Pension Dashboard - Distressing to see the reasons behind the interminable delays

hugheskevi

poseidon1 said:

For those who have been awaiting  the eventual launch of this service, best bet for the duration is to try and identify,  collate and where possible control your various pension benefit pots yourself ( as far as you are able! ).

If Dashboards are live to the public in 2027 with coverage of the vast majority of pension pots, that would be a considerable achievement given the current position. The Dashboard Programme started in April 2019...

The cost of the Pension Dashboard Programme is around £300m. In pension terms, that is not very much, for example, the 2015 Remedy following the botched change of pension schemes in 2015 is costing public service pension schemes (ie taxpayers) around £17,000m in additional liabilities, plus increased administration costs to deal with Remedy and additional costs for decades to come.

However, there are also considerable additional costs for every pension provider to comply with the regulations.

It is rather frustrating though that those who do not need the Dashboard are forced to pay for it nonetheless, both through higher taxes and also due to higher pension charges.

Pensions get hit with so much regulatory burden around protecting members from themselves, whereas ISAs have no barriers to taking everything out and using it however you wish - even LISAs, which benefit from tax relief in much the same way as pensions.

ProDave said:

This looks to me like a solution looking for a problem to fix.  

I've always found the ambitions of the Dashboard interesting. If it was simply a modern pension-finding tool then it may have a role - although modern member tracing services used by providers are very good at finding anyone as long as they are in the UK, and the cost per trace is low in most cases.

But the ambitions are much wider than that, with a desire being to enable individuals to see their 'Estimated Retirement Income' aggregated across all their different types of pensions. Forecasts are assumption-ridden, and one of the key benefits of either advisors or individuals doing this sort of forecast themselves is that they understand the assumptions, sensitivities, risks and so forth involved in pension planning whereas this is all captured in a single aggregated figure regardless of whether it comes from DB, DC or State.

Then there is the aim to enable the Dashboard as a tool for financial advisors, able to access all pension details with the permission of their client. Again, this is a cost borne by all that will not benefit those with no need of this service as they are capable of managing their own finances. Although additions such as this should be a trivial cost in the overall design.

Beddie said:

A pensions dashboard is a very difficult project to deliver. Look at all the issues people mention on here with companies and their legacy systems. So I am going to cut them some slack, as they are up against an almost impossible task.

They should perhaps try to get 90% of providers on board, by volume, and accept that the last bit will be too difficult. That should help with younger people who move jobs more often, but will usually be with the main providers of pensions to businesses now.

It would be a lot simpler if the Dashboard had been designed and rolled out alongside the start of automatic enrolment, covering the Mastertrusts and large Group Personal Pension providers forming the bulk of the AE market, then bringing in public service pension schemes, and then any new pension schemes established after a particular date, as well as any other scheme that wished to participate.

Bringing them in as a bolt-on to established systems is going to be difficult.

It is quite disturbing to see pot follows member resuscitated at Budget 2023. It is not a bad approach as such, I think New Zealand uses it successfully, but is again something you build into the fundamental design of automatic enrolment, not bring in later. It rather illustrates a lack of planning and focus on the desired eventual end-state of workplace pension policy and delivery.

Marcon

hugheskevi said:

It is quite disturbing to see pot follows member resuscitated at Budget 2023. It is not a bad approach as such, I think New Zealand uses it successfully, but is again something you build into the fundamental design of automatic enrolment, not bring in later. It rather illustrates a lack of planning and focus on the desired eventual end-state of workplace pension policy and delivery.

Even then it would be a huge issue. Members could nominate schemes which were either net pay or relief at source -  can you imagine the payroll chaos, not least because the terminology is so hopelessly confusing?

Add in options such as salary sacrifice, the definition of earnings used by each employer, possible contribution matching...the comms alone would be a nightmare.

Quoting @dunstonh from a previous thread: '10,000 employees using individual plans would be 10,000 direct debits.Most individual pensions require an employer payer form every time the amount changes.  So, that is 10,000 forms per month.   Employers cannot input details into individual schemes.  So, they would need each employee to input it.  Individual schemes are not auto-enrolment compliant.  So, each employee would need to opt out.  However, many individual pension providers will not accept opt-outs without an adviser taking liability.'

TheBanker

Beddie said:

A pensions dashboard is a very difficult project to deliver. Look at all the issues people mention on here with companies and their legacy systems. So I am going to cut them some slack, as they are up against an almost impossible task.

They should perhaps try to get 90% of providers on board, by volume, and accept that the last bit will be too difficult. That should help with younger people who move jobs more often, but will usually be with the main providers of pensions to businesses now.

You could draw parallels with Open Banking. Lots of different banks, with lots of different legacy systems, but all of the main banks managed to implement it (as they were mandated to by regulators), and a lot of the smaller/newer banks have been able to join.

Is a Pensions Dashboard really much more complex? It doesn't even need to be real time, given the nature of pensions. 

Sarahspangles

TheBanker said:

You could draw parallels with Open Banking. Lots of different banks, with lots of different legacy systems, but all of the main banks managed to implement it (as they were mandated to by regulators), and a lot of the smaller/newer banks have been able to join.

From the point of view of the user interface the outcome is similar, if information provided by the user system is matched by the provider system then data is presented.

The underlying datasets are the issue, with the Pension Dashboard the user system will typically present incomplete information so more data fields will be required to improve the chance of a confident match. From an IT project management perspective I can see multiple ways for the system and data architecture to be set up, all with different data governance and security issues.

Marcon

TheBanker said:

Beddie said:

A pensions dashboard is a very difficult project to deliver. Look at all the issues people mention on here with companies and their legacy systems. So I am going to cut them some slack, as they are up against an almost impossible task.

They should perhaps try to get 90% of providers on board, by volume, and accept that the last bit will be too difficult. That should help with younger people who move jobs more often, but will usually be with the main providers of pensions to businesses now.

You could draw parallels with Open Banking. Lots of different banks, with lots of different legacy systems, but all of the main banks managed to implement it (as they were mandated to by regulators), and a lot of the smaller/newer banks have been able to join.

Is a Pensions Dashboard really much more complex? It doesn't even need to be real time, given the nature of pensions. 

Vastly more complex!

robatwork

TheBanker said:

Beddie said:

A pensions dashboard is a very difficult project to deliver. Look at all the issues people mention on here with companies and their legacy systems. So I am going to cut them some slack, as they are up against an almost impossible task.

They should perhaps try to get 90% of providers on board, by volume, and accept that the last bit will be too difficult. That should help with younger people who move jobs more often, but will usually be with the main providers of pensions to businesses now.

You could draw parallels with Open Banking. Lots of different banks, with lots of different legacy systems, but all of the main banks managed to implement it (as they were mandated to by regulators), and a lot of the smaller/newer banks have been able to join.

Is a Pensions Dashboard really much more complex? It doesn't even need to be real time, given the nature of pensions. 

I had a very similar thought. Never felt the need to use open banking - probably 25% a "not entirely trusting the process/particular bank" and 75% "just don't see the need".  If it was a centralised established system (which is how I envisage the pensions dashboard would be) rather than each bank having the dashboard I may feel differently.  

Nobody I know uses open banking in any case - I wonder how many people on this thread make use of it.

hyubh

Points from a slightly different view to that predominating on this thread...

• The Pensions Dashboards model is not one of a central store of data. So the concern of @LHW99 is much less general than it may appear. Instead, the model is for public-facing dashboards to send matching data to an outsourced central identity service, which verifies the information received then sends it on to all registered schemes (or more exactly, all registered scheme data providers). The scheme provider gives a response (positive match; possible match; no match), and actual pension data is only requested when a positive match is communicated.
• The amount of central government effort and expenditure so far shouldn't be exaggerated, so I possibly disagree with @poseidon1. While I certainly don't wish to defend central government procurement practices in general, some amount of central direction is inevitable for the pensions dashboards project, and it's really unclear to me how the PDP that actually existed could have been smaller. It was/is all about setting standards, not actually implementing anything.
• Competent people from the pensions industry were hired for PDP on fixed term contracts, but those contracts ended before the project was completed. The 'reset' was (in my view!) rather suspiciously aligned with a government coming to the end of its term and so not wanting to have a 'government IT disaster' in the papers (which, with a longer view, it might have been OK to ride out).
• @Marcron asserts that Dashboard 'not even a system which is ever going to work properly'. That's just hyperbole. Data standards have been published, and while they have been periodically fiddled with, they are what actually-existing software providers have been working towards. Perfection in the data standards isn't a necessary condition for them to be workable, just them being good enough and centrally defined. 
• @hugheskevi over-eggs the ERI issue, in my view. For DC, we're just looking at SMPI figures, and for DB actives, we're just looking at statutory ABS figures. This leaves DB deferreds. For them, statutory ABSs have been talked about for at least twenty years without being introduced. However some DB schemes do deferred ABSs anyway (either of their own accord, or LGPS funds, because of a scheme-specific statutory requirement), and for those that don't, software providers will be very keen to sell the service. Minor point: while I would agree that statutory annual DBSs before Dashboard would make more logical sense, being out of order makes the Dashboard ISP market more competitive in principle.
• @hugheskevi also suggests that schemes only created for AE would inherently be in a better place for Dashboard than old closed DB schemes. This I think belied by the awful data issues of (e.g.) NEST (only some of which has been made fully public). Conversely some one old closed DB schemes administered on ancient software (hello, Profund Classic!) are actually in an OK position because the software provider will have a clean-room solution for Dashboard.

hugheskevi

• for DB actives, we're just looking at statutory ABS figures. 

It is not just statutory ABS figures for actives, as they do not include forecasts of future accrual that are required for ERI so new processes have to be developed. As they are forecast under prescribed Dashboard assumptions around future earnings (necessary for consistency), these are going to be inconsistent with other things such as scheme-specific benefit modellers, 2015 Remedy modellers and such-like. Depending on the default settings of the other modellers, the differences compounded over many years may be significant and confusing to members.

There are also issues for schemes planning to use data from last ABS where an ABS does not yet exist, as data has to be supplied no more than 3 months after joining when a new member seeks view data, and an ABS may well not exist for such members. That is a wider issue than just ERI of course.

LHW99

Instead, the model is for public-facing dashboards to send matching data to an outsourced central identity service, which verifies the information received then sends it on to all registered schemes (or more exactly, all registered scheme data providers).

Capita? ......... Sopra Steria? ........... Experian?

https://www.bbc.co.uk/news/technology-65746518

https://www.express.co.uk/finance/city/1898079/China-attacks-mod-supplier

https://www.techradar.com/news/identity-thieves-crack-major-experian-security-flaw-access-customer-credit-reports

westv

robatwork said:

TheBanker said:

Beddie said:

A pensions dashboard is a very difficult project to deliver. Look at all the issues people mention on here with companies and their legacy systems. So I am going to cut them some slack, as they are up against an almost impossible task.

They should perhaps try to get 90% of providers on board, by volume, and accept that the last bit will be too difficult. That should help with younger people who move jobs more often, but will usually be with the main providers of pensions to businesses now.

You could draw parallels with Open Banking. Lots of different banks, with lots of different legacy systems, but all of the main banks managed to implement it (as they were mandated to by regulators), and a lot of the smaller/newer banks have been able to join.

Is a Pensions Dashboard really much more complex? It doesn't even need to be real time, given the nature of pensions. 

I had a very similar thought. Never felt the need to use open banking - probably 25% a "not entirely trusting the process/particular bank" and 75% "just don't see the need".  If it was a centralised established system (which is how I envisage the pensions dashboard would be) rather than each bank having the dashboard I may feel differently.  

Nobody I know uses open banking in any case - I wonder how many people on this thread make use of it.

I use one click to download transactions from my bank to my finance app. I'm not sure if that's OB or not but, if it is, them I use it.