Although both positive and negative DR are simply specific cases of trade creation/diversion,
there are two reasons why it is helpful to distinguish them separately. One is that the overall
impact of any given trade policy change will be determined by the relative importance of
‘diversion reversal’ effects. The second is that this relative importance will be influenced by
the number of countries that are covered by any particular regime.
To illustrate these differences Table 1 sets out in simplified form the effects of each relative
to the others on the level of imports, the location of the producers who will have to adjust to
increased competition, and the effects on consumers in the importing countries. The effect on
the level of imports is the determining change - the others flow from it.
Table 1: Trade creation, trade diversion and diversion reversal
|
_____Reshuffling effect |
______Level of imports______ |
Adjustment impact |
____Consumer impact____ |
|
Trade creation_______________ |
Large increase_______________ |
Domestic producers_________ |
Significant gain________________ |
|
Trade diversion______________ |
No change________________ |
Efficient exporters a___________ |
Small gain or nil b______________ |
|
Positive diversion reversal |
Small increase_______________ |
Inefficient exporters a__________ |
Small gain___________________ |
|
Negative diversion reversal |
Reduction or no change______ |
Inefficient exporters___________ |
Nil or lossl b_______________________ |
|
Note: (a) And, possibly, domestic producers. (b) Depending on how government replaces lost tariff revenue. | |||
A policy change that results only in trade creation will result in the largest relative increase in
imports for any given level of tariff cut and price elasticity of supply and demand. By
definition trade diversion will result in no increase in imports; if the relative costs of supply
are such that the tariff cut allows an increase in imports from the preferred suppliers that
exceeds the fall in those from the non-preferred the net increase will be classified as trade
creation. It is possible that consumers may gain from trade diversion if prices fall as a result
of the tax cut, but that partly depends on whether the government seeks to offset the tariff
reduction by increasing other taxes (or by reducing expenditure that benefits certain
consumer groups). In either case there are likely to be distributional effects.
Positive DR can be thought of as a sub-category of trade creation. Imports are likely to
increase but only in proportion to the preference margin that is eroded and in relation to the
supply capacity of the beneficiaries. The only element of any constraints on imports that will
have been removed is the tariff disadvantage of those competitive suppliers that have been
granted improved access vis à vis less efficient preferred states. For many products the pre-
existing margin of preference is quite small - typically less than 10% for horticulture. An
extension of preferences may not fully remove this margin and, unlike multilateral
liberalisation, it may not apply to all suppliers. If the ‘newly preferred’ states are not globally
the most efficient (merely more efficient that the previously preferred ones) the effect will be
smaller than the ‘pure’ trade creation envisaged in the literature.
This is why the table shows a small consumer gain relative to the entry for trade creation.
There will be a need for adjustment by formerly preferred, less efficient exporters. Whether
or not domestic producers also have to adjust depends on the extent to which the new
preferences remove import taxes and the number of suppliers that they affect. Whereas trade
creation affects domestic producers by definition, the extent of positive DR is a matter for
empirical observation. It is possible to envisage cases in which the increase in imports is so
small as to avoid any significant displacement of domestic production.
Negative DR will normally result in a fall in imports: the exports of the inefficient supplier
can be expected to decline but no change would have happened to cause an offsetting rise in
12