1=c(I) units of whichever manufacture i. The higher the investment I ,the
lower is c(I). Hence, the higher is the efficiency of the plant in transforming
inputs into outputs. Investment expenditures consists of φ(I) units of labor.
Consider a MNE that has decided to locate its production facility for good i
in country h (possibly its own home country). It has to solve a two-stage game.
In the first stage, the MNE chooses how much to invest in order to reduce its
unit production costs; in the second stage, the price for its good is set.
As for the second stage, recall that each MNE can produce and sell a given
good i at a world-wide level using its IRS technology (trade is free). However,
the same good can also be supplied by competitive firms belonging to any coun-
try using a CRS technology. We restrict the analysis to cases in which the labor
supply L is sufficiently big to ensure that world demand of any good i is partly
satisfied by the competitive fringe (recall that the employment size of MNEs
is fixed). By choice of units, we assume that one unit of each good i is pro-
duced by competitive firms using one unit of labor in all locations. This implies
that wages must be equalized across countries. Using labor as the numeraire,
each manufacture i will be supplied by competitive firms at a unit-price. From
the assumption of Cobb-Douglas preferences, monopolistic MNEs would like to
charge an infinite mark-up over marginal costs in all markets. Hence, the only
solution to the second stage of the MNE problem is the limit-pricing one: MNEs
will fix a price equal to unity.
Turning to thefirst stage, the MNE sets its investment by equating marginal
costs and marginal returns of investment. The costs are due to plant-level
fixed labor expenditures. Once the investment is made, plant-level fixed costs
are sunk. So, the investment “ties” the MNE to country h. The benefits of
the investment are due to increased efficiency in production that results into a
higher mark-up. The returns of investment are not fully appropriated by the
MNE. Countries are able to appropriate part of the rents of MNEs. We assume
that, because of unforeseeable contingencies, negotiations between MNEs and
countries’ representatives concerning rent division can only occur ex-post, once
investment costs are sunk (incomplete contracting).14 Thus, in deciding about
the investment, the MNE knows that country h will extract a fraction βh of its
operating profits. This fraction represents the bargaining power of country h;
and bargaining power differs across countries. We think of this rent extraction
in a broad sense. Obviously, it can simply be interpreted as a tax on profit
repatriation, but it can also include more indirect measures of rent sharing like
performance requirements or other regulatory restrictions. 15
Since perfect symmetry holds across sectors, we can omit index i and express
the profits of a MNE in country h as follows
¼h = (1 - βh) (1 - c (I)) C(I) - Φ(I): (2)
14See also Schnitzer (1999) for a formalization of the hold-up problem faced by MNEs
vis-à-vis host countries.
1 5 Alternatively, in a broader framework, it can be thought as a share of MNE rents accruing
to local factors when property rights are not fully enforced (see, e.g. M arkusen (1998a)).