The Treasury

Global Navigation

Personal tools


Monthly Economic Indicators

Special Topic: Tradable Sector in New Zealand

In early 2011, the Treasury published a working paper on the drivers of the exchange rate and how movements in the dollar impact on the tradable sector.[1]  This note takes a more detailed look at the tradable sector measure used in that working paper and elsewhere by the Treasury.

What is the tradable sector?

The tradable sector produces output that can readily be traded internationally (at a reasonable price) so is the part of the economy most exposed to global competition.  As a result, it is more likely to experience stronger productivity growth than the non-tradable sector.  The tradable sector is also more able to contribute to growth in a country's exports (because it includes industries that export a large proportion of their output) or to reduce a country's imports (as it includes import-competing activity).  Given New Zealand is a small and relatively open economy and given recent modest growth in both productivity and exports, there is considerable interest in the performance of the tradable sector.

How do we estimate tradable sector output?

There is, however, no ideal measure to gauge the tradable sector’s performance (unlike consumer prices, which have an official measure of tradable prices).  Over the last three years, the Treasury has used an estimate of tradable sector output building on previous work by the IMF and Reserve Bank of New Zealand and similar to estimates made by the ANZ Bank and Savings Working Group.[2]  Our standard estimate uses a simple approach that tracks changes in tradable sector output based on industries judged most likely to produce tradable output – the primary sector (agriculture, forestry and logging, fishing and mining) and manufacturing.  The volume of value-added production GDP in these industries is then added to the volume of service exports (expenditure GDP) using quarterly seasonally-adjusted data.

Service exports are used to proxy services that are tradable because there are difficulties in isolating tradable service activity.  For example, tourism cuts across industries and education is dominated by domestic students.  It is not recommended to add service exports (a measure of final demand) to value-added industry output, but to reduce this problem we only refer to index numbers and not levels.  Non-tradable output is the residual with total GDP.  Using these measures, the tradable sector has seen output diverge from the non-tradable sector (Figure 1). 

Figure 1 – Tradable and non-tradable output
Figure 1 – Tradable and  non-tradable output.
Source:  Statistics New Zealand, The Treasury calculations

Are there other approaches?

We can test this simple and arbitrary approach using Input-Output tables from Statistics NZ, the last official set being for 1995/96, to choose which industries had exported a large share of their final demand or imported a large share of their inputs.  The difficulty is choosing the share to be used as a cut-off for an industry to count as tradable.  Using 30%, the tradable sector is agriculture, forestry, fishing, mining, five manufacturing industries (food, beverage and tobacco, textile and apparel, wood and paper products, metal products, and petroleum, chemical, plastic and rubber) wholesale trade, accommodation, restaurants and cafes and transport and storage.  These are similar to the goods industries above, although some industries within manufacturing are excluded as they mostly service the domestic market, and we now have three service industries.  They are also the industries typically noted in overseas studies.[3] 

The tradable sector using a 30% or 40% share is similar to our simple measure.  A 20% share brings in printing, publishing and recorded media, machinery and equipment and furniture and other manufacturing, electricity, gas and water, communications and property and business services.  Using this definition, the tradable sector has performed somewhat better than most of the narrower definitions as it includes some faster growing service industries (Figure 2).  The 50% share also performs relatively well as it excludes non-food manufacturers, which have contracted in recent years.  Clearly, the tradable definition using this approach still depends on a choice about how much an industry is exposed to overseas trade.

Figure 2 – Tradable sector sensitivity
Figure 2 – Tradable  sector sensitivity.
Source:  Statistics New Zealand, The Treasury calculations

What has happened to the tradable sector?

Tradable sector output has diverged from the non-tradable sector over the 2000s no matter which measure above is used.  The divergence is less using high and low shares as there is a double impact from shifting a high-growth industry (eg, communications) from non-tradable to tradable or shifting a low-growth industry (eg, textiles) the other way as the tradable sector benefits and the non-tradable sector is dampened (Figure 3).

Figure 3 – Tradable/non-tradable divergence
Figure 3 –  Tradable/non-tradable divergence.
Source:  Statistics New Zealand, The Treasury calculations

One factor often identified as contributing to this divergence is the high level of the NZ dollar, which reduces the returns for firms exporting and/or competing with imports, all else equal.  There does appear to be some relationship between the exchange rate and tradable sector output (Figure 4).  The main exception is the fall in the NZ dollar in 2008/09 when tradable output failed to rise, but this was during the global financial crisis. 

Figure 4 – Tradable/non-tradable and the TWI
Figure 4 –  Tradable/non-tradable and the TWI.
Source:  Statistics New Zealand, RBNZ, The Treasury calculations

What happened within the tradable sector?

Most of the industries identified as tradable saw output fall over the last five years, partly reflecting a high exchange rate.  This is not to say the exchange rate is the only factor driving the recent poor performance of the tradable sector.  The global financial crisis clearly had an impact in 2008/09, particularly manufacturing and tourism, and drought had a negative impact on the primary sector at a similar time before recent good growing conditions.  Import-competing manufacturing firms faced additional factors such as competition from China and trade liberalisation.  Also, the expansion of government-dominated industries over much of the 2000s may have crowded out resources from the tradable sector. 

Output in agriculture, forestry and mining has risen over time, encouraged by higher world prices.  Service exports have been weak owing to travel spending, which is particularly sensitive to a high exchange rate.  Industries on the verge of being tradable such as communications have performed better, but even communications has been flat in recent years owing to fewer call minutes.

What are the issues with these approaches?

This analysis suggests the “standard” tradable measure is an approximation to using reasonable export and import shares.  However, there are still several issues to be aware of.  Lower and higher shares – and the share chosen is arbitrary – can show somewhat different pictures as industries that are on the cusp of being tradable come in or fall out of the equation.  The time period chosen for analysis can also influence the comparisons. 

A more significant difficulty is gauging shifts over time within an industry towards producing tradable output.  For example, retail trade may still be dominated by non-tradable activity such as grocery stores, but new technology has significantly changed the retailing of books, music and movies to make them more tradable.  The simple demarcation of an industry as either tradable or non-tradable for long periods of time can miss important changes within an industry.[4] 

Areas for further work

This Special Topic outlines how we constructed a measure of tradable output and notes the many caveats, in particular, measuring tradable services.  Newer Input-Output tables and the move to the Australian and New Zealand Standard Industrial Classification (ANZSIC) 2006 classification could help refine our tradable measure (eg, ANZSIC 2006 has an improved breakdown of services).  There are other measurement issues, including adding together chain-linked seasonally-adjusted data, treating non-tradable output as a simple residual with GDP and better measuring import-competing industries, which could be improved on.

This note has not examined other aspects of tradable sector performance, including productivity, employment, capital intensity, or prices.  A fuller analysis of relative tradable and non-tradable prices, beyond the exchange rate, could shed light on why resources appear to have shifted from the tradable to the non-tradable sector.  However, it is important to look at other measures such as exports and imports, industry productivity, unit labour costs and exchange rates (as in February 2012’s Special Topic).  These also show a fall in competiveness and build-up of imbalances in New Zealand over the 2000s.


Page top