Monday, 10 March 2014

Purchasing power parity? I don't think so

The Commerce Commission, as I blogged a little while ago, has come out with its latest benchmarking report on retail mobile telecommunication services. Produced for it by Teligen, it uses OECD-sourced power parity (PPP) exchange rates to convert overseas currency prices into NZ dollar equivalents, because, as the report says (p7), "The use of PPP rates was preferable to using spot exchange rates as PPP rates are a more stable measure".

I've got no problem with that - quite the reverse. It's the right thing to do, and it's absolutely standard operating procedure for international benchmarking reports like this.

But I do have an issue with one of the OECD's calculations. I don't think its PPP estimate of the NZ$/A$ exchange rate is right.

Before I go any further, what's your estimate, however rough and ready, of the NZ$/A$ PPP rate? If, for example, you were to go to Australia for a week's holiday, would prices in Oz seem about the same as at home if the Kiwi/Aussie rate was 70 Aussie cents? 80? 90? Parity? Above parity? What was your instinct?

My own guess is that things in Australia start looking on the cheap side when the Kiwi is somewhere into the higher 80s, and are definitely looking cheap over there when the Kiwi is above 90 cents. And if you told me I could swap my Kiwi dollars for Aussie dollars one for one, I'd be in Sydney every other weekend. Or to put it another way, my estimate of the PPP rate is around the 80-85 Aussie cent mark. Maybe I should go all Bayesian and call it my "prior".

Right. Now let's have a look at the PPP A$/NZ$ rate that Teligen used. Or must have used, since the link they gave in the report to the OECD data ( is broken and I've had to look for the OECD PPP data some other way. I found them here.

Here's a summary of the OECD PPP calculations for the Aussie and Kiwi dollars. The first two rows show the PPP exchange rate in terms of local units per US$, which is how the OECD likes to write them, whereas we like to use US cents per local dollar, which you can see in the next two rows. And the final row is the OECD's estimate of the A$/NZ$ PPP rate, again expressed the way we usually do.

On the OECD's calculations, the Kiwi dollar's PPP exchange rate is - wait for it - approximately, parity with the Aussie dollar.

Frankly, I don't believe a word of it. It's way too high.

As a practical matter, if one of your interests in reading the Commerce Commission's reports has been to see how much better or how much worse prices are in Australia, I would put up a large health warning against the bilateral comparisons. The PPP rate is making things in Australia look less expensive than they really are. Something that costs A$100 will supposedly cost NZ$100: no way. At what I'd say is a decidedly more realistic 85 cents PPP, the cost in Australia would actually be NZ$117.65.


  1. There are definitely relative price differences between the countries - which makes comparing bundles a pretty fraught experience! From a tourist side I agree with you, but there supermarkets over there are pretty crazy expensive, and even at parity their alcohol is obscenely expensive!

    I like to use the Aus/NZ PPP as an example of where our priors about exchange rates can be a bit out of whack - so ultimately I'd put this down as an issue that we need to be careful with, and deserves further investigation ;)

  2. Sure. I appreciate that it's a rather subjective take on where PPP might lie, though I'm also aware that people who have tracked a long-term (50 year) series of the actual A$/NZ$ rate have found that it consistently lies below the OECD PPP rate, which is somewhat odd and leads you to think that either there's some wedge that isn't closing, or that the PPP rate is out of whack (or maybe both).

  3. This comment has been removed by the author.

  4. Purchasing Power Parity :- World Economics is an organisation dedicated to producing insight, analysis and data relating to questions of key importance in understanding the world economy.


Hi - sorry about the Captcha step for real people like yourself commenting, it's to baffle the bots