Export Intelligence Reports

MBIE has produced a series of one-page Export Intelligence Reports that analyse the export performance of 21 different product categories identified from within the Food & Beverage Information Project. The export commodities range from relatively unprocessed products, such as cherries and avocados, to products with a larger manufactured component like fruit juices and biscuits.

Read the Export Intelligence Reports [PDF 2.8MB]

Data source

The data is sourced from Statistics New Zealand's Overseas Merchandise Trade datasets. This publicly available data source, contains more than 15 years of detailed merchandise trade data. Observations are classified by destination country (or, in the case of imports, origin country) and product, classified by the New Zealand Harmonised System Classification (NZHSC), at the very detailed 10 digit NZHSC level. There are slightly more than 8 million rows of data.

The Overseas Merchandise Trade data is extremely granular and contains more than 96 million data elements. The main reason for this level of detail is that the data is organised into an finely-grained international standard called the "Harmonised System" (HS). This hierarchical classification system contains more than 15,000 different product categories.

Value, quantity and price

The source data does not contain export prices. Price information is derived using basic accounting relationships for commodities whose units of measure are consistent across their NZHSC codes. Derived prices have enabled us to decompose total export value - measured in New Zealand dollars - into price and quantity components. This is useful to help identify premium export markets, where New Zealand producers earn export premiums.

For example, measured by quantity, beef exports to Australia are 60 per cent larger than compared to the United States. But the actual export revenue obtained from the United States is 42 per cent higher than compared to Australia. The reason is that the United States pays much more for New Zealand beef than compared to Australia.

Making sense of data variability

For some product categories, the month-to-month data contains considerable variation. This can make it difficult to identify longer term value, quantity or price movements. As a solution, trend lines were fitted to the data using a smoothing technique. Additionally, extremely low or high values that have the potential to distort conclusions were excluded using an outlier detection algorithm.

Markets and diversity

For each product category, the countries that received the largest value of exports were identified. For some product categories, total exports are dominated by only one or two countries.

In order to measure export market diversity, a monthly time-series of the number of export markets was constructed. For example, during the past 15 years, the number of countries to which beef is exported has steadily increased from 12 to more than 25 countries.

Future work

In order to analyse this data from an economic perspective, various HS classifications were aggregated to form more consumer-orientated segments. For example, even a seemingly homogeneous product category such as Salmon consists of 26 underlying HS codes. Currently, the Export Intelligence Reports only reflect commodities whose NZHSC codes share a common unit of measure; for example, kilograms or litres. Where NZHSC commodity codes include multiple units of measure, prices cannot not be easily derived. Developments are underway to expand the Export Intelligence Reports to include multi-unit-of-measure commodities using price index measures derived from the export data and consistent with price index theory. We are currently developing a more sophisticated technique using index calculations that will help us estimate prices across different commodities.