two commodity groups separately. Of the available cluster algorithms this paper uses
the Ward method to determine the number of clusters and the K-Means algorithm for
the final partition. Clusters will finally be interpreted and labeled.
To this point, analysis is restricted to the trade performance of developing countries to
OECD countries. A positive development of the trade performance in meat and fruits/
vegetables is interpreted as an indication of successful adjustment to the requirements
posed by standards - or at least a compensation of losses arising from them. However,
data interpretation has to occur carefully since trade performance has several
determinants - a general positive trend observed in world trade, agricultural growth in
developing countries, price and trade policies, the ongoing trend of liberalization, as well
as naturally volatile production and trade volumes - and cannot be fully attributed to
the effect of standards.
To get a clearer idea about relations with standards, additional information on border
rejections and donor investments are collected (compare section 3). This information not
only captures the general increase in importance of standards, but allows to
differentiate between countries: Which countries have more rejections and investments?
Available data on border rejections only provide a broad picture (compare section 3).
They are limited either in terms of yearly availability (for the EU they are only available
since 2001) or in terms of the reported categories (data are either classified according to
the products country of origin or according to the product group). Since EU rejections
are heavily determined by meat and fruit/ vegetable rejections it seems appropriate to
take the total number of rejections notified by country into account 10. In the analysis we
use the unweighted border rejections from the EU and the US. Meaningful weighting is
complicated since no information on quantities rejected is available and the product
groups differ or might be very specific (e.g. nuts in Iran). When looking at results for
rejections weighted by the export quantity of fruits and vegetables no very clear results
were found.
The value of investments from STDF (find explanation in section 3) in US$ is employed
in the analysis, again not weighted with regard to the trade volume. STDF investments
are limited to those countries that receive foreign aid at all so in a developing country
sample it can be used without a systematic bias. It does not reflect, however, national or
private investments in the upgrading or setup process of the national safety and quality
management.
5 Export performance of developing countries in meat and
fruits/ vegetables - patterns of winners and losers
Before turning to the formal analysis as described above, we will briefly introduce
general developments in agricultural trade.
Even though the share of agricultural products in total merchandise trade is shrinking,
international exports in agricultural and food products almost doubled between 1993
and 2004. The traded value increased from nearly 340 billion to more than $ 600
billion [4]. The share of developing countries’ agricultural exports in world trade is with
around 30% relatively constant in the last decade. The average share of meat and fruit/
vegetable products in total agricultural trade is 17.8% and 9.5% during that period [4].
The markets for fruits/ vegetables and meat have a specific importance for many
developing countries. Especially fruits and vegetables have additionally to their
economic importance a high social relevance due to their labor intensity. The average
share of developing countries in these two markets ranged from 35 to nearly 40% market
share in fruit/ vegetable products and around 16- 18% for meat products with a slight
upward tendency since 1998.
Both sectors are dominated by only very few major players. In the fruit/ vegetable sector
Mexico and China already count for 30% of all OECD third country imports in the
average of the years 2001- 2004. This is 5% more than during the time period between
1993 and 1998. Mexico and China are followed by another ten countries with an export
share between 10 and 2%. This group supplies more than 50% of the OECD imports. All
other developing countries have a share of less than 2% on total fruit/ vegetable exports
to OECD countries.
10 The only other groups which are not included in the analysis, but are important sectors for
rejections, are fish and seafood products. With respect to the very low quality of data fish and
seafood products have not been included in the analysis.
8