∏" = π(0) -y+(n-l)π(r) (2)
The difference between this and the profits from exporting can, from (1), be written as:
∏w-∏x = γ(r√), where: (3t,f) ≡ π(O) -/"π(r) (3)
+ -
The function γ(t,f) measures the net gain from tariff-jumping into an individual union market:
supplying it from a local plant rather than incurring the unit cost t of supplying it from
outside. It is clearly increasing in t and decreasing in f. Other things equal, the multinational
will engage in FDI if and only if γ(t,f ) is positive.
However, under the assumptions made so far, it does not make sense to locate in only
one union country. The total profits from locating plants in m member countries (m<_n),
denoted by ΠFm, equal:
∏λm = (4 [π (O) (4) + (4-т) π (4 (4)
Now the firm enjoys improved access to m markets (though it must incur the fixed costs of
constructing m plants), and unchanged access to the other n-m. Extending the logic which
led to (3), the profit gain from establishing an additional plant is:
∏⅛-1F>-ι = γ5)J) (5)
Clearly, if it pays to set up a plant in one union country (i.e., if γ(t,f ) is positive), it pays to
set up in all. The only form which FDI takes in this case is a different plant located in each
of the n member countries. This leads to total profits denoted by ΠFn:
(Iirn = n[π(0)√] (6)
Because all union countries are identical, and because there is no production cost advantage
to locating inside the union, it never pays to produce with less than n plants.