How big is the hole in Key’s “$17b hole”?

The “show me the money” costings war is a credibility test for both John Key and Phil Goff

Phil Goff says he will “show John the money” today, providing Labour’s costings for its various policies. He says he will borrow about $2.6b more than National over the first three years. Key’s costings put that figure at $11.7b, increasing to $17.2b over one additional year.

Yesterday, commentators were saying that Goff’s reputation was on the line with these forthcoming costings. If they didn’t stack up, he would not be credible. But what about National’s credibility with its costings? Isn't National’s credibility on the line as well? It is making very public estimates about what some policies will cost, too.

Having looked at National’s spreadsheet (available here on its specialist Attack Labour website), I think there are some serious problems with it. Following up on John Pagani’s insightful posts, I found $8b worth of clear errors in National’s spreadsheet. And I’m sure I didn't find everything there is to find. These are errors that have the effect of roughly doubling the impact of Labour’s manifesto on net debt. That isn’t much good for National’s credibility.

Contributions to the NZ Superannuation Fund.

Labour wants to resume contributions. National doesn’t. Labour’s added contributions total $6.1b over four years, an amount National says is new net debt. Except Labour isn't giving the money away or burning it or buying Drachma, it is putting it in a high-quality investment account that the government owns. The money increases a government financial asset, which means it cannot also be new net government debt. If you borrow $10,000 and then put it in the bank, have you increased your net debt by $10,000? Of course not. But National says you have. This is crazy, and a glaringly obvious error.

Running total of errors: $6.1b

GST removed from fresh fruit and vegetables

National says this policy will cost $1.1b over four years, topping at $350m in year four. That level is based on a guesstimate that we will spend $2.7b a year on fresh fruit and vegetables that year. But we won’t. Stats NZ figures show that NZ households only spent $1.7b a year on all fruits and vegetables in 2010, including fresh ones, dried ones, canned ones, and frozen ones. The public figures don't break it down any further, but generously assuming that 50% of that is fresh, then the annual bill from removing GST tops out at only $110m.

Then you have to add in fresh fruit and veg bought by restaurants and turned into other stuff. Last year restaurants in New Zealand had GST exclusive sales of $4b (Stats NZ figures). Food cost in restaurants are usually about 30% of sales. If households are anything to go by, fruit and veg makes up about 20% of food costs. We’re assuming fresh is about half of that. And the GST 15% of that. Which is another $19m a year. So the total cost tops out at $130m a year, not $350m. Applying that correction across the other years gives a four-year cost of $400m, not $1.1b

Running total of errors: $6.8b

Labour’s tax reforms

National’s figures completely exclude Labour’s policy to close tax loopholes. Apparently, closing tax lopholes is completely impossible. Except, of course, when National closes them. Labour says it will gain up to $300m a year from these. Let’s assume they’re dreaming, and they actually only get half of that. $600m over four years.

Running total of errors: $7.4b

Finance costs

$1.1b of the $17b hole National is claiming comes from extra finance costs (i.e. interest) on all the additional borrowed money from Labour’s other debt-incurring projects. Of course, if that new debt is fictional, then so are the finance costs. Given the errors I listed above, that saves a further $525m.

Running total of errors: $7.9b

And these are only the obvious errors I can easily quantify online. There are plenty of other, ahem, opaque guesses in National’s figures. For example, National says that stopping asset sales will increase net government debt by $2.9b over four years. How did they get that number? No idea. It seems to include cash received from the sale and interest saved on the retired debt and lost dividends, all rolled together into one big messy meta-guess.

Similarly, for some reason the impact of the R&D tax credit has been rolled up with the impact of bringing forward agriculture’s entry to the ETS, even though they have nothing to do with each other. That makes assessing the accuracy of each individual number impossible.

Once Goff releases Labour’s costings, there will no doubt be a weekend full of dorky people throwing spreadsheets at each other and bemused reporters who took journalism instead of maths trying to play umpire. The most important thing to remember is that this is a credibility test for both Key and Goff, not for Goff alone.