HMRCs new PS tool

[northernladuk]

My testing seemed to happily generate outside answers without RoS. I don't think its that strong an emphasis.

Phew!!

[Jessica@WhiteFieldTax]

OK, I just tried a few test cases.

I think they're going to struggle to achieve what they want (90% caught) if the current weighting matrix is anything like the final matrix. It's not difficult to achieve an outside position. Basically, from what I can tell, if you have a legitimate RoS or a "reasonable" lack of D&C, you're going to be outside. I tried both the actual terms of some recent contracts and a few variations thereof, and they all exited very quickly with an outside position. I suspect you'd need a lack of RoS coupled with a moderate to high degree of D&C to have an inside determination. Just an initial impression though, as I only tried a few cases (back to proper work now )... but, so far, I haven't managed to generate an "inside" position

That said, I'm not sure it really matters. Are PSBs really going to use this with an open mind? Doesn't it really come down to whether the PSB actually wants a contractor or a deemed employee? In other words, they might use the tool to evidence a position, but probably not to determine one.

I think your last paragraph is pertinent, also serious questions about who will complete it at the client and hw much understanding they will have of the nuances.

[teapot418]

My testing seemed to happily generate outside answers without RoS. I don't think its that strong an emphasis.

Has your testing thrown up any inside answers?

[Jessica@WhiteFieldTax]

Has your testing thrown up any inside answers?

Yes, based on weaker SDC factors.

[Maslins]

That bit that I've put in bold hinges on HMRC having a record of exactly what the questions were at the time of asking, what the client answered, and linking that back to what HMRC have said will be an anonymous tool. I'm not convinced that HMRC have considered a mechanism for storing that information, and I'd put money on them not having a mechanism for retrieving that information at any point in the future.

I haven't been given access to the beta, but I imagine whatever you can print at the end would include all the questions and answers you gave?

The anonymous thing does seem to work in the contractor/public sector body's favour. Ie they do the test once, it says inside, HMRC don't know about that (or at least don't know who it was/where they work etc). They ignore this result. They then re-do it with different answers (which hopefully they can justify) until they get the end result they want. At that point, they print, sign and date it, basically so it's no longer anonymous. Apparently this will then be binding.

[RonBW]

I haven't been given access to the beta, but I imagine whatever you can print at the end would include all the questions and answers you gave?

The anonymous thing does seem to work in the contractor/public sector body's favour. Ie they do the test once, it says inside, HMRC don't know about that (or at least don't know who it was/where they work etc). They ignore this result. They then re-do it with different answers (which hopefully they can justify) until they get the end result they want. At that point, they print, sign and date it, basically so it's no longer anonymous. Apparently this will then be binding.

A printed copy isn't going to mean very much if it can easily be faked, so you'd need to have a record that cannot be manipulated by anyone as an audit trail. I could get an out judgement and knock up something with questions on the bottom to indicate what was "asked" to get there quite easily and fabricate a printout in not much time. HMRC would also need to have a copy of what questions were asked and answered for every challenge that is made to an inside declaration otherwise the tool is meaningless.

The people that I've spoken to in the past have indicated that HMRC had not considered that there may ever be a need to accurately report what questions were asked and what answers were given. This may have changed in the past few weeks, but I'm not confident that HMRC will have built anything reliable, scalable and reportable this calendar year to allow for any historical reporting of what was asked and answered - and if the questions change in the future (as HMRC have already indicated that they will) then there will be no way to validate any inside or outside determination by the tool.

There may be trouble ahead etc etc

[jamesbrown]

I haven't been given access to the beta, but I imagine whatever you can print at the end would include all the questions and answers you gave?

The anonymous thing does seem to work in the contractor/public sector body's favour. Ie they do the test once, it says inside, HMRC don't know about that (or at least don't know who it was/where they work etc). They ignore this result. They then re-do it with different answers (which hopefully they can justify) until they get the end result they want. At that point, they print, sign and date it, basically so it's no longer anonymous. Apparently this will then be binding.

It does, and it also includes a unique reference number, so there's a permanent/traceable record; it's only anonymous until they join the dots upon a specific status inquiry (if you're claiming outside).

Iterative doesn't necessarily matter if the PSB really wants to fill a contract as outside, but the fee payer will be on the hook if it's all BS, so they better be satisfied too. Hence my comment above about using the tool to evidence a position rather than decide it (in most cases).

[WordIsBond]

A printed copy isn't going to mean very much if it can easily be faked

HMG may be very, very poor at strategic thinking, but HMRC are not stupid at tactical thinking.

They will archive and date any changes to the tool and will be easily able to prove a printed copy couldn't be produced by the tool, and is faked. Then, they'll not only nail you for tax evasion but for perverting the course of justice, and a few other unpleasant things, if you've tried to use a faked printout.

I hope some people DO try it on. It will be nice to have those people stupid enough to try this end up with massive fines to help support my state pension in coming years. Might even end up saving my kids a few pence in taxes over the years.

[RonBW]

HMG may be very, very poor at strategic thinking, but HMRC are not stupid at tactical thinking.

They will archive and date any changes to the tool and will be easily able to prove a printed copy couldn't be produced by the tool, and is faked.

I refer you to the post I made some moments ago - according to people close to the project, until not so long ago the prospect of needing to retain an audit trail of what questions were asked and answered had not occurred to HMRC. According to the same people, there would be a number and an outcome available, and a release note of what version was in operation at the time of completing the tool, but no record of what answers had been given. And there would be no way for the client to get a copy of the questions asked and answered in the event of future need.

That may have changed in the past month, but I don't have faith in HMRC to have suddenly produced it out of thin air.

[jamesbrown]

Yes, based on weaker SDC factors.

As far as I can tell, it needs to be fairly weak though, as in the PSB telling you what to do, when and where to do it, and how to do it and moving you across projects as they deem fit. If all of those apply then you're clearly caught in reality anyway (on D&C). But I haven't tried moderate combinations of all of them (because each question is multiple choice, and there are generally ~4 choices). Moderate combinations of D&C across the above elements may lead to an indeterminate or caught scenario in the absence of a legitimate RoS (which immediately terminates the process as outside).

My sense is that they're really only using this tool to catch the very low-hanging fruit and that the rest of what they want will be achieved by: 1) ensuring that PSBs make a blanket determination upfront (primarily this); or 2) they dispute a claim in the usual way. Even on (2) they're winning, because there will be fewer claims (see 1) and they will have a permanent record of what was claimed, upfront, and they'll be able to check that systematically at the beginning of a dispute (less of a blank canvas). However, the basic assumption underlying this is that the beta is similar to the final version, and that the decision matrix/weighting isn't wholesale changed, otherwise it could catch several higher hanging fruit (not that I'm calling NLUK a fruit, you understand ).