Kate Upcraft pulls apart the long awaited post-implementation review of RTI, and considers the lessons to be learned for the MTD project.
The publication of the post-implementation review (PIR) on RTI has not been as speedy as its introduction in 2013.
The Office of Tax Simplification and Admin Burdens Advisory Board asked HMRC to progress this review in spring 2015, as the RTI project had moved to business as usual in October 2014. Feedback was not sought until spring 2016 and some wondered if the PIR would ever be published publicly.
It’s testament to those of us who have worked with HMRC since 2010 that seven years later our feedback has led to a set of actions.
The scope of MTD is even more radical than RTI so sharing the lessons learned from this project across government is vital.
The report, to its credit, does reflect the widely differing views from stakeholders on costs versus benefits. It acknowledges that there has not been the predicted £292m savings for employers from RTI, but rather transitional and ongoing costs of £292m – quite an interesting juxtaposition. I find it hard to understand how the deployment of RTI can be cost-neutral for employers and agents.
Given the cost of MTD for businesses is still widely believed to be understated, this a salutary lesson to government that requires addressing.
I take the issue with the assertion that it was HMRC’s flexibility that led to the deferral of automated penalties for late submissions, a situation which continues after nearly five years.
The true reason for the deferral is that the HMRC data wasn’t robust enough to deploy automatic in-year penalties. It concerns me that relying on a risk-based penalty regime (human overview) has allowed some employers to ‘game’ the system at the expense of those who have incurred additional costs in an effort to be truly compliant with the ‘on or before’ reporting deadline.
Actions to delivered by 2019
We are told that HMRC is ‘putting a programme in place to deliver …. recommended changes ….by 31 March 2019’. I wonder why they are still putting a programme in place when there have been obvious, widespread and ongoing problems with data and systems since 2013.
The action plan emphasises HMRC’s view that employers are not getting the message about failures to operate RTI correctly. Payroll professionals have been asking since 2013 if HMRC could outline what those failings are. I’m still not clear what specific mistakes employers are supposed to be making, and whether these are process or software driven errors.
The RTI Customer User Group, which I was part of until HMRC disbanded it, has constantly urged HMRC to acknowledge that simply telling employers to do something that doesn’t fit with commercial practice was a pointless exercise. For example, many large businesses want to keep one employee payroll number throughout an employee’s tenure with them and changing payroll software, agents and the structure of businesses is part of business as usual.
HMRC should have therefore advanced the work it began on a unique payroll ID separate from an employee number or NINO. That would have helped considerably in reducing duplicated records, which often arise after a restructuring of the business or a change in payroll software.
The PIR says HMRC has set up a core communications team to focus on employer-related issues. I’ve seen no evidence of this team meeting that need, which is a pity as HMRC still says communications will be a major solution to ongoing issues.
The Employer Bulletin is still the only employer communications vehicle and that carries messages on a wide range of subjects, which don’t all emanate from HMRC. It would be good to see a dedicated communications vehicle for RTI issues that uses the expertise of the Employment and Payroll Group (EPG) to ensure the information is practical and accurate.
Disputes about data are now to be referred to as ‘discrepancies’ as HMRC says stakeholders find that a more neutral description. Whatever they are referred to, the promised action in the report to ‘facilitate amendments both during after the end of the tax year’ is not progressing and needs prioritising.
In December 2014, HMRC gave a senior-level commitment to re-design the Earlier Year Update (EYU), to allow employers to correct end of year errors they had discovered themselves. The redesign would also allow the employer to adjust the data when requested by HMRC staff to correct corruptions introduced to the data by the HMRC central systems.
This promised redesign of the EYU has not happened three years later, and the PIR offers the solution of an API to allow payroll software to display the HMRC data to the employer or agent, not to correct it.
In the last few weeks, software developers have been asked if they would develop such an API if it was just scheme level data, not employer-level data. This will not assist in employers in establishing where any data discrepancy lies, we can see scheme level data on the Business Tax Account (BTA) already, and we have been told that since its recent upgrade the BTA is now an accurate reflection of the data HMRC holds.
Of the on average 40,000 formal disputes logged each year, the report says ‘78% of the outstanding debt was spurious’. That must be a concern for the Treasury given the need for accurate tax receipts to inform public borrowing.
Equally 40,000 disputes only relate to the schemes that have logged a formal dispute, many employers with a payroll agent will have no knowledge that there is a data discrepancy, as their agent has no access to the client's BTA.
The report notes that 68% of disputes are from small employers. Many of these disputes will only be initiated when the tax agent needs to react to inappropriate compliance activity by HMRC and has to register a dispute to protect their reputation with the client.
Despite the IT fixes that have been deployed to reduce the numbers of duplication issues the PIR says ‘0.5% of records of the employee population are affected by a duplicate’.
Does that mean 0.5% of 40m taxpayers, 0.5% of 590m records received p.a. (as confirmed in a recent Parliamentary question) or 0.5% of 100m transactions per month as detailed in this report? Even the lowest end of that range is 200,000 employees.
In December 2016, Jane Ellison (then chief secretary to the Treasury) told the DWP committee that after IT fixes in January 2017, duplicates would be ‘minimal’. Saying 99.5% of employee records are not affected by duplicates is surely not 'minimal'.
In my view, there should have been a quicker focus on other solutions such as unique payroll IDs, a robust way to correct end of year data, and most importantly better data validation between payment data and the Full Payment Submission.
The report references the fact that DWP and HMRC have had to establish a joint team to resolve data discrepancies between RTI data and payment data into bank accounts. It is hard to see as UC rolls out at scale how this approach can be sustainable. It shouldn’t be necessary for there to be manual intervention to ensure RTI data is fit for UC purposes, let alone for tax credits or the collection of tax and NI.
The key sentence in the 40-page report is: ‘We fully accept the lessons learned on consultation, communication and implementation’. It’s good to see stakeholder views reflected in print, particularly where they differ from HMRC’s and that actions are allocated to some of these views.
However, actions speak louder than words and so far, the pace of acknowledging and then fixing issues with RTI is too slow. Now there is a published action plan we need to press HMRC to deliver this by the promised date of March 2019 - six years after RTI went live.