Blame game won’t solve duplicate employment records
Duplicate employment records lead to double trouble, but getting to the root of it is down to communication, not blame, says Ian Holloway.
Replies (10)
Please login or register to join the discussion.
A useful synopsis of some ways in which HMRC's systems can process data (that has been submitted correctly) before being transformed into incorrect data within their own databases ... in this case fake 'duplicate' employments.
HOWEVER I dispute that it (or more importantly the advice copied from HMRC's Employer Bulletin) is remotely practical:
Case 1 - The new starter -
"first FPS will be the first time HMRC’s systems are aware of the new employee in this employment.
Yet, the first FPS should be submitted after the previous employer has submitted their final FPS containing the leave date.
How do we check if the previous employer has submitted their final FPS?"
Consequence: it doesn't matter what you (the new ER) knows/guesses, you will usually be stuffed (to use an old technical term).
Say EE leaves oldCo on 9th Jun and starts with newCo on 12th Jun ... oldCo only processes final payment on 30th Jun (their regular payday); whereas newCo processes first payment on 26th Jun (their regular payday).
Result = newCo's FPS is submitted before oldCo's FPS, so HMRC are unhappy;
OR = newCo holds back first payment to new EE to the next month, so EE is unhappy.
[HMRC would like oldCo to process a leaver payment on final day (outside of their usual pay cycle), but are too scared to suggest this as not all software is designed to cope with doing this].
Case 2 - A leaver under TUPE -
"HMRC’s view is that this is a leaver under the old PAYE reference and a starter on a new reference.
This is often recognised by employers who will generate a P45 but will not give it to the employee.
Yet, even though there may be a need to retain the original start date in software, say for seniority purposes, this is a leaver and starter process and must be reported as such."
Consequence: there is a direct mis-match between what the RTI data item is labelled (such as 'Date employment contract ended or state pension or taxable benefit ended') and what HMRC actually want you to report ('Date payments under this PAYE scheme ceased').
Of course most Payroll software doesn't exist in a vacuum and tries to avoid duplication of any data being entered (a cardinal sin in software design), so the designers will where possible 're-use' an existing data item (such as 'employment end date') - in order to make life simpler & intelligible to users - which was sanctioned by HMRC all through the pilot stages of RTI and the next decade.
Result = a perception that HMRC is asking ERs to 'lie' in order to enable them to process the FPS in the way that they'd assumed (but not made clear to developers) ... and all because they have no concept of employment legislation or of integrated systems used by ERs who are seeking greater efficiency.
I could go on (and on) despite pleadings from the now comatose audience ... but fully endorse Ian's closing comment:
"it still does not take away the fact that employers are using up-to-date, 21st-century software that is 'speaking' to legacy HMRC systems" ... whilst adding that those HMRC legacy systems are being maintained by people who appear to be clueless that Payroll is no longer a stand-alone system (in terms of software or indeed as part of business & tax reporting processes)!
How to reduce my maunderings, whilst retaining the essence, in just a dozen words ... brilliant!
On reflection, I could have got away with 9 words. "Just a thought" are somewhat redundant to the point. Though perhaps HMRC will then double count them and put the total up to 15.
Commonsense goes out of the window. How often does someone on your payrolls double their basic monthly pay? In my own case - processing around 100 staff per month for the last 14 years, so that is 16,800 or so monthly payslips - the answer is zero.
How hard can it be? Simply build in a sense check fo a monthly paid person increasing his or her pay in a month by more than, say, 180%. (Note this does not apply to those paid by the hour where big fluctuations are not unusual.) Then check back with the employer. Oh, and finally, actually take responsibility for the fact that your stupid system double counted the pay, and hence it is down to HMRC to fix that, and not point the finger at Tom, Dick, Harry, Uncle Tom Cobley and Uncle Tom Cobley's dog.
Commonsense goes out of the window. How often does someone on your payrolls double their basic monthly pay? In my own case - processing around 100 staff per month for the last 14 years, so that is 16,800 or so monthly payslips - the answer is zero.
How hard can it be? Simply build in a sense check fo a monthly paid person increasing his or her pay in a month by more than, say, 180%. (Note this does not apply to those paid by the hour where big fluctuations are not unusual.)
Doesn't this depend on whether people get bonuses as a significant part of their remuneration? I suspect there are a lot of people who do see their monthly pay double (and more) from one month to the next during bonus time.
Quite. And when the payslip is checked, the line "Bonus £5,000" or whatever will appear. And you will also spot that only 1 payslip was received for the month.
Putting this another way, in the world of commonsense when your system receives two payslips each for exactly 3,525.00 what is your first thought?
If it is such a wide spread issue (we have multiple clients affected) why isn't there a dedicated resolution team? We're getting HMRC blaming software providers and software providers blaming HMRC and no one providing an actual fix.
It should be blatantly obvious where there are two employments in the same name and something has gone wrong but we've had to resort to employees signing 64-8s and mass letters to get one of ours resolved despite the Dispute Resolution team being involved.
People claim tech will run the world. The people who code the stuff can't get the basics right, so what chance they have with anything more complex, I don't know. Still, keep hoping for your self driving cars - and your hoverboards.