4. Export Concentrationand Diversification
A number of publications have commented on the degree of diversification in the product and destination mix of New Zealand exports. This stems at least in part from concerns over the heavy reliance on the primary sector. Various commentators have suggested that economic growth may be hindered by a reliance on primary products – the so called "Curse of Natural Resources".10
Lederman and Maloney (2003) find that while there is little evidence for a negative effect on growth associated with an abundance of natural resources, there is some evidence for a negative effect from export concentration. Heavy reliance on any one export creates a vulnerability to changes in demand for that good, and may raise issues of political economy which can be harmful to growth (for example, through lobbying by interest groups). At the same time, geographic concentration of export destinations (and import sources) leaves exporters vulnerable in the case of rapid changes in the political or economic situations of their key trade partners.
Conversely, increasing concentration may be a sign of reallocation of resources towards the sectors in which New Zealand has a competitive advantage. This can have a positive effect on growth both through a level effect, as resources move to more efficient uses, and a growth effect, through dynamic feedback loops and learning-by-doing.
Existing studies of New Zealand's merchandise export composition suggest that there has been substantial diversification, and imply that this has been largely positive for the economy. Easton (1997, 2004) suggests that between 1965 and 1980 New Zealand's rate of product and export market diversification was faster than in any other OECD country. However, this analysis used only 2 digit industries, rather than specific products, and is hence significantly influenced by New Zealand's high level of concentration in agriculture at the beginning of this period.
Nees and Nixon (2005) present a detailed discussion of the diversification of New Zealand's merchandise exports, at the level of 1 digit SITC classifications for New Zealand's top export partners. They show a substantial diversification in New Zealand exports between 1964 and 2003, concluding that "New Zealand's exports can no longer be deemed to be concentrated in either markets or commodity types" (Nees and Nixon, 2005, p.i)
Grimes (2006) shows that diversification in merchandise exports (also measured at the 2 digit level) has helped to reduce the volatility of New Zealand's terms of trade. Lower volatility is in turn associated with faster economic growth, although Grimes finds that New Zealand's diversification away from meat and dairy products (as a share of exports) since 1972 had the proximate effect of reducing the level of the terms of trade, relative to the early 1970s (ie. making us worse off).
Black et al (2003) note that while there was relatively little change in the broad structure of exports (2-digit level) over the 1990s, significant changes have occurred at a more disaggregated level. Compositional changes have been particularly strong within the categories involving less transformation, suggesting that structural change is continuing within New Zealand's primary based industries.
This view is supported by Vercoe's (2007) description of recent export patterns using the new Level of Processing classification developed by Statistics New Zealand. This classification distinguishes between unprocessed and processed primary products and simply transformed and elaborately transformed manufactured goods. Over the period 2002 to 2006, exports of processed primary goods have grown strongly, and by 2006 had overtaken unprocessed primary products as the top value group. The growth in processed primary exports was driven by a combination of continuing growth in the dominant product within that category, milk powder, and rapid growth in a number of smaller product groups including malt extract, wine and dairy products other than cheeses, milk powder and butter. Similarly, while exports of simply transformed manufactures trended downwards over the period, there was strong growth in exports of elaborately transformed manufactures.
In this section, we document the degree of diversification of NZ export products and destinations at the SITC 5 and HS10 levels, using the full range of countries to which New Zealand exports. Several measures of diversification are possible, but for brevity we focus only on the standard Herfindahl Index: 
, in which si is the share of product (country) i and n is the total number of products (destinations).11 This gives an index ranging from 1/n to 1, with smaller numbers indicating greater diversification. Merged products are excluded from the HS10 results on products and relationships, but included in the country level results.
While the focus of this paper, and those mentioned above, has been the diversification of merchandise exports, it is important to recognize that they reflect only one side of the export diversification story. One of the key drivers of diversification in New Zealand's overall export earnings has been the rise of services exports. By 2007, services accounted for over a quarter of New Zealand's total export earnings (Statistics New Zealand 2007). The composition of New Zealand's services exports are discussed in Ballingall and Stephenson (2005).
4.1 Product and Destination Concentration
Figure 4 shows the Herfindahl index for the diversification of products over the 18 years to 2005. Both the HS10 and SITC5 product classifications are shown. The greater diversification for the HS10 products reflects the greater number of product groups over which products are split in the more detailed data. While there was significant diversification between 1988 and 1995, this had effectively tailed off by the mid 1990s, and the period from 1999 to 2001 saw increasing concentration in export value. This appears to be caused by a rise in the value of dairy exports, which make up a significant proportion of New Zealand's aggregate export value throughout the period.
There has also been some diversification in the geographic destination of New Zealand exports, although this has been quite uneven over time (Figure 5). The overall diversification in both products and markets implies a similar diversification at the level of product-partner relationships, particularly over the years from 1988 to 1996 (Figure 6).
4.2 Firm Level Concentration
At the level of the firm, it is well recognized that New Zealand's exports are concentrated among a small proportion of firms. This is even more obvious when the focus is on merchandise exports, thereby excluding service exports which are received onshore, such as tourism and international education, and "weightless" exports such as financial and business services and licenses which do not involve the transport of physical materials over New Zealand's customs border.
In any given year between 1996 and 2005 no more than 2.1% of firms in New Zealand show up as having any "traditional" exports, in the sense of goods flowing across the border. This is not however surprising, as many firms are in industries which show little opportunity or need to export. For example, most service industries, many local retailers, and branches of larger international enterprise groups are unlikely to show significant merchandise exports. In the analysis that follows we therefore focus on three industries which have significant merchandise export potential – agriculture, forestry and fisheries (AFF), manufacturing (MANU), and wholesale trade (WST). Together, these three industries account for over 85% of goods export value in nine of the ten years covered (Table 3).12
The outlier year is 2002, in which nearly 25% of value is attributed to enterprises outside of the three main groups. This is due to a spike in the share of value attributed to firms whose listed industry groups are Property and Business Services, and Finance and Insurance. Potential explanations for this spike revolve around lags between firms changing their main activities, and this being reflected in their reported ANZSIC codes, and changing roles of enterprises within multi-enterprise groups which have not been immediately captured in the allocation of export value.
We therefore allocate firms to their modal industry – the ANZSIC which is most commonly associated with that enterprise. The main effect this has on the data is through shifting some value away from Finance and Insurance and Property and Business Services, and into the three main merchandise exporting industries.
The last ten years have seen a substantial increase in exporters, both in absolute number and as a proportion of active firms in each industry. Within the period 1999 to 2004,13 the percentage of firms exporting increased by around 22% in Manufacturing, and 30% in Agriculture, Forestry and Fishing, and Wholesale Trade (Table 4).
Differences in export proportions across industries are determined by market structure, as well as the degree of trade in specific goods. For example, while goods produced by the Agriculture, Forestry and Fishing industries are highly traded, this trade is often handled by either intermediaries (whose main activity would commonly be listed as wholesale trade) or processing companies (which would be listed as manufacturers).
However, export value remains highly concentrated, with the top quartile of exporters contributing over 98% of export value, and the top 10% of firms accounting for around 94% (Table 5). There has been a slight increase in the share of the top quartile over time, but no obvious trends at the very top of the distribution.
As would be expected, the probability and value of exporting are strongly related to firm characteristics, including firm size and ownership. Table 6 looks at export propensity and the share of total export value by firm size. Firms in the top quartile of total sales have a much greater propensity to export, and account for over 99% of total export value in every year over the observation period.14
Foreign ownership is also strongly related to export activity (Table 7). Information on foreign ownership and control is only available for between 22 and 35% of firms in any given year, for the years 1996 to 2005 (the years covered by the LBF), as it is sourced from IR4s. Overall, foreign controlled companies have both a greater probability of exporting, and a higher average value of exports per firm among those firms which do export, such that they account for a disproportionate amount of export value relative to their number. However, it is notable that foreign controlled companies are also substantially larger than New Zealand firms on average so their higher exporting may be associated with this larger size, as much as the ownership factor.15 Future work on the determinants and impacts of exporting on firm size, ownership and performance will look in detail at the causality of these relationships.
Back to Top