6. Conclusion
In this paper we have presented a range of descriptive statistics about New Zealand's merchandise trade relationships. Here we summarise the main findings in each section and provide some initial discussion on what these might indicate about New Zealand's export performance to date. We also raise some of the questions suggested by our findings and indicate where we might need to look further in order to inform the policy debate.
Our first finding is that over the last 18 years, New Zealand has experienced a significant amount of diversification in both the products it exports and the countries with which it trades. However, most of this diversification occurred prior to 1996, and the last ten years have seen little change in the degree of concentration at any level.
At the firm level, the very high concentration of exports among a small number of firms has raised concern among a number of commentators. However, there are positive signs in terms of substantial increases in the number of firms which have begun to export over the past ten years.
As noted at the beginning of Section 4 it is difficult to establish whether product diversification should necessarily be seen as a good thing, particularly at the aggregate level. On the one hand, a more diverse portfolio of products and partners implies that we are less at risk from shocks to particular product markets or regions. Alternatively, however, concentration may be seen as a positive sign of increasing focus on the products which New Zealand has existing comparative advantage in, and may suggest greater opportunities for learning-by-doing through clustering of industry and supporting activities. A better understanding of the sources of changing composition is required to judge whether New Zealand's experience in this area should be seen as encouraging.
In order to begin to understand the sources of changing product composition, we therefore decomposed export value growth over time according to whether it came from the introduction of new products, new market entry, or the entry of new firms, or whether it was rather through increases in the value of existing trade relationships. The SITC 5-digit analysis suggests that New Zealand has seen little entry of new products, and that almost all of our export growth has come from increasing exports of existing products (less than 1% of export growth can be attributed to new products). In contrast, a significant amount of export growth came from the geographical spread of existing products to new markets.
There also remains room for substantial change within product groups, through increasing variety and quality. Future work will examine this in more depth by considering changes in the unit value of export products.
The results for the HS10 data are not easily comparable to that for the SITC5 data, due to differences in the time period considered and the definitions of entry and exit. These results would suggest somewhat greater change in export composition at the product and relationship level, with a significant amount of churn as the value generated by the entry of new products has been largely balanced by the exit of previous export products.
Individual export spells at the product-market level are relatively short on average, but the probability of failure drops dramatically after the first few years. There are substantial differences across a range of product and market characteristics, with higher value relationships and those involving OECD countries tending to last longer than others. There are also differences across product types, with agricultural and forestry products tending to maintain longer export spells than manufactured goods.
At the firm level we see a lot of firms have only very short spells of exporting, such that they are classed neither as entrants nor exits. However, over time there has been significant growth in both the number of exporters in any given year and the share of active firms with exports. Many of the firms which are observed to enter or exit from export markets are firms which also begin operation or cease over the period considered.27
Firms tend to alter their product and partner mix within continuous spells of exporting, such that export spells tend to last longer than the relationships that underlie them. Among continuing exporters, product changes appear to outweigh partner changes in determining the dynamics of changing export relationships. Many new firm-product-market relationships involve a firms sending a product which they had not previously exported to one of their existing partners. Conversely, firms often maintain trade with a certain destination but stop sending a particular product.
The duration of firm level export relationships also differs across industries and product types. While manufacturers tend to have relatively long export spells, compared to firms in the agriculture forestry and fishing or wholesale trade industries, they are more likely to alter the mix of products and destinations which underlie these relationships than are primary producers. Interestingly, firms whose main activity is wholesale trade tend to have relatively short export spells, both at the firm and the relationship level.
This paper has outlined a number of patterns observable in New Zealand's export data. In all cases, these results are merely descriptive, and further work is required to determine the appropriate areas for policy development, both through improving the data to capture better link enterprises over time, and to better understand what is driving the observed patterns.
Back to Top