Is it time for a proper root and branch independent review of our official statistics provider, including what it does, what it provides, how it provides it and the timeliness and accuracy of that information?

Either it wasn't great timing, or it was actually very good timing.

At a stage in its life when it is gradually getting more heavily questioned on the quality of data it is producing, Statistics NZ's admission of what I think was a truly calamitous error relating to food prices saw our official statistics provider turning the spotlight brightly on itself.

(The nub of the food price issue appears to have been that "due to human error" about 13 days worth of January data from one of the supermarket groups just simply wasn't included. Incredible.) 

An official statistics provider is a very easy target. And for this reason (well, I would suspect it's the reason), economists and the like don't tend to publicly question, raise doubts, or criticise Stats NZ and the information it provides.

I can imagine that providing statistical information could easily feel like a thankless task. Produce spot-on data that nails the subject matter and people say, yeah fine. But produce something that looks like it has glitches and you become open to criticism.

So, to openly criticise Stats NZ is to seem, well, mean. Let's face it we are talking about people with feelings and professional pride. How do you feel if someone criticises your work?

All I can say though is, that as someone who has used the data provided by Stats NZ on a very regular basis, I have become increasingly frustrated and, yes, doubtful, about what is being produced.

One area of Stats NZ's work that has received a reasonable amount of discussion is its GDP figures. This is important stuff. Any number of people and organisations need to know what the economy is doing. 

A key user of the information is the Reserve Bank, since it needs to know what its monetary policy - its Official Cash Rate settings - are doing to the performance of our economy. 

Since a very bad mess-up in May 2024 when the RBNZ misread the tealeaves completely and gave a super 'hawkish' statement - at about the time our economy was actually starting to fall of the cliff - it's been noticeable that there's been much more dependence by the central bank on 'high frequency' data, which includes such things at the BNZ-Business NZ Performance of Manufacturing Index and and Performance of Services Index, among other things.

The outcome has been that the RBNZ hasn't messed up again, but it's not a satisfactory situation. Our central bank SHOULD be able to trust the official data sources.

Vital decision making processes depend on this

I use this example to stress that the information produced by Stats NZ is not just 'nice to have'. In many instances it is information that people are relying on to make big decisions. If the information is not good, decisions made might not be good - and that could have adverse consequences for us.

In May 2024 the RBNZ was talking about the possibility of interest rate hikes - when it should very much have been thinking about cuts. The subsequent huge U-turn in commentary by the bank in July, followed by the first OCR cut in August 2024 were based on what the central bank was gleaning from high frequency data - not from what the GDP data was saying, which at that point was missing the very sharp recession we had fallen into.

That's one negative example.

Here's a (generally) positive one. The Selected Price Indexes (SPI) monthly publication began in 2023 in response to a push, particularly from the RBNZ, for more timely inflation information - given that we as a country are an  outlier in not having a monthly inflation read. To have just four definitive snapshots of inflation in a year is nowhere near adequate.

Anyway, the SPI started in 2023 and, to be honest, it all looked a bit rushed and the initial iteration was short on commentary and was just something of a data dump. BUT from the get-go there's no doubt this has been a very useful addition to the information provided by Stats NZ. And it seems to have been done by our data provider without any obvious additional resources made available. And that's probably important to bear in mind, given this week's events.

It seems to me that economists' forecasts of the future level of the Consumers Price Index, the official measure of inflation, have been a lot nearer the mark since the SPI was introduced. The economists like the SPI (which covers about 47% of the CPI's ingredients), because it covers some of the more volatile ingredients. Consider the fact that domestic air fares rocked 12.8% in the month of February. Economists can now factor that into their understanding of where the CPI may sit for the March quarter.

The Government of course tipped some more money in for Stats NZ in the Budget last year in order for the development of a monthly CPI measure. That's going to start next year and will be welcome.

An independent stock take

But I didn't really like the ad hoc way in which that was done. "Here's a fiver, go get us a monthly CPI." And indeed, I would certainly say the way in which the SPI was introduced was pretty ad hoc as well. And did that in any way contribute to the awful error Stats NZ admitted to this week?

I think it would be helpful, to say the least, to have an independent stock take of what it is Stats NZ produces, why it produces it, who it is for and whether it is fit for purpose. 

The Government of course would have to play a key role. And obviously there's plenty of room for debate about which work Stats NZ does is particularly 'important'. I'm biased - I think the economic stuff is crucial. But others would argue strongly about the more socially focused work.

I think one vital area of focus for any inquiry should be to examine data that others - such as the RBNZ - lean on for making key decisions that have a key impact on NZers.

Right at the moment we definitely have a situation where the likes of the RBNZ are looking elsewhere for guidance. And if we end up in that situation, why spend around $250 million a year for a statistics provider at all? And no, I'm not suggesting getting rid of it. The question is rhetorical. We need to get better bang for the buck and, as far as I'm concerned, if one way of doing that is actually giving the stats provider MORE bucks, then so be it.

I think it is that important.

We need good official information. I think our official information has gone backwards in terms of quality and reliability in recent years. This trend needs to be arrested. An all-embracing independent assessment of what Stats NZ is, what it does, how it does it, and what it should be doing is needed. Sooner rather than later.