Lecture 12:
Contracts and Vertical FDI
Instructor: Thomas Chaney
Econ 357 - International Trade (Ph.D.)
In the models we have seen so far, …rms are de…ned as technological
(or property right) entities. In the simplest Krugman model, a …rm cor-
responds to a variety. Only one …rm has the ability to produce a given
variety (because of the existence of …xed cost of developping new vari-
eties, there will never be two …rms producing the same variety). One can
either think that this …rm is the only one with the know-how to produce
this speci…c variety, or that it is the only one owning the property right
of this variety. In the extended Melitz model, a …rm corresponds both to
a variety, and a technology of production. Once again, this is merely a
technological de…nition of a …rm. A …rm corresponds to the entity that
is able to produce this speci…c variety, and it has a given technology to
produce this variety, none of which are transferable between …rms. In
the Helpman, Melitz and Yeaple model of multinational …rms that we
saw, once again, a …rm corresponds to a variety, and a technology to pro-
duce that one variety, with the extra possibility of splitting production
between di¤erent countries. The technological boundaries of …rms have
been extended, but we still have nothing to say about the legal bound-
aries of …rms. Whether a …rm own foreign facilities, or whether is sells
o¤ a licence to produce its variety with its productivity is indeterminate.
In all those models, there was no strict de…nition of the legal bound-
aries of …rms. If one were to open up …nancial markets, and allow the
purchase of …rms, the price of a …rm would exactly equal the expected
discounted sum of future pro…ts generated by this …rm, so that the "le-
gal" ownership would be indeterminate.
Strictly speaking, this is actually not exactly correct. In a world
with asymmetric random shocks hitting di¤erent economies, such as in
Ghironi and Melitz, households in a country will be subject to some
risk, and they may be willing to diversify away those risks by investing
in foreign …rms. In such a case, the ownership structure may no longer
1
Figure 1: Share of Intra…rm Imports, in Industries with Di¤erent Capital
Intensities.
be indeterminate. Actually, even in Ghironi and Melitz, there is no
such motive for portfolio diversi…cation. The authors do not solve for
the optimal allocation of portfolio in a world with uncertain productivity
shocks, they only solve for the transitional dynamics with no uncertainty,
after the realization of an unexpected shock. However, even in such
a world, because we have always assumed competitive labor markets,
households may want to diversify the international portfolio of assets,
but they would always own all …rms in a given country. They may hold
more of less of assets from di¤erent countries, but never more or less
of an individual …rm. So even in such a case, the legal boundaries of
individual …rms would be indeterminate.
1 Antras (2003)
0n 1 0n 11
ZY ZZ
U = @ y (i) diA @ z (i) diA ; ; 2 (0; 1)
0 0
1=(1 )
y (i) = AY pY (i)
1=(1 )
z (i) = AZ pZ (i)
2
Figure 2: Share of Intra…rm Imports from Countries with Di¤erent Cap-
ital/Labor Ratios.
RY (i) = pY (i) y (i) = A1Y y (i)
RZ (i) = pZ (i) y (i) = AZ1 y (i)
y (i) = xY (i)
z (i) = xZ (i)
1
Kx (i) k
Lx (i) k
xk (i) =
k 1 k
k 2 fY; Zg
1> Y > Z >0
1
fr kw k ; k 2 fY; Zg
…nal good producer: F
supplier: S
lump sum transfer Tk (i) from S to F
= + (1 )>
max Rk Kx ~ ; Lx ~ rKx ~ wLx ~ f r k w1 k
~ 2f ; g
8
< Kx ~ = arg maxKx ~ Rk Kx ; Lx ~ rKx
s:t:
: Lx ~ = arg maxLx 1 ~ Rk Kx ~ ; Lx wLx
3
1 r k w1 k r k w1 k
pk;V = 1
> = pk
k k
1
1 r k w1 k r k w1 k
pk;O = > = pk
k
(1 )1 k
F
k;V + f r k w1 k
( k) = F
k;O + f r k w1 k
k
(1 ) (1 2 k)
1
( k) = 1+ 1+ (1 ) 1
1 (1 k) + (1 2 k) (1 )
^ =1
Y >^ > Z
E = rK + wL
X
nk (Kx;k + Kf;k ) = K
k2fY;Zg
X
nk (Lx;k + Lf;k ) = L
k2fY;Zg
1 ~Y + (1 ) 1 ~Z
w L K K
= =
r 1 L L ~ + (1 )~ L
Y Z
~Y = Y 1+ (1 Y) 2 1
~Y = Z (1 + (1 Z ) (2 1))
E = rK + wL
X
nk (Kx;k + Kf;k ) = K
k2fY;Zg
X
nk (Lx;k + Lf;k ) = L
k2fY;Zg
1 ~ + (1 ) 1 ~
w L K Y Z K
= =
r 1 L L ~ + (1 ) ~Z L
Y
4
Figure 3: Patterns of Production (2 country case).
~ = 1+ (1
Y Y Y) 2 1
~Z = Z 1+ (1 z) 2 1
~ >~
Y Z
X
njk Kx;k
j j
+ Kf;k = K j ; 8j
k2fY;Zg
X
njk Ljx;k + Ljf;k = Lj ; 8j
k2fY;Zg
M ij = si njY pY y + njZ pZ z = si sj (rK + wL)
ij
Mintra = si sjY pY y
~ Z (1 Kj ~Z K
1 L ) Lj LL
ij
Sintra =
~ ~ (1 Kj
L ) Lj + K
LL
Y Z
2 Antras (2005)
5
Figure 4: Volume of Intra…rm Imports.