152 The Multinational Enterprise as an Economic Organization
the very large specific and durable investments in facilities also invoke the problems
of long-term contracts that were identified earlier. Finally, Stuckey gave some
weight to Arrow’s model of vertical integration as a route to securing information:
Nobody knows more about future bauxite supplies and exploration than an existing
bauxite producer.
A good deal of evidence also appears on vertical integration in the oil industry.
The ambitious investigations have addressed the U.S. segment of the industry, but
there appears to be no strong difference between the forces traditionally affecting
vertical integration in national and international oil companies. These studies give
considerable emphasis to the costs of supply disruption faced by any nonintegrated
firm in petroleum extraction or refining. Refineries normally operate at capacity
and require a constant flow of crude-oil inputs. Storing large inventories of input
is quite costly, and so backward integration that reduces uncertainty about crude
supplies can save the refiner a large investment in storage capacity. It also reduces
risks in times of “shortages” and “rationing,” when constraints some-where in the
integrated system (crude-oil supplies are only the most familiar constraint) can
leave the unintegrated firm out in the cold. The hazard of disrupted flows translates
into a financial risk, as vertically integrated firms have been found to be able to
borrow long-term funds more cheaply than those with exposure to risk.
Country-based studies of the foreign-investment process have also underlined
vertical MNEs as the outcome of failed arm’s-length market transactions. Japanese
companies became involved with extractive foreign investments only after the
experience of having arm’s-length suppliers renege on long-term contracts, and
they also experimented with low-interest loans to independent foreign suppliers
as a way to establish commitment.
Vertical Integration: Other Manifestations
The identification of vertically integrated foreign investment with extractive activities
is traditional and no doubt faithful to the pattern accounting for the bulk of MNE
assets. However, it gives too narrow an impression of the role of vertically subdivided
transactions in MNEs.
First of all, it neglects a form of backward integration that depends not on
natural resources but on subdividing production processes and placing abroad
those that are both labor-intensive and footloose. For example, semiconductors
are produced by capital-intensive processes and assembled into electronic equipment
by similarly mechanized processes, both undertaken in the industrial countries.
But, in between, wires must be soldered to the semiconductors by means of a
laborintensive technology. Because shipping costs for the devices are low relative
to their value, it pays to carry out the labor-intensive stage in a low-wage country.
The relationship of the enterprises performing these functions in the United States
and abroad must obviously be a close one, involving either detailed contractual
arrangements or common ownership. This subdivision of production processes
should occur through foreign investment to an extent that depends again on the
transactional bases for vertical integration.