Page 88 - DECO503_INTERNATIONAL_TRADE_AND_FINANCE_ENGLISH

International Trade and Finance

Notes 7.4 The Balassa and Grubel-Lloyd Indices

Balassa (1966) proposed the first index of intra-industry trade that measured the degree of trade
overlap—simultaneous import and export—of goods within an industry :
X − M i
i
B j = (X + M ) i . ... (1)
i
where i ≡ commodity within industry j. This index, the ratio of net trade to gross trade, ranging from
0 to 1, with 0 representing “perfect” trade overlap, and therefore pure intra-industry trade, while 1
represents pure inter-industry trade. In order to calculate the degree of intra-industry trade for all
industries (country level), Balassa took an unweighted average for each B :
j
1
B= ∑ B ... (2)
n j
where n ≡ number of industries. This can be generalized to be a weighted index :
∑ w B
B= j j ... (3)
j
where w ≡ industry j’s share of total trade.
j
Though the essence of this index has remained intact to this day, an index that measured intra-
industry trade that gave pure intra-industry trade a value of zero was not intuitively appealing.
Grubel and Lloyd (1975) proposed an alternative index :

(X i + M i ) − X i − M i X i − M i
GL = = −1 = 1 – B ... (4)
(X i + M ) i (X i + M ) i j
where i ≡ commodity within industry j, that assigned pure intra-industry trade a value of 1 and pure
inter-industry trade a value of 0. As with the Balassa Index, the Grubel-Lloyd Index has been calculated
as an (un) weighted average to measure the degree of intra-industry trade at the country level.
This class of index has been criticized for suffering from categorical/sub-group aggregation issues.
These issues have two basic forms that bias the index towards 1 : the grouping of two products in the
same industry that should not be classified together, the canoe and tanker example above; and trade
imbalance. The grouping of two, or more, categories together that should not be in the same industry
is best explained using the following table :
Table 1 : Simple aggregation bias in the GL Index
Category X M |X – M | (X + M ) GL Index
i i i i i i
3-Digit 150 160 10 310 0.968
Sub-Group 5-Digit 0 160 160 160 0.00
Sub-Group 5-Digit 150 0 150 150 0.00

Suppose we have one 3-digit “industry” that contains 2 sub-groups and each sub-group is
independently engaged in (pure) inter-industry trade. We can see that the Grubel-Lloyd Index is
zero for each of these sub-groups, so if we took an average, weighted or unweighted, of the two, the
Grubel-Lloyd Index would still be zero. If, however, the import and export values are summed to
form the 3-digit category, it appears that we have almost pure intra-industry trade with a Grubel-
Lloyd Index of 0.968. Though this is an extreme example, it should be clear that aggregating across
improper categories can lead to a misrepresentation of the degree of intra-industry trade.
The simple aggregation bias example above is a particular case of trade imbalance bias—trade
imbalance, however, can occur when sub-groups are appropriately aggregated. This problem arises

82 LOVELY PROFESSIONAL UNIVERSITY