Wednesday, 21 May 2014

Fixed costs in small countries, again

The Productivity Commission explains New Zealand's high prices:
The overall messages from this analysis are:
  • Modelling the determinants of non-tradables prices using the extended FG model works fairly well in general, and especially for a group of OECD-Eurostat countries. 
  • This supports the hypotheses that: 
    • Non-tradables prices are higher in association with higher values of capital and unskilled labour endowments, and trade deficits; 
    • Non-tradables prices are lower in association with higher values of skilled labour, and population;  
  • Tradable prices also tend to be higher where non-tradables prices are higher both because this raises the (non-tradable) input costs for tradables, and because high tradables prices are impacted by „trade impediments‟ and indirect taxes. 
  • The model suggests NZ‟s relatively small population and high tradables prices tend to raise its non-tradables prices relative to other countries. NZ‟s lower than average capital endowments should serve to counter-act these high prices, but modestly. (NZ has the 28th highest ranked capital endowment, and 25th highest capital/labour ratio, out of 43 countries). 
  • For labour endowments, taking a ratio of skilled labour (differences) to population (differences), NZ is ranked 29th out of the 43 OECD-Eurostat countries in this ratio, so that its relatively low skilled labour per capita also serves to raise its service prices.
  • NZ's relatively high price of non-tradables is estimated to add around 40% to the domestic consumer price of tradables, compared to border or factory gate prices. This is also around the OECD-Eurostat average cost share. 
  • However, after accounting for this 'cost share of non-tradables' contribution, NZ's tradables prices remain fairly high by international standards. A "good candidate" to explain this are the transaction costs associated with NZ‟s distance from markets. Though we have not examined direct evidence on this here, numerous other studies have suggested the importance of such factors; see, for example, McCann (2003, 2009). 
  • Based on "adjusted" tradables prices that removes the "cost share of non-tradables" element, NZ's tradables prices are around 6th highest in the 43 country OECD-Eurostat sample – behind such countries as Iceland, Norway and Japan (see Figure 6). These are also countries relatively distant from many of their key markets. (For Japan at least, other protectionist measures may also be relevant). However, Australia is ranked 19th out of 43 countries in its adjusted tradables price, suggesting that to the extent that there are "disadvantages of distance", Australia manages partially to avoid or overcome these.
  • As Figure 6 shows, NZ is like a number of other small countries with a high adjusted tradables prices (e.g. Cyprus, Malta, Denmark, Finland, Israel, in addition to the "small and distant" economies of Iceland and Norway). This suggests that size or other characteristics of domestic markets/populations may be important in ways not already accounted for by the FG model's variables.
Labour has been exercised lately by the high costs of an increasing population: new migrants come in and bid up the price of housing where housing supply cannot expand to meet increased demand. They might worry instead about the longer-term costs of being small if they block migration. We can't do much about distance. But we can get bigger.

Other points worth thinking about:
  • Our relatively high cost of alcohol and tobacco is particularly noted by the Commission; this does run contrary to the usual story we're told about alcohol being too cheap in New Zealand.
  • I wonder whether results would be affected much by taking PriceSpy import prices instead of domestic landed prices for some consumer goods. How much does PriceSpy, and related services, constrain domestic prices?
  • Gemmell's prior work showed footwear and clothing to be expensive here. I note that footwear still here attracts a 10% tariff. We could fix part of our stupid-high-prices problem by getting rid of the tariff that helps make one of those price categories more expensive than it needs to be. Sir Roger tried back in 2010; National vetoed the change.


  1. I rather expect that small European countries, and Singapore, do very well by being very close to larger markets. The effective size is larger than the population.

  2. Note Aaron Schiff's amazing story of why NZ has no IKEA

    IKEA alone would move the needle on prices of durable goods, but imagine how many other stories we don't know about.

  3. Singapore isn't in the OECD-Eurostat sub-sample. But in any case, there is also no link between price level and popultion across the entire ICP sample (with lots of small Caribbean or African countries).
    One reason to be cautious about the results is that this is a snapshot, and we know that the passthrough of exchange rate changes to consumer prices is quite slow these days (not just in NZ) and there is lots of pricing to market. If the data were available (which I gather they aren't) it would be fascinating to see how NZ compared averaging a period such as 20000/01 when the exchange rate was very low with the 2005 results (when the exchange rate had recently moved sharply higher, to well above average levels).

  4. Indeed. I'd hoped that we might have gotten an Ikea for Christchurch post-quake. Would have made a lot of sense, what with a whole city needing affordable flat-pack furniture.

  5. Yup. I'd also be really keen to see the differences between "products that face PriceSpy competition" and those that don't.