An odd approach to tax policy

Pattrick Smellie has had about the best summary on the messes in the last round of proposed tax changes.

Earlier this week, the government put up a pile of tweaks to tax policy. This sort of thing is usually pretty bland. But included in the mix was a change to the GST treatment of management fees on Kiwisaver funds. 

Pattrick writes:

A regulatory impact statement, which it’s not clear if most ministers had read or been briefed on before waving this proposal through at a cabinet meeting, warned that maybe $103 billion of KiwiSaver savings would not occur as a result of the changes. 

Stuff’s Rob Stock broke the story yesterday afternoon and all hell broke loose on a policy whose introduction broke every rule in the political book. 

Quite apart from apparently attacking the KiwiSaver scheme's basic premise – to build funds for the future – it was introduced without so much as a press statement. A properly prepared government would at least have tried to control how such a political hand grenade was first reported. 

No such effort was made. 

Cynics will think this was the government sneaking it in and hoping no one would notice. 

Well, maybe, but look what actually happened. 

It is equally likely that somehow, amazingly, this proposal didn’t cause even a momentary flicker on anyone’s political radar.

It's a really strange one because the change looks like it could have been defensible. 

The most recent Tax Working Group laid out the usual case for exempting financial services from GST. There can be huge valuation problems in sorting out what's the value of the service provided and what's the value of the underlying traded thing when it all gets bundled together.

I'm hardly an expert on this stuff, but I know that motivated people sometimes point to the exemption for financial services when they try arguing for exemptions on other stuff. So I'd looked at it at that point, saw that there was darned good "It's just too impracticably difficult to levy GST here so we're not going to" reasons for it, and concluded that the "let's exempt meritorious things" people were being disingenuous in pointing to the financial services exemption. 

So when news broke on a plan to set GST on financial services here, it seemed odd. Hadn't IRD already concluded this stuff is just too hard? I checked back in the papers from the TWG Secretariat and they hadn't foreshadowed any areas that could be pulled from the actually-too-hard basket. 

But those difficulties shouldn't apply when it comes to explicit fees levied in financial services. Whether it leads to other distortions and new structures to try to turn fees into margins, that's beyond me. It's at least possible and possibly even probable that the proposed change had made sense on a basic-tax-principles assessment. 

It seems the kind of thing where, if the RIA on it has, at paragraph 49, $103 billion in reduced Kiwisaver fund balances by 2070, they might have considered better preparation on this one. It's plausibly defensible, but got killed within a day through bad comms. 

The same bundle brought GST changes for platform service providers like Uber, and this one could turn into a right mess. It's very much a what-sucks-least problem, and all options are going to suck. I'm just not convinced they've picked an option that sucks least.

Recall that there's a de minimus regime around GST where if an outfit has less than $60k in revenue it isn't required to file for GST because the time and hassle for everyone involved is greater than the amount of tax that might be collected. Makes sense, right?

So what then happens if a tech innovation means a lot of new small part-time entrants are able to enter a sector, many of which will be under the de minimus threshold because it's a part-time deal for them? IRD might start worrying about base erosion and about distortions favouring the small-time operators. 

But how can you do anything about it that doesn't make things worse? All the options are bad too. 

IRD canvasses some in the RIA, and came up with the following. 

For driver-partners who work with platforms like Uber and who are GST registered, no changes. They charge GST, claim back expenses; Uber pays them a GST-inclusive price from riders and claims back the GST that they've charged. All fine. 

But for drivers below that threshold, the platform will charge GST on the ride. 

Now that's a problem because the drivers will have paid GST on their fuel, oil, maintenance, vehicle - all the inputs where GST would normally be claimed back. Remember that part of why the de minimus threshold works is because minimus is smaller than you might have thought. At the same time as they're not charging GST, they're also not claiming back GST on expenses. So the net is a lot smaller than you might have figured. 

If Uber, or Lyft, or whoever, is charging GST on the full cost of a ride, and the suppliers aren't claiming GST on inputs, then you've double-charged GST. And that's a big problem. 

IRD proposes a workaround. The platform would collect 15% GST on the full cost of the ride. It would submit 6.5% up to IRD and send 8.5% back to driver-partners as a deemed input cost proportion. 

I don't know where the split came from - whether it's some overall average of how this stuff nets out, or one specific to transport, or something else. 

But it will wind up requiring the platforms to implement a pile of new accounting to track things, which might need runway to sort out.

Suppose I drove for a few ride-share companies. Would I be able to set up one GST-registered company where all my driving for one platform gets accounted, have it take more than its fair share of the costs of fuel, maintenance, and everything else, and have my driving for the other platforms come under the deemed cost regime? Possibly isn't worth the hassle to set up, but IRD might need to watch that driver-operators aren't trying it on. 

GST normally avoids this kind of problem; folks claim back GST on expenses while paying GST on sales. But a deemed-expenses kind of set-up would break that. 

Maybe the thing is defensible if it really is less bad than potential base erosion and distortions where platforms change industry structure, but I'd hope that they'd talked with the platforms about practicabilities and implementation. It sounds like it could be tricky.