Taxes and Entry Mode Decision in Multinationals: Export
and FDI with and without Decentralization
Yoshimasa Komoriyay
Søren Bo Nielsenz
Chuo University
Copenhagen Business School
Pascalis Raimondosx
Queensland University of Technology
November 1, 2016
Abstract
We use a duopoly model to simultaneously consider both foreign direct investment (FDI)
and decentralization decisions, featuring transfer pricing. To investigate the relationship
between corporate taxes and the optimal entry mode for the foreign market, we examine
and compare …rms’pro…t in each of three entry modes: export, FDI with decentralization,
and FDI with centralization. We …nd that, even when a host country does not have any
advantage, FDI can be undertaken because of a strategic incentive. We also …nd that trade
liberalization promotes decentralization of multinationals and that a tougher regulation of
transfer pricing may expand the region of no centralization.
Keywords: decentralization; foreign direct investment; transfer pricing
JEL Classi…cation: F21, F23, H25
Very preliminary! This is one of our ongoing research projects undertaken while Komoriya was visiting the
Department of Economics in the Copenhagen Business School as a visiting scholar. Komoriya would like to thank
the department for its hospitality.
y Faculty of Economics, Chuo University, Hachioji, Tokyo 192-0393, Japan; E-mail: [email protected]
z Department of Economics,
[email protected]
x Economics and Finance,
[email protected]
Copenhagen Business School,
QUT Business School,
Frederiksberg 2000,
Brisbane QLD 4000,
Denmark;
Australia;
E-mail:
E-mail:
pas-
1
Introduction
Transfer pricing is one of the pro…t-shifting instruments for multinationals. In addition to the
pro…t-shifting role, transfer pricing may have other roles1 . In oligopoly settings, multinationals
have an incentive to manipulate their intra-…rm prices for a¤ecting competition that their subsidiaries are facing. This manipulation of intra-…rm price is denoted the strategic role of transfer
pricing.
The strategic role of transfer pricing is widely known after Schjelderup and Sørgard (1997).
The motivation of their research is as follows: “In the IO-literature on MNE behaviour both
delegation of authority to local a¢ liates and strategic interaction are discussed. [t]he focal point
in the IO literature on MNEs, however, is on foreign direct investments, not on pro…t shifting through transfer pricing (Schjelderup and Sørgard, 1997, pp.278)”2 . With this motivation,
Schjelderup and Sørgard (1997) …nd this new role of transfer pricing. However, foreign direct
investments (FDI) is taken as a given in their analysis. Therefore, in this paper, we consider both
of FDI decision and pro…t shifting, focusing on the strategic motive of transfer pricing.
Decentralization is the key of the strategic role of transfer pricing. In the model of Schjelderup
and Sørgard (1997), in addition to FDI, decentralization is also taken as a given. Although Nielsen
et al. (2008) examine decentralization decisions, that is, they investigate whether multinationals
decentralize or not, FDI is taken as a given even in their analysis. Therefore, in this paper,
we simultaneously consider both FDI and decentralization decisions of multinationals, featuring
transfer pricing.
In the literature, Porter (2012) analyzes e¤ects of a corporate tax on choices between FDI and
export. In a duopoly setting, taking …rms’choices between export and FDI into account, Porter
(2012) shows the possibility that an increase in the host country’s corporate tax can create its
desirable situation where the more e¢ cient …rm undertakes FDI and the less e¢ cient …rm does
not. Although, as far as we know, Porter (2012) is one of the rare papers that focus on e¤ects
of corporate tax on choices between export and FDI, she does not consider pro…t shifting and
transfer pricing.
Amerighi and Peralta (2010) investigate choices of modes for serving two countries. A …rm
has two alternatives in their model. One is to have a plant in one country and to serve both
the local and foreign markets from the plant. The other alternative is to have two plants and to
serve markets from the local plants. They combine the proximity-consentration trade-o¤ with a
tax-saving incentive. Although, as far as we know, Amerighi and Peralta (2010) is the …rst paper
that shows the relationship between the pro…t shifting and FDI incentive, they do not focus on
strategic incentives for undertaking FDI because they use a monopoly model.
In our paper, we …rst examine multinationals’pro…t in each of three entry modes into a foreign
market: export, FDI with decentralization, and FDI with centralization (without decentralization). Second, to investigate FDI incentives in decentralization and centralization cases and a
1 See
2 See,
for, example, Göx and Schiller (2007).
for example, Horstmann and Markusen (1992) and Motta (1992).
2
decentralization incentive, we compare after-tax pro…t between entry modes. Finally, we show
the relationship between corporate tax rates and the optimal entry mode for the foreign market
by simultaneously considering FDI and decentralization decisions.
First, we …nd that, even when a host country does not have any advantage, that is, even when
there is no di¤erence in marginal production cost between the home and host countries, there is
no tax di¤erence, and trade costs are not incurred on export, FDI can be undertaken due to a
strategic incentive. Second, we also show that centralized multinationals do not appear in free
trade at least in our current setting and that trade liberalization promotes decentralization of
multinationals. And …nally, we …nd that a tougher regulation of transfer pricing may expand the
region of no centralization.
Section 2 introduces our model and shows pro…ts and welfare in each of the three modes and
Section 3 investigates FDI incentives in decentralization and centralization cases and a decentralization incentive. Section 4 shows the relationship between corporate tax rates and the optimal
entry mode for the foreign market. Section 5 discusses e¤ects of trade cost and Section 7 discusses
that of transfer pricing cost. Section 7 o¤ers conclusions.
2
The Model
In this section, we introduce our model. There are two countries (home and foreign) and two
…rms (…rm 1 and …rm 2). Only …rm 1 serves the home market and both …rms serve the foreign
market. The two …rms compete in Cournot fashion. The output levels of …rm 1 in the home
and foreign markets are, respectively, q1 and Q1 , and the output level of …rm 2 in the foreign
market is Q2 . We use linear demand functions, p = a
in the foreign market, where q
q1 and Q
bq in the home market and P = A
BQ
Q1 + Q2 . p and P are, respectively, the market
prices in the home and foreign markets. Marginal costs for production of the two …rms are c1
and c2 . We assume c1 = c2 = 0 for simpli…cation. Firm 1 has three entry modes for serving the
foreign market: Export, FDI with decentralization, and FDI with centralization. The following
subsections show each of the three modes.
2.1
Export Case
In this mode, …rm 1 is an exporting …rm and it competes with the rival (…rm 2) in the foreign
market. Governments in the home and foreign countries tax locally earned pro…t at th and tf ,
respectively. After-tax pro…ts of …rm 1 and …rm 2 are, respectively,
where
1
=
(1
th ) [(a
bq1 )q1 + (A
2
=
(1
tf ) [(A
BQ1
BQ1
BQ2 )Q1
BQ2 )Q2 ] ;
is a trade cost. We …nd
q1 =
A 2
A+
a
; Q1 =
; Q2 =
2b
3B
3B
3
Q1 ] ;
in the equilibrium. The after-tax pro…ts are
"
2
1
=
(1
a2
(A 2 )
th )
+
4b
9B
2
=
(1
tf )
#
;
(1)
2
By assuming
= 0, we have
1
and de…ne (2) as
2.2
(A + )
:
9B
EX
1
= (1
th )
a2
A2
+
4b 9B
;
(2)
in the following part3 .
FDI with decentralization
In this mode, …rm 1 is a decentralized MNE4 . Before-tax pro…ts from production and sales on
the part of …rm 1 in the home and foreign countries and …rm 2 are, respectively,
h1
=
(a
bq1 )q1 + mQ1 ;
f1
=
(A
BQ1
BQ2 )Q1
f2
=
(A
BQ1
BQ2 )Q2 ;
mQ1 ;
where, m is a intra-…rm price (transfer price). The transfer price m can deviate from the marginal
production cost (of zero). If it is set di¤erent from zero, the company is assumed to incur transfer
pricing cost. After-tax pro…ts of …rm 1 and …rm 2 are, respectively,
1
=
(1
th )
h1
2
=
(1
tf )
f 2:
The third and fourth terms of
1
+ (1
tf )
f1
u 2
m
2
F;
are the transfer pricing cost and the FDI …xed cost. The
parameter u governs the change in transfer pricing cost, if the deviation from zero is enlarged.
In this case, the headquarters of …rm 1 sets the transfer price before its subsidiary and …rm 2
choose their output levels. We solve this model using backward induction. In the second stage,
we have
q1 =
a
A 2m
A+m
; Q1 =
; Q2 =
:
2b
3B
3B
In the …rst stage,
d 1
dm
d
dm
d 1
dm
=
=
1
A (4tf
9B
4
2tf
9B
3th
1)
9
3th + 1 + Bu
4
3 We
4 In
4m 2tf
9
3th + 1 + Bu
4
;
< 0:
relax this assumption in section 5.
this subsection, we assume 4tf 5th + 1 + 3Bu > 0 for obtaining non-negative output levels of both …rms.
If Bu >
4
,
3
we can obtain non-negative output levels.
4
The second order condition holds with the assumption of 4tf
5th + 1 + 3Bu > 0. In equilibrium,
the level of the transfer price is
m=
A (4tf 3th 1)
:
4 2tf 3th + 1 + 94 Bu
Given corporate tax rates, the absolute value of the transfer price jmj decreases in u. If tax rates
are not vastly di¤erent (tf is not much greater than th ), m will lie below marginal cost of zero, in
that case, the transfer price increases in u. Generally, the more pronounced are transfer pricing
costs, the closer m is to zero. If th = tf = t,
A
9ABu
A
+
>
:
4
4
16 1 t + 94 Bu
Although this equation shows the existence of the strategic motive of transfer pricing, the strategic
m=
motive is muted because of the transfer pricing cost5 . And the transfer pricing costs are present,
even though there is no tax rate di¤erence.
Equilibrium outputs are
Q1
=
=
Q2
=
=
and the after-tax pro…ts are
1
=
2
=
(1
(1
h
i
2
A2 (1 th ) + 2Bu (1 tf )
th ) a2
+
4b
8B 2tf 3th + 1 + 49 Bu
tf ) A2 (4tf
16B 2tf
If th = tf = t,
1
We de…ne (3) as
2.3
DC
1
A [2 (1 th ) + 3Bu]
4B 2tf 3th + 1 + 94 Bu
A [8 (tf th ) + 3Bu]
A
;
2B
8B 2tf 3th + 1 + 94 Bu
A (4tf 5th + 1 + 3Bu)
4B 2tf 3th + 1 + 94 Bu
A
A [8 (tf th ) + 3Bu]
+
4B
16B 2tf 3th + 1 + 94 Bu
= (1
F;
(3)
2
5th + 1 + 3Bu)
3th + 1 + 94 Bu
"
a2
A2 (1
t)
+
4b 8B 1
2
t + 2Bu)
t + 94 Bu
> 0:
#
F:
(4)
in the following part.
FDI with centralization
In this mode, …rm 1 is a centralized MNE6 . Before-tax pro…ts of …rm 1 in the home and foreign
countries and …rm 2 and after-tax pro…ts of the two …rms are the same as those in the decentralization case. Unlike in the previous case, the headquarter of …rm 1 chooses both its transfer
price and output level, and …rm 2 chooses its output level simultaneously.
m= A
, Q1 is equal to the output level of the Stackelberg leader.
4
2
this subsection, we assume
tf th + Bu 1 tf > 0, for non-negative output levels of both …rms.
5 When
6 In
5
We have
@ 1
@m
@ 1
@Q1
@ 2
@Q2
=
(tf
th ) Q1
mu;
=
(1
tf ) (A
2BQ1
=
(1
tf ) (A
BQ1
BQ2 ) + m (tf
th ) ;
2BQ2 ) ;
and
@ 1
@Q1
@
@Q1
@2 1
@m@m
(tf
2
th ) + 2Bu (1
@2 1
@Q1 @Q1
@2 1
@m@Q1
@
@ 2
@Q2 @Q2
=
2B (1
=
(tf
=
2B (1
tf ) < 0;
2
2
th ) + 2Bu (1
tf ) > 0 holds with the assumption of
tf ) > 0;
tf ) < 0:
2
(tf
th ) + Bu (1
tf ) > 0.
From F.O.C., we have
(tf
th ) Q1
;
u
(1 tf ) (A BQ2 ) + m (tf
2B (1 tf )
A BQ1
:
2B
m =
Q1
=
Q2
=
th )
;
Thus, in the equilibrium, transfer price and outputs are, respectively,
m =
Q1
=
A (tf
th ) (1
tf )
2
tf )
2
tf )
2 (tf
th ) + 3Bu (1
Au (1 tf )
2 (tf
th ) + 3Bu (1
;
2
=
Q2
=
=
A
2A (tf th )
h
+
2
3B
3B 2 (tf th ) + 3Bu (1
h
i
2
A
(tf th ) + Bu (1 tf )
h
i
2
B 2 (tf th ) + 3Bu (1 tf )
A
3B
3B
h
2
A (tf
2 (tf
i;
tf )
th )
2
th ) + 3Bu (1
i:
tf )
As already mentioned in Nielsen (2014), the output of …rm 1 is consist of two parts. If th = tf = t,
because m = 0, the output of each …rm equals a Cournot output level. And we also have q1 =
6
a
2b .
The after-tax pro…ts are
h
i
2
2
2
A
u
(1
t
)
(t
t
)
+
2Bu
(1
t
)
f
f
h
f
(1 th ) a
+
h
i2
4b
2
2 2 (tf th ) + 3Bu (1 tf )
h
i2
2
(1 tf ) A2
(tf th ) + Bu (1 tf )
> 0;
h
i2
2
B 2 (tf th ) + 3Bu (1 tf )
2
1
=
2
=
If th = tf = t,
= (1
1
C
1
We de…ne (5) as
3
t)
a2
A2
+
4b 9B
F;
(5)
F:
(6)
in the following part.
Comparison of three modes
In this section, we investigate FDI incentives in each of the two FDI modes and decentralization
incentives of multinationals.
3.1
Export and FDI with decentralization
In this subsection, we compare
EX
1
DC
1
and
for investigating the FDI incentive for decentral-
ization case. The condition for FDI is
DC
1
>
,
If
(1
th ) (16tf
15th
EX
1
[ (1
1)
th ) (16tf 15th 1) 18Bu (tf
72 2tf 3th + 1 + 94 Bu
18Bu (tf
th ) < 0, that is, tf
th )]
>
th >
F
:
(A2 =B)
(1 th )2
2[8(1 th )+9Bu] ,
…rm 1
chooses Export. This means that …rm 1 chooses Export when tf is su¢ ciently high relative to
th . If tf
th <
(1 th )2
2[8(1 th )+9Bu]
1 th
16
=
not depends on the relative size of the
1 th
< 1 16th , whether …rm 1 chooses FDI or
1 th + 98 Bu
…xed cost, A2F=B . When F is low enough, when A is high
7
enough, and/or when B is low enough, a …rm 1 chooses to be a decentralized multinational …rm .
Otherwise, …rm 1 chooses Export. If th = tf = t, from (2) and (4),
DC
1
>
EX
1
"
#
8
>F
9
,
A2
8B
,
A2 (1 t)
> F:
72B 1 t + 94 Bu
(1
1
t + 2Bu)
t + 94 Bu
This means that …rm 1 chooses FDI when F is low enough and in particular when F = 0.
7 Because
we use P = A
BX for the demand function in the foreign market, A (A=B) governs pro…t
opportunities in that market.
7
When A = B = 1, F = 0, u = 1, and th = 0:3, the condition for FDI (
185 584tf
> 0:
72 (47 + 40tf )
DC
1
>
EX
1 )
is
Below, Figures 1 has been drawn using these values for A, B, F , u, and th .
FDI premium
0.06
0.04
0.02
0.00
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
t_f
-0.02
-0.04
-0.06
-0.08
Figure 1. Additional pro…t of FDI with decentralization (th = 0:3)
In Figure 1, …rm 1 chooses FDI when tf < tef
185
584
' 0:31678. Because we assume th = 0:3,
we can …nd that, even when tf > th , can …rm 1 chooses FDI.
Proposition 1 Even in a case without the foreign country’s production cost and tax advantages
and trade cost, FDI may take place. Even when the tax rate in the foreign country is higher than
that in the home country, FDI can be superior to export because of a (purely) strategic incentive,
if the tax di¤ erence is small.
Whentf > th , there is no FDI incentive coming from tax-saving incentive. However, in the
above …gures, …rm 1 chooses FDI if the tax di¤erence is small. Although Amerighi and Pelarta
(2010) shows a tax-saving FDI incentive, they do not consider this strategic incentive for FDI.
3.2
Export and FDI with centralization
In this subsection, we compare
EX
1
and
C
1
to investigate the FDI incentive for the centralization
case. The condition for FDI is
C
1
>
EX
1
A2 (tf
,
th )
n
8 (1
th ) (tf
3
th ) + 3Bu (1 tf ) [(3tf 8th + 5) (tf
h
i2
2
18B 2 (tf th ) + 3Bu (1 tf )
8
th )
6Bu (1
o
tf )]
> F:
If th = tf and F > 0, from (2) and (6), …rm 1 does not prefer FDI with centralization to Export
C
1
because
=
EX
1
F.
When A = B = 1, F = 0, u = 1, and th = 0:3, the condition for FDI (
(10tf
52410tf + 11325t2f + 18250t3f + 25236
3)
90 90tf + 100t2f
2
C
1
>
EX
1 )
is
> 0:
141
Figure 2 has been drawn using these values. Because, with our parameters, tf 2 [0; 0:77468)
meets the assumption of
2
(tf
th ) + Bu (1
tf ) > 0, we cut o¤ the rage above tf = 0:7 from
the …gure.
FDI premium
0.05
0.04
0.03
0.02
0.01
0.00
0.1
0.2
0.3
0.4
0.5
-0.01
0.6
-0.02
-0.03
-0.04
-0.05
Figure 2. Additional pro…t of FDI with centralization (th = 0:3)
In Figure 2, when th = tf ,
C
1
=
DC
1
because of F = 0. And FDI with centralization is
better than Export only in the case where the tax rate in the host country is lower than that in
the home country (tf < th ).
3.3
FDI with decentralization and FDI with centralization
In this subsection, we compare
DC
1
and
C
1
for investigating the decentralization incentive8 .
The condition for decentralization is
DC
1
>
,
8 This
C
1
8B 2tf
h
A2 4 (1
2
4
th ) (tf
3th + 1 + 94 Bu
subsection is based on Nielsen et al. (2008).
9
th )
h
2 (tf
Bu (1
2
tf )
i
th ) + 3Bu (1
i2 > 0;
tf )
0.7
t_f
where
4 (1
th ) ( tf
2
th + 2) (tf
th )
Bu (1
8tf + 6th + t2f
tf )
6t2h + 6tf th + 1 .
As Nielsen (2008) showed, decentralization is preferable to centralization when the two countries
have the same tax level. If th = tf = t, from (4) and (6),
DC
1
"
C
1
=
A2
8B
=
A2 (1 t)
> 0:
72B 1 t + 94 Bu
(1
1
t + 2Bu)
t + 94 Bu
8
9
#
Thus, FDI with decentralization is preferred to FDI with centralization when the two countries
have the same tax level.
When A = B = 1, F = 0, u = 1, and th = 0:3, the condition for decentralization (
is
184880tf + 179260t2f
32800t3f + 4000t4f + 46187
40 (40tf + 47) 90tf + 100t2f
2
DC
1
>
C
1)
> 0:
141
Figure 3 illustrate the additional pro…t from decentralization, using the parameter values above.
Dec. premium
0.01
0.00
0.1
0.2
0.3
0.4
0.5
-0.01
-0.02
Figure 3. Additional pro…t of decentralization (th = 0:3)
In Figure 3, …rm 1 chooses FDI with decentralization when tf < tf ' 0:38163
10
0.6
0.7
t_f
4
Optimal entry mode for the foreign market
In this section we explore the optimal entry mode of …rm 1. We directly compare the following
three pro…ts,
EX
1
=
DC
1
=
C
1
=
th ) a2
(1 th ) A2
+
;
4b
h9B
i
2
2
A
(1
t
)
+
2Bu
(1
t
)
2
h
f
(1 th ) a
+
F;
4b
8B 2tf 3th + 1 + 94 Bu
h
i
2
2
2
A
u
(1
t
)
(t
t
)
+
2Bu
(1
t
)
2
f
f
h
f
(1 th ) a
+
h
i2
4b
2
2 2 (tf th ) + 3Bu (1 tf )
(1
F;
when a = 2, b = 1, A = B = 1, F = 0, u = 1, and th = 0:3 in this section. We have
EX
1
DC
1
C
1
=
0:77778;
184tf + 313
=
;
8 (40tf + 47)
=
48582tf
18865t2f + 26700t3f + 11500t4f + 32608
2 90tf + 100t2f
Figure 4 shows these pro…ts (
EX
1
2
:
141
is black thin straight line,
DC
1
is red thin line, and
C
1
is
blue bold line).
Profit
0.84
0.83
0.82
0.81
0.80
0.79
0.78
0.77
0.76
0.75
0.74
0.0
0.1
0.2
0.3
0.4
0.5
Figure 4. Comparison of pro…t in the three modes (th = 0:3)
11
0.6
0.7
t_f
Firm 1 prefers decentralized FDI to centralized FDI when tf < tf ' 0:38163, prefers decentralized FDI to Export when tf < tef = 185 ' 0:316 78, and prefers centralized FDI to Export
584
when tf < tbh = 0:3. The order of the three pro…ts are
EX
1
DC
1
DC
1
C
1 >
>
>
C
1
>
EX
1
for
185
584
DC
1
< tf < 0:38163,
>
EX
1
EX
1
C
1
C
1
>
>
>
DC
1
for tf > 0:38163,
for 0:3 < tf <
185
584 ,
and
for tf < 0:3. Thus, we have the following proposition.
Proposition 2 Given the parameter values, in the case of no foreign country’s advantage and
no trade cost, when the …rm chooses FDI, the …rm also implements decentralization.
This proposition implies no centralized multinationals appear in free trade. Because the FDI
DC
1
…xed cost decreases
C
1
and
(shifts the red thin line and blue bold line downward) and
it does not a¤ect the decentralization incentive, this proposition would remain valid even when
F 6= 0.
5
Trade cost
In this section, we relax the assumption,
= 0: For checking and enforcing robustness, we
investigate how and to what extent trade costs a¤ect the result in the previous section9 . When
6= 0, from (1), the after-tax pro…t of …rm 1 is
"
EX
1
With our parameters,
= (1
= 0:05, we have
EX
1
#
:
7
2 +2 2+5 :
45
= 0:763. In Figure 4, the green bold straight dotted line shows
EX
1
When
2
a2
(A 2 )
th )
+
4b
9B
=
this pro…t. Although, in this case, the trade cost does not a¤ect the decentralization decision,
the cost a¤ects FDI decisions. Firm 1 prefers decentralized FDI to Export when tf < tef = 102 '
235
0:43404 and prefers centralized FDI to Export when tf < tbf ' 0:45327. The order of the three
pro…ts are
C
1
DC
1
>
EX
1
>
>
EX
1
C
1
>
DC
1
for tf > 0:45327,
for 0:38163 < tf <
102
235 ,
and
C
1
DC
1
>
>
EX
> DC
for 102
1
1
235 < tf <
C
EX
for tf < 0:38163.
1 >
1
0:45327,
Thus, as
the corporate tax in the foreign country increases, …rm 1’s entry mode to the foreign market shifts
from FDI with decentralization to FDI with centralization, to Export. This implies FDI with
centralization is feasible when the trade cost is su¢ ciently high. By comparing the results in the
previous and current sections, we can …nd that trade liberalization accelerates decentralization10 .
At tf ' 0:38163,
DC
1
tf
(1
9 In
C
1 t
f
=
th )
"
' 0:76933. We call the trade cost
2
a2
(A 2 )
+
4b
9B
#
=
DC
1
tf
=
C
1 t
f
, at which
:
this section, we consider a case where the trade cost is speci…c. We will show an ad valorem case in our
future research. We will also consider the case where …rm 1 incurs trade costs even in FDI cases.
1 0 In our future research, we will discuss a relationship between trade costs and decision lights with anecdotal
evidence.
12
And we obtain
' 0:027934. When
<
, the relationship between the optimal entry mode
and the foreign country’s tax rate is the same as the case of
When
>
, the relationship is the same as the case of
= 0 (in the previous section).
= 0:05.
We can consider the possibility that …rm 1 has a domestic subsidiary and delegates an output
decision to the subsidiary and the subsidiary competes with the local rival. If F = 0, the net pro…t
of …rm 1 with this domestic decentralization overcomes the pro…t in Export case. However, even
when we replace the normal Export pro…t by this decentralized export pro…t, the replacement
only makes the critical trade cost rate higher (
0
11
do not change .
6
' 0:054932 >
' 0:027934) and our results
Transfer pricing cost
We examine the e¤ect of u on the optimal entry mode. It could be considered that the higher
u is related to the tougher transfer pricing regulation. Before that, we consider how u a¤ect the
transfer price. We …nd that an increase in u a¤ects the transfer prices in both cases as follows,
This means
dmdc
du
dmdc
du
=
dmc
du
=
4 2tf
Q 0 when mdc R 0 and
2 (tf
dmc
du
9B
mdc ;
3th + 1 + 49 Bu
3B (1 tf )
mc :
2
th ) + 3Bu (1 tf )
Q 0 when mc R 0. Figure 5 shows the transfer prices
of decentralized MNEs and centralized MNEs using our parameters. The solid lines show the
case of u = 1 and the dotted lines the case of u = 1:2. The red thin lines show the decentralized
cases and the blue bold the centralized case. A change of u has no e¤ect on the transfer price of
decentralized MNEs when tf = 0:475, that is, when mdc = 0 and no e¤ect on that of centralized
MNEs when tf = 0:3, that is, when mc = 0.
1 1 We
do not use any transfer pricing cost in the domestic decentralization case.
13
TP
0.14
0.12
0.10
0.08
0.06
0.04
0.02
0.00
-0.02
0.1
0.2
0.3
0.4
0.5
0.6
0.7
t_f
-0.04
-0.06
-0.08
-0.10
-0.12
-0.14
Figure 5. E¤ect of u on transfer price
DC
1
We …nd that an increase in u decreases
d
DC
1
du
d C
1
du
=
=
(4tf
2 (8tf
1 2
A (tf
2
3th
and
2
12th + 9Bu + 4)
2
as follows,
2
1) A2
th ) (1
C
1
2
tf )
0;
2 (tf
2 (tf
2
th ) + 5Bu (1
2
th ) + 3Bu (1
tf )
3
0:
tf )
Although a tougher regulation (an increase in u) reduces the FDI pro…t in both cases, when
tf =
1+3th
4 ,
because decentralized MNEs always set m = 0, a change of u has no e¤ect on the
pro…t of the …rm and when tf = th , because centralized MNEs always set m = 0, a change of u
has no e¤ect on the pro…t of the …rm.
14
Profit
0.80
0.79
0.78
0.77
0.76
0.75
0.74
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
t_f
Figure 6. E¤ect of u on pro…t
When a = 2, b = 1, A = B = 1, F = 0, u = 1:2, th = 0:3, at tf ' 0:38267,
C
1 t
f
' 0:76912. The critical trade cost rate is
' 0:028649 >
DC
1
tf
=
. This implies that a
tougher regulation (an increase in u) may expand the region of no centralization.
7
Conclusion
We investigated FDI incentives, taking into account the strategic role of transfer pricing and
decentralization. It has been found that, even when a host country has neither production cost
advantage nor tax advantage, FDI can be undertaken by a (purely) strategic incentive. We also
found that trade liberalization and a tougher transfer pricing regulation boost decentralization
of multinationals.
For checking and enforcing robustness, we should examine how and to what extent an ad valorem trade cost a¤ect our …ndings and also consider the case where …rm 1 incurs trade costs even
in FDI cases. We leave those investigations to future work. In the model, we assume a quadratic
form of transfer pricing cost. This cost implicitly re‡ects transfer pricing regulation implemented
by governments. Investigating more speci…cally the e¤ects of transfer pricing regulation on FDI
incentives is also left to future work.
15
References
[1] Amerighi, Oscar, Susana Peralta, 2010, “The Proximity-concentration Trade-o¤ with Pro…t
Shifting?,” Journal of Urban Economics 68, 90-101.
[2] Göx, Robert F., Ulf Schiller, 2007, “An Economics Perspective on Transfer Pricing,” In:
Chapman, Christopher S., Anthony G. Hopwood, and Michael D. Shields (Eds), Handbook of
Management Accounting Research, 673-695, Elsevier.
[3] Horstmann, Ignatius J. and James R. Markusen, 1992, “Endogenous Market Structures in
International Trade (natura facit saltum),” Journal of International Economics 32, 109-129.
[4] Motta, Massimo, 1992, “Multinational Firms and the Tari¤-jumping Argument,” European
Economics Review 36, 1557-1571.
[5] Nielsen, Søren Bo, 2014, “Transfer Pricing: Roles and Regimes,”CESifo Working Paper Series
No. 4694.
[6] Nielsen, Søren Bo, Pascalis Raimondos-Møller and Guttorm Schjelderup, 2008, “Taxes and
Decision Rights in Multinationals,” Journal of Public Economic Theory 10, 245-258.
[7] Porter, Lynda A, 2012, “Asymmetric Oligopoly and Foreign Direct Investment: Implications
for Host-Country Tax-Setiting,” International Economic Journal 26, 229-246.
[8] Schejelderup, Guttorms and Lars Sørgard, 1997, “Transfer Pricing as a Strategic Device for
Decentralized Multinationals,” International Tax and Public Finance 4, 277-290.
16