external barriers are equal. If it serves union markets by exporting only, its profits are:
(19)
∏x = nπ[t,(n-l)t]
By contrast with equation (1) in Section 2, the multinational now faces competition from all
n union firms in each of the union markets. However, n—1 of these face the same access cost
t as the multinational, so the sum of rivals’ access costs, t-k, equals (n—1 )t. From the fourth
row of Table 1, the multinational’s total profits in the linear case are:
∏x = n
l-2rl2
n+2
(20)
Thus the prohibitive tariff t in this case is 0.5. Indeed, with equal internal and external
barriers, the same threshold is sufficient to eliminate all intra-union trade.
Suppose instead that the multinational establishes a single plant in the union. In the
market where it sets up (call it market i) it has preferential access, under-cutting and hence
out-selling the firms from other union countries. For sufficiently high tariffs, these firms will
be squeezed out of that market.8 From the fifth row of Table 1 (recalling that τ=t in this
sub-section), this happens under linear demands ift exceeds one third. In that case, the union
market in which the multinational has located becomes a duopoly, with only the local firm
providing competition. For the present, defer consideration of that case: i.e., assume that t
is less than one third. The multinational faces no trade barriers in competing against the local
firm and n-1 partner-country firms in market i. However, it continues to face a tariff of t
in the other markets. Hence its total profits are:
8 With the multinational located in market i, firms from partner countries face more intense
competition there. A firm from country j earns profits of π[t,(n-1)t] on its sales in market
i without FDI, but only π[t,(n-2)t] if FDI occurs. This implies a threshold tariff on intra-
union sales which is lower than the value of tt implied by (19).
12