Alex Tarrant asks whether IRD could even implement changes if the govt embarked on widespread tax reform

By Alex Tarrant

How about this for a conspiracy theory: The government actually wants to implement widespread tax reform with a total shake-up of how New Zealanders are taxed, but the Inland Revenue Department is reluctant because its systems can't handle it.

What if the Finance Minister took his grand plan to the Inland Revenue Department? They'd likely shake their heads and say, "sorry Bill, no can do. We'd love to help, but seeing as we're still using early 1990s technology, which is at breaking point by the way, you're out of luck. Come back in four years, and if we get around to upgrade our systems (and you've given us the money for it), we'll see what we can do then."

What's more, and just for argument's sake, let's say Paula Bennett is really focussed on reforming the welfare payments system through a big shake-up of Working for Families so that it actually targets the poorest and most in-need of New Zealand's children. She's come up with a fantastic, world-leading system for New Zealand welfare to replace WFF. However, because something like this has never been done before, it's extremely complicated - much more than WFF. But Bennett thinks it's much fairer than the status quo, so it has to be implemented.

But again, our IRD boffins shake their heads. They even laugh at the Minister this time, because it's just so not-doable right now it's painfully funny to them. "Working for Families was hard enough," they say. "Give us four years, and a few hundred million dollars, and maybe we can sort something out."

That'd be pretty nuts, right? A first world country not able to implement new policy because it still uses computer systems from when John Key began his money-trading career and when Bill English was a young, bright-eyed whippit about Bolger's Cabinet table learning from Ruth Richardson how not to be a Finance Minister.

That step change in the economy we've all been talking about, that rebalancing that we so desparately and quickly need - yeah we can do it, but it won't be quick because it can't be quick - that's why we've just been tweaking and fiddling. Great conspiracy.

But...

In its 2008 briefing to incoming (and former) Revenue Minister Peter Dunne, the IRD expressed concerns about its ability to deal with policy changes "quickly and simply". You see, the system it uses (FIRST - Future Inland Revenue System Technology) was introduced in 1992 and designed to process tax requirements, and that only:

Inland Revenue’s wide portfolio of functions has increased the diversity and complexity of interactions with the public and resulted in increasingly complex technology systems. The department’s core tax processing system (FIRST) was designed for processing and administering tax requirements; it was not intended to fit social policies which have subsequently been overlaid.  In simple language, successive “work-arounds” and “boltons” have been added to FIRST.  That is an expensive way to do business and it limits our ability to make changes quickly and simply.

Hilariously, three years earlier in it's 2005 briefing to the incoming Minister, the IRD expressed concern FIRST had reached a tipping point:

At that time [when FIRST was introduced in 1992], the department’s role was purely tax administration and FIRST was designed accordingly.  Since then the department’s responsibilities have been significantly expanded to include many social policy activities including family assistance, child support, and student loans.

The technology support for these activities has been integrated into FIRST, creating a more complex system than was originally designed. This complexity continues to increase as our systems are modified to incorporate policy changes, resulting in both increased cost and time to implement further changes.

Over the past 12 months there have been indications that our systems are reaching a “tipping point”, where consideration will need to be given to replace systems rather than make further changes to existing systems. This is a particular concern for the administration of the student loans system.

So surely something's been done?

Er, well, not quite. In fact, included in Treasury's dump of Budget 2011 advice to Ministers yesterday were a few little gems from the heavily censured Vote Revenue four-year Budget plan for 2011. Gems like this:

Without future investment in our  transformation programme Inland Revenue will find  it challenging to deliver further process and system technology efficiencies to deliver future policy changes effectively. Opportunities to improve customer compliance behaviours and reduce customers’ compliance burdens will be compromised and pose the risk that current system failures will increase with time.

And this:

In the early 1990s the Crown  made a significant investment in Vote Revenue to implement the core components of the tax system (FIRST). While Inland Revenue has been able  to leverage this investment to incorporate new functionality, such as eServices and changes in Government policy, there has only been limited investment in the core system and additional internal funding has been required to meet growing customer contact levels  and increased demand for service delivery activities such as Working for Families.

The IRD is basically saying they've been doing it tough, and please don't introduce anything too complicated, because we can't handle it.

So much for being able to change our step quickly.

Now the 'four year' figure I refered to above wasn't just picked out of thin air. Indeed, a senior government figure made reference of it to me a few weeks ago. At the same time as other OECD countries are spending millions of dollars upgrading their taxation systems, we've been applying band-aids to a system introduced when the same time the Maastricht Treaty was signed.

So I thought I'd ask Peter Dunne what was going on. This was his response:

Inland Revenue's FIRST system is ageing technology and solutions are being worked on to ensure that the department is in the future able to deliver on all its existing and future programmes.

There is nothing to suggest that Inland Revenue cannot deliver on its current work programme intiatives. The reason for the recent focus by Inland Revenue on seeking to update its technology is to allow the department to respond to future change and/or the need to adapt in a changing environment.

More IT band aids

Recent focus. What recent focus? The only thing I can think of is when the IRD announced it wasn't going ahead with a new student loan processing system, despite having spent NZ$21 million of the NZ$35 million it was given for the 'Oracle' project. After spending NZ$21 million figuring out it didn't want a new system, IRD applied a band-aid to FIRST instead.

Guess what the IRD's excuse for not implementing the new Oracle system was? It was too complicated to do, given the government wanted to change student-loans legislation quickly. IRD found it was more  difficult and time consuming than expected to extract information from FIRST and put into the new Oracle system.

So even though they were given money to upgrade to a new system, all we got was a band-aid on the old one. Here's some quotes from the IRD earlier this year explaining the situation (from here and here):

“It would not have been possible to build and populate a new system in time for important changes to student loans expected to come into effect next year.”

This single sentence shows the tension between government wanting to introduce new policy, and the IRD's ability to allow the policy to be introduced. What happens when student loans legislation changes again? No new system to handle it, and the current one is at breaking point. Great position to be in.

“While the new student loans system was intended to be the first component of a new computer system for the department, the complexity of student loans made this unrealistic given the legislative time frames.”

“Unlike a normal bank loan, for example, repayments depend on factors including the loan-holder’s income, whether he or she is resident in New Zealand, and the need to apply these rules retrospectively, in some cases back to 1992. All of these factors are programmed into FIRST, and proved far more difficult to extract and migrate than expected.”

So what was to be the starting block for a new computer system for the IRD was disbanded after NZ$21 million was spent on nothing. Also great to learn that it's really hard to extract information from the current system.

Doesn't give you much confidence for when some government really wants to make some substantial tax or welfare payment changes does it?