks have arisen—for example, the risk regarding the timing of the Scottish budget. The document is definitely dynamic.
An example is the third point in the current register, which is:
“There is a risk that - the Project does not keep abreast of wider transformational change which could result in the solution not sitting within the HMRC operating model circa 2016.”
Looking back at the 2014 figures that you have given us, I note that that risk was showing as 13, on the amber scale, and it is now showing as 14. I appreciate that the change is probably marginal, although I do not know whether it relates to increased probability or increased impact. However, the forecast is that that area will be A-okay and the risk indicator will be green by the end of the process. What leads you to that conclusion when the movement that has taken place appears to have involved increased rather than decreased risk?
The increase in the risk relates to the fact that we have a spending review coming up and our financing—and, in a sense, HMRC’s operating model—is always up for review at that time. There has been increased uncertainty over the past six or seven months about our operating model and the structure of the department as we move forward.
I remain confident that the spending review will give us a sustainable and stable transformation model that will bring that risk down, but until the review is out there and we have settled our internal budgets for our transformation and plans for the next five years—or at least the next three years—I cannot be certain of that. That element therefore remains in the document. It has been a bit more worrying during the year because of the uncertainty around the spending review, but I remain confident that it will come down on the right trajectory.
That teases out some of the difficulties and subtleties within those risks. In a sense, it is good to have the challenge, but what is important to me as the accountable officer in the department is that the risk register is maintained and that each of the risks is—as I feel it is—being actively managed. If we have red ones, I worry about them, but I am more concerned to know that there is someone in Sarah Walker’s team or in the delivery teams who is on top of the individual risks and is driving them down, and I am satisfied that that is happening.
I appreciate that. Getting the deeper information would allow me to look at probability versus impact.
My only other question is about the control effectiveness. I assume that the reason for the very low control effectiveness in relation to the timing of the letters is that it is determined by the timing of the Scottish Government’s budget, which is not something that you have control over. Where it says “Moderate”, is that because there are some factors that are outwith your control and some that are within your control?
I am looking at the ones that are moderate. Yes—absolutely.
To ensure that there is a smooth transition to the SRIT, it is important that there is the right exchange of data—and ease of exchange of data—between you and the Scottish Government, particularly at an official level. How are you ensuring that that takes place? How often are you meeting? What discussion channels do you have? Are you satisfied that you have all the data that you need and that there is willingness to co-operate? Is that being reciprocated, or are you confident that it will be reciprocated?
We have a very good relationship with Revenue Scotland and the Scottish Government. Indeed, I am going to see them this afternoon, as I generally do when I come up here. However, I am not sure that, when it comes to running the SRIT—which, although it is a devolved tax, will be administered by HMRC—we are dependent on the Scottish Government for a huge amount of data exchange. Sarah Walker might correct me.
We are not.
I am thinking of things such as national health service records, which you mentioned, and the timescale for the budget. There might be issues that the Scottish Government wishes to discuss.
We cover both of those issues. The position on NHS data that is held by the Scottish Government is the same as with any other data set holder. We say, “Can we look at the data? It will help us.” The fact that we have not done that in the case of NHS records does not mean that there are not good relations.
The timing of the budget is the timing of the budget. We work closely and share our proposals for publicity on the SRIT all the time, and we have been closely in touch with the Scottish Government on the devolution of the other two taxes that Scotland has already taken control of.
On the point about regular contact, the Scottish Government is formally represented on our programme and project boards, which meet monthly. In addition, I have a formal catch-up with Sean Neill, who was here earlier, at least fortnightly and often more frequently than that.
We have an open dialogue and I am happy with the co-operation and help that we are getting from the Scottish Government. For instance, it is important for us to co-ordinate the messages that we are giving about the Scottish rate with other publicity and communications that the Scottish Government carries out.
There is a chance that people who have questions about the Scottish rate will phone up Revenue Scotland rather than us because they think that it will be operating the tax. We therefore work closely with Revenue Scotland to ensure that it has the right information and can pass inquiries on to the right place in HMRC.
That is helpful.
That appears to conclude questions from the committee. Are there any further points that you wish to make?
No. I just look forward to hearing about the timing of your budget.
We all want to hear about that. Thank you for answering our questions so comprehensively.
At the start of the meeting, the committee agreed to take the next four items in private.
12:23 Meeting continued in private until 12:38.