(Update for readers in November 2011: This was written in July, in response to Joyce's claims about Labour's tax policies. It doesn't address subsequent issues such as KiwiSaver and NZ Super Fund changes. You can find that here.)
Labour says their policy will pay off $7.8b of debt by 2025. National says it'll add $15b of debt. That's $22.8b credibility gap.
Allow me to fill it in for you.
We already saw Labour's efforts last week, but over the weekend, David Farrar put this up on Kiwiblog, introduced with "Steven Joyce has put Labour's numbers through the Treasury calculator".
On Twitter, Farrar cited these same numbers as "numbers from Treasury". Let's hope he doesn't say the same thing in his columns or his blog, because that would be greatly misrepresenting these numbers and besmirching Treasury's reputation as neutral, competent public servants.
Joyce got someone to download a publicly available spreadsheet and punch some numbers into it. That's it.
(Those spreadsheets are horrifying abscesses of public finance. You barely want to touch the stuff that oozes out when you poke it, let alone stick your head inside. I did that some time ago. I say "some time" because there's a three week gap in my timesheet that started a month ago, when I started ripping one of these fiscal models apart. I managed to retrace my steps for two out of the three weeks, but the third week simply vanished. My mind may have blocked out the horrors I saw inside the endless LOOKUP() recesses.)
Wrong spreadsheet - $3.1b
Joyce reckons that Labour grossly underestimated the cost of their tax-free threshold, and overestimated the revenue from their new top tax rate. Unfortunately, Joyce used the wrong spreadsheets. In particular, he missed the bit which said:
Larger changes require more complex modelling than is shown here.
But that's just some nerdy ass-covering technical caveat, right? Well, Joyce punched in the numbers and came out with an estimate that costs will grow rapidly, doubling by 2025.
Let's think about how this would work in the real world. Everyone earning over $5000/year would get the benefit of the whole tax-free threshold. That's pretty much everyone in the workforce. So if everyone already gets something, how would more people get it?
The cost of a tax-free threshold only grows when new people enter the workforce.
So unless Joyce thinks there will be 3 million more people in the workforce in the next decade, this is a patently stupid and ridiculous result. Common sense would tell you that it is impossible.
This one goes firmly in Labour's favour.
Religious differences - $3.3b
Last week, I said it was Left dogma to believe that nasty tax dodgers can be squeezed indefinitely, and that more money will keep coming out. I contrasted it with the Right's belief that public servants can be squeezed indefinitely, and more efficiency will keep coming out.
What I didn't count on was that Joyce had another belief as well: We have reached Peak Tax Enforcement. He says that "The Government is already putting significant resources into this area" and decides that the amount of additional money you can get from tax dodgers is $0. Forever.
Labour's position here is just arbitrary and unfounded, but Joyce's position is kinda ridiculous. (Draw.)
(Actual Reason!) Behavioural Changes - $1.6b
Some people will be able to avoid the loss ring-fencing (can't be bothered explaining this) measures, so Joyce reckons it'll only get 50% as much revenue as it could in theory.
This one goes to Joyce.
Climate Change Denial - $4.3b
Joyce uses Nick Smith's estimates, which reckons Labour's ETS changes will net a mere $483m, rather than the $4.8b that Labour expects.
How much it costs to subsidise farmers' carbon footprint (and how much we can save by stopping it earlier) depends on how much carbon costs.
The Parliamentary Commissioner for the Environment commissioned a report on this last year.
If there's little international action on climate change, the report estimated that the price of carbon would reach $35/tonne by 2020 and rise to $50/tonne by 2030. If there is a moderate international commitment to reducing emissions, then the price of carbon would reach $50/tonne by 2020 and $100/tonne by 2030.
Australia has already set its carbon price at $29.55 NZD.
Punchline: Nick Smith's model assumes the carbon price to be $25/tonne and not to rise at all.
Unless climate change is a hoax, this is an unreasonable assumption.
I'm Out of One-Liners - $1.3b
Labour reckons (using "Modelling based on officials") R&D tax credits will cost $2.9b. Joyce uses the 2008 costs with some adjustments. Given that no information is available on Labour's methods, and Joyce's seem quite reasonable, this should go to him.
Whatevs - $1.7b
Joyce reckons the CGT is only worth $14.6b, rather than Labour's $16.2b estimate. Farrar calls this "roughly accurate". I agree – for a big fat unknown like CGT, any estimate is a rough ballpark, and a $1.7b difference between estimates is not worth arguing about.
Double Counting - $7.5b
Then Joyce adds on another $7.5b of interest on everything. That would have been fine, except that most of those things which accrued the interest are bullshit.
--
There's about $4-5b of legitimate gripes here. And it crystalises my view of Labour's tax package: Using CGT to pay off debt is a rock solid policy core, but everything else about the package is haphazard, with a lot of question marks and not much rationale.
At the launch, I asked Goff and then Cunliffe whether their debt payment track was a commitment, or whether it was simply what they would do if they had the money. Both avoided the question, and Cunliffe said he was confident that their estimates were so conservative that they would only have more money, not less.
National's attacks (even though the "$15b new debt" figure is plain ridiculous) raise important questions. I still don't know how committed Labour is to paying off debt; if forced to choose between paying off debt and tax cuts, what would a Labour government choose?
Until they answer that question, they're not a government-in-waiting.