This proves that, as claimed above, the non-monotonicity described in Propo-
sition 2 is entirely due to contractual incompleteness. Note also the opposite
role of parameter β. Under complete contracts, for mode O to arise as an equi-
librium configuration, both β and τ have to be small (specifically, β < (1 - τ2)).
Intuitively, as already pointed out, the joint surplus from outsourcing is inde-
pendent from β . For small values of τ the exporting modes are expensive, and
for small values of β the Y -firm’s share of surplus and therefore its payoff from
outsourcing are large. Under incomplete contracts, if mode O arises, τ has still
to be small but β has to be large (precisely, β > τ /2). This is due to the fact
that the joint surplus from outsourcing depends on the ex-post bargaining pow-
ers of the two parties. In particular, it drops to zero as β goes to zero. The
reason is that, under incomplete contracts, if its ex-post bargaining power is
negligible, the supplier does not produce any amount of intermediate for fear of
being held up.
These results are summarized in Figure 3, which shows that for large trade
costs (small τ) and small MNE’s shares of surplus from outsourcing (large β), Y -
firms choose between final exports and FDI plus intermediate exports favoring
the former over the latter when market size is small. For small trade costs
(large τ) and large MNE’s shares of surplus from outsourcing (small β), Y -firms
choose between final exports and FDI plus intermediate exports favoring the
former over the latter when market size is small. Lower trade costs promote
the choice of X over O, while trade cost changes have no impact on the choice
between X and E.
Comparing Figures 2 and 3 points out that contractual incompleteness al-
ters someway dramatically the choice of supply mode made when contract are
complete. In particular, the trade-off between modes E and O is completely
reversed. Indeed, for low trade costs, all the rest given, MNEs may shift from
intermediate exports for local assembly to local outsourcing due to improve-
ments in the writing and enforceability of outsourcing contracts.
7 Conclusions
We have introduced contractual incompleteness in an otherwise standard model
of MNEs based on the trade-off between proximity and concentration. This has
been shown to alter the results of the original set-up from both a positive and
a normative points of view.
In terms of positive implications, we have shown that, for large markets,
incomplete outsourcing contracts can account for the emergence of FDIs not
only when trade costs are large (as predicted by the proximity-vs-concentration
set-up) but also when trade costs are small (as pointed out by empirical ob-
servation). The reason is the positive effect that lower trade costs have on the
ex-post bargaining position of MNEs with resect to local subcontractors.
In terms of normative implications, we have shown that contractual incom-
pleteness alters someway dramatically the choice of supply mode made when
contracts are complete. In particular, for low trade costs, all the rest given,
18