|
Division Cluster of sample number |
No. of |
Countries |
|
Small 1 vegetable |
18 |
Bahrain, Burkina Faso, Dominica, Gambia, Guinea, Guyana, Indonesia, |
|
~ |
7^^ |
French Polynesia, Nigeria, Saudi Arabia, Togo, Trinidad and Tobago, |
|
___________3 |
_________4 |
Barbados, Mozambique, Nepal, Qatar |
|
4 |
24 |
Bangladesh, Belize, Benin, Bolivia, Cuba, El Salvador, Fiji Islands, |
|
____________ 5 ~ |
6 ~ |
Cote d'Ivoire, Cameroon, Dominican Republic, Egypt, Ghana, Kenya |
|
Large 10 |
______________1 |
Peru |
|
vegetable 20 exporters |
9" |
Argentina, Brazil, Colombia, Costa Rica, Ecuador, Guatemala, India, |
|
__________ 30~ |
3 ~ |
Chile, China, Mexico |
|
Outlier 100~ |
1 ~ |
Thailand |
Source: own calculation
According to the particular characteristics of the groups which can be read from the
averages of the three clustering variables in each cluster each group got a specific label
which is displayed in Table 2. To complement the interpretation the “average” value of
fruit and vegetable exports in 2002- 2004 is displayed.
Table 2: Cluster labels - fruit and vegetable exports
|
Cluster |
Cluster label |
Exanples |
No cf |
Mean of clustering variables |
Descriptive | ||
|
Difference |
Ratio 02∕04∕93∕95 |
Coefficient |
Average" | ||||
|
3 |
Very small exporters, |
Barbados, |
4 |
-2150 |
0.4 |
0.85 |
145 |
|
2 |
Vety small expertes, |
FrenehPolynesia |
7 |
6398 |
4.4 |
0.26 |
6258 |
|
4 |
Small exporters, |
Bdivia, |
24 |
11935 |
1.9 |
0.24 |
16229 |
|
1 |
Small and medium |
Panama, |
18 |
-16693 |
0.8 |
0.19 |
19481 |
|
5 |
Mediumexpcrters, |
Kenya, |
6 |
102857 |
2.0 |
0.19 |
237522 |
|
10 |
Large exporter, |
Peru |
1 |
375310 |
3.2 |
0.22 |
543993 |
|
20 |
Large exporters, |
Costa Rica, |
9 |
427361 |
1.6 |
0.11 |
1050800 |
|
100 |
Large exporter, |
Thailand |
1 |
-239517 |
0.8 |
0.19 |
1128955 |
|
30 |
Very large exporters, |
China, |
3 |
1577467 |
1.8 |
0.16 |
3985954 |
a Median
Source: own calculation
Clusters are arranged by the median of the average trade value. Albeit this variable was
not entered in the clustering, it is used as additional information to describe the
clusters. The overall growth of agricultural exports was taken into consideration when
interpreting the cluster. Thus when comparing the country ratios to the ratio of all
exports of developing countries to the OECD, a ratio of fruit/ vegetable exports in 2002-
04 compared to 1993- 95 above 1.4 indicates an increase of exports above the average.
All clusters are discussed in detail as follows:
Cluster 3 (very small exporters, strong losers, instable): Cluster 3 consists of four very
small exporters with an average trade value of fruit/ vegetable exports in the second
period in thousand $ (hereafter “average”) below 10,000 (e.g. Barbados, Mozambique). It
is characterized by strong losses (an average ratio of only 0.4) and strong instability of
export values in the second period. This country group can clearly be labeled as “losers”.
Cluster 2 (very small exporters, strong winners): The second group of seven very small
exporters (including e.g. Uganda and French Polynesia) is characterized by strong
10