- 1 45:“ 1&1- A
‘ ' ' u
x -'

_"

. I ‘

V I in?

‘ ‘I

4‘". ‘_.~ ’1‘

Ln:- ‘x‘xf’fi

I’kﬁfo'gm
r

_~1r' . 1‘1““
+35%]Jagﬁhs “32541.. a ‘ . . -.,,. ‘ .
.- .-~Jt'...'u1:~§;,-":., ‘ ‘ _ A vy - y ' ".17 .- _
a , I v ‘2‘ 9"}1" I “4;; _ --' . ‘ a, I
I ‘ v . .
' "' {-3. - 33..
at. 35%%

.
.
r.

..- .‘;

. ‘3 ,‘ ' 1 y y . ".‘4

.L-w - - _ . , (l - A ‘ a

FJ’fu ' ‘ 5 J. ‘A 1‘ '
'.

—4-—-—. ‘J—L- -—

|

I. «11* SEX». .'
1‘. 3.. 3,

I,

.
‘ "H I '
- I
-. .l T.
. , J.

M
l

v

u’, v 4 A;
y. . . V .
o J . ‘

_ I
'."'
It!
1:4
2‘

m._.<

9435""

V .4
..-.-.< r. _

.‘ .v

.A...Ac—V_ fg'

.
T . _ w
$9.2" 3f. .'

ﬁr

Yaw

. I
.' , -."_'1' ﬁg" ‘

assg‘i’ 23{='zgj§ilr-L'-3m,. ,,
1 Egg“ EQN;'¢“’¢">§”"-!i$ 'jiv
." '0 y ",I' ~ 1" [{v‘
. .ﬁ . . c V Es”
Mam ! '
7! mm.“ .
a” .h'

.._.2
“ﬁzz;

.i-‘h
., If?"
.'L~ ‘ 23..

“.533
3:4,? ’ uri
.. u

-&

i‘ 1.1:
J.
.51":

i
1 .

.92. was
QU‘OC‘I

...~.,;-.

0v ,
,‘1

,.
~¢vs ‘
, .
“gnu.
v
var

.
.
a!

‘ 3'3."
.’-—’.J
z. .
. .A V
' h

.-..:—. ‘.

K .

 

thias

Ifllllllllll

 

Illlllllllllllllllllllllllllllllllllll

3 1293 017076

This is to certify that the

dissertation entitled

TAX INCENTIVES AND THE COST OF CAPITAL
ON FOREIGN DIRECT INVESTMENT:
A PERSPECTIVE OF THE U.S. SUBSIDIARY IN KOREA

presented by

Younghoon Sea

has been accepted towards fulﬁllment
of the requirements for

Ph . D . degree in Economiea.

aka KW

Major professor

 

Date Ifﬂlcyf /C} ?X

MS U is an Afﬁrmative Action/Equal Opportunity Institution 0- 12771

 

LIBRARY '

Michigan State
University

 

 

 

PLACE IN RETURN BOX
to remove this checkout from your record.
to AVOID FINES return on or before date due.

 

DATE DUE DATE DUE DATE DUE

 

JUL ti. 11 “£334
‘ c i u-

A CI “1
Val"

FEB 0 5 2000

 

 

 

 

 

 

 

 

 

 

 

 

 

ma W969.“

TAX INCENTIVES AND THE COST OF CAPITAL
ON FOREIGN DIRECT INVESTMENT:

A PERSPECTIVE OF THE U.S. SUBSIDIARY IN KOREA

By

Younghoon Seo

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Economics

1998

ABSTRACT

TAX INCENTIVES AND THE COST OF CAPITAL
ON FOREIGN DIRECT INVESTMENT:
A PERSPECTIVE OF THE U.S. SUBSIDIARY IN KOREA

Younghoon Seo

This paper develops the transnational cost-of-capital model for the U.S.
subsidiary operating in Korea, taking into account various transnational ﬁnancial
methods and the international taxation. It also considers all tax beneﬁts prevailing in
Korea, with a particular emphasis on tax holidays. Furthermore, I incorporate the real
data on the exchange rate and interest rates, and different tax bases into the cost-of-
capital model, for more adequate calculation.

Our cost-of-capital calculations show that the cost of capital is very sensitive
to price changes, interest rates, and the exchange rate. In addition, the cost of capital
varies signiﬁcantly to different sources of ﬁnancing. The U. S. subsidiary and
domestic ﬁrm have a cost advantage over the Korean domestic ﬁrm. Most authors
conclude that equity ﬁnancing is more costly than any other ﬁnancing. However, this
is not the case for the U.S. subsidiary operating in Korea. The result shows that the

cost of capital for the U.S. subsidiary ﬁnanced by new-share issues is not much higher

(and is sometimes even lower) than those ﬁnanced by retained earnings, local
borrowing, and an intercompany loan.

I extend the cost-of-capital model to capture tax savings from tax holidays
correctly. I ﬁnd that the cost of capital under the tax holiday varies signiﬁcantly to
different ﬁnancing methods. While the beneﬁt of the tax holiday for the investment
ﬁnanced by retained earnings or an intercompany loan is very substantial, the
investment ﬁnanced by new shares is penalized under the tax holiday.

Finally, my econometric results indicate that the Korean tax policy has an effect
on total FDI and Japanese direct investment, but not on U.S. direct investment in
Korea. The only signiﬁcant factors in determining the U.S. direct investment are the
government regulation and tariffs. The tariff variable work well in regression models,
as a proxy of the degree of openness on trade. This estimation result implies that
capital-exporting countries, including the U.S. and Japan, have invested in Korea for
the purpose of supplying the Korean market, and importing raw materials from third

countries and exporting their products to third countries.

Dedicated to

My father, Seong-Ryul Seo
My mother, J eong-Ok Kim

iv

ACKNOWLEDGMENTS

I would like to express appreciation to all committee members, Charles
Ballard, Steven Matusz, and Leslie Papke. Their suggestions and guidance were
invaluable in completing my dissertation. I am particularly grateful to Charles Ballard,
my dissertation chair person, for his thoughtful and extensive comments on my
dissertation. Without his patience, support and encouragement, the completion of my
dissertation would have been impossible.

Friends at the university, particularly, Hee-Chan Lee, Soo-Jong Kim, and Su-
Ryeol Ryu deserve my thanks for ﬁiendship and discussion on economic topics.

I am especially grateﬁil to my wife, Hyekyung Na, for her unconditional love
and understanding over the years. My parents-in-law have been consistently gracious
and helpﬁil for the completion of my dissertation.

Finally, I would like to express my best thanks to my mother, who is sick in
bed, for her invaluable support and prayer. Without my father and mother’s

encouragement and countless sacriﬁces, I could not have ﬁnished my dissertation.

TABLE OF CONTENTS

LIST OF TABLES ........................................................................... viii
LIST OF FIGURES ............................................................................. x
CHAPTER 1 INTRODUCTION ......................................................... 1
1-1. Literature Review for the Domestic Cost-of-Capital Model ................. 2
1-2. Literature Review for the Cost-of-Capital Model Including Risk .......... 6
1-3. Literature Review for the Transnational Cost-of-Capital Model ............ 9
CHAPTER 2 FOREIGN DIRECT INVESTMENT IN KOREA .................. 17
2-1. Overview of the Tax Policy and the Position of FDI in Korea ............ 17
2-2. The Structure of FDI in Korea ................................................ 21
2-3. Tax Incentive Provisions in Korea ............................................ 25
2-3-1 Investment Tax Credits ............................................... 25
2-3 -2. Income Deduction from New-Share Issues ....................... 26
2-3-3. Special Depreciation .................................................. 28
2-3-4. Immediate Expensing ................................................. 29
2-3-5. Tax-Free Reserves .................................................... 30
2-3-6. Tax Holiday ........................................................... 31
CHAPTER 3 THE DESCRIPTION OF THE MODEL ............................ 33
3-1. Post-Tax Rate of Return ........................................................ 35
3-2. Transnational Cost of Capital .................................................. 40
CHAPTER 4 THE MODEL WITH TAX HOLIDAYS ........................... 66
4-1. Transnational Cost-of-Capita] Model with Tax Holidays .................. 67
4-2. Effective Transnational Cost-of-Capital Model with Tax Holidays ....... 76
CHAPTER 5 SOURCES OF DATA .................................................... 83
5-1. Data Sources for Korean Firms ............................................... 83
5-1-1. Corporate Tax Rates ................................................ 83
5-1-2. Personal Tax Rates .................................................. 84
5-1-3. Depreciation Allowances ........................................... 89
5-1-4. Parameters Associated with Price and Financial Data .......... 90
5-2. Data Sources for U.S. Parent Firms ......................................... 92
5-2-1. Corporate Tax Rates in U.S. ...................................... 92
5-2-2. Personal Tax Rates in the U.S. ................................... 92
5-2-3. Depreciation Allowances
and Investment Tax Credits in the U.S. ......................... 94

TABLE OF CONTENTS (cont’d)

5-2-4. Parameters Associated
with Price and Financial Data in the U.S. ....................... 97
5-2-5. Data Sources for Transnational Investment ....................... 97

CHAPTER 6 CALCULATIONS OF EFFECTIVE TAX RATE ‘

AND COST OF CAPITAL ...................................... 98
6-1. Comparison of Effective Tax Rates in Korea and the United States ...... 98
6-2. Comparison of Costs of Capital in Korea and the United States ......... 107

CHAPTER 7 DETERMINANTS OF FOREIGN DIRECT

INVESTMENT IN KOREA .................................... 138

7-1. Empirical Model and Results ................................................ 142
CHAPTER 8 CONCLUSIONS .......................................................... 163
APPENDIX .................................................................................... 166
REFERENCES ................................................................................ 174

vii

LIST OF TABLES

Table 1. Difference between My Model and Those of Other Authors .................. 16
Table 2. Foreign Capital Inﬂows and Major Economic Indicators ....................... 19
Table 3. Composition of FDI in Korea by Sector .......................................... 23
Table 4. Composition of FDI in Korea by Investing Country ............................. 24
Table 5. Income Deduction from New-Share Issues ....................................... 27
Table 6. Different Notations between King and Fullerton and Seo ....................... 36
Table 7. Required Rates of Return under Different Financing Policies ................... 65
Table 8. Required Rate of Return under Tax Holidays .................................... 71
Table 9. Corporate Tax Rates in Korea ..................................................... 86
Table 10. Personal Tax Rates in Korea ...................................................... 88
Table 11. Tax and Price Parameters for the United States ................................ 93
Table 12. Depreciation Method Rules in the United States ............................... 95
Table 13. Estimation Results for Total FDI in Korea .................................... 148
Table 14. Estimation Results for Manufacturing of Total FDI in Korea ............... 149
Table 15. Estimation Results for Service of Total FDI in Korea ........................ 151
Table 16. Correlation Matrix of ETR ....................................................... 153
Table 17. Effects of Corporation Tax Rate and Tax Incentives ........................ 155
Table 18. Estimated Results for U.S. Direct Investment in Korea ...................... 159
Table 19. Estimated Results for Japanese Direct Investment ............................ 160

viii

LIST OF TABLES (cont’d)

Table 20. Classiﬁcation of Industry ......................................................... 167
Table 21. Economic Depreciation Rates in the U.S. and Korea by Asset .............. 168
Table 22. Weights of Financing Methods ................................................... 169
Table 23. Interest Rates in the U.S. by Industry .......................................... 170
Table 24. Interest Rates in Korea by Industry ............................................. 172
Table 25. Data Used in the Regression Model ............................................. 174

ix

LIST OF FIGURES

Figure 1. Cost of Capital for a New Investment under Tax Holidays .................... 77
Figure 2. Capital Accumulation under Tax Holidays ...................................... 78

Figure 3. Cost of Capital as the Ratio of Investment to Taxable Income Varies

in Each Financing Method .......................................... 166
Figure 4. Effective Tax Rates in Korea by Asset ......................................... 102
Figure 5. Effective Tax Rates in Korea by Major Industry .............................. 103
Figure 6. Effective Tax Rates in the U.S. by Asset ....................................... 104
Figure 7. Effective Tax Rates in the U.S. by Major Industry ............................ 105
Figure 8. Coefficient Variation of ETRs in the U.S. and Korea ......................... 106
Figure 9. % Change of Price in the United States and Korea ............................ 117
Figure 10. Weighted Average of Interest Rates in the U.S. and Korea ................ 118
Figure 11. Trend of the Exchange Rate .................................................... 1 19
Figure 12. Comparison of the Weighted Average of Costs of Capital ............... 120
Figure 13. Comparison of Costs of Capital Financed by Retained Earnings ......... 121

Figure 14. Comparison of Costs of Capital Financed

by Debt or Intercompany Loan ........................................... 122
Figure 15. Comparison of Costs of Capital Financed by New Share Issues .......... 123
Figure 16. Sensitivity of the Cost of Capital in a Deﬁcient-Credit Position

according to the Length of Deferral ..................................... 124

Figure 17

Figure 18
Figure 19
Figure 20
Figure 21
Figure 22
Figure 23
Figure 24
Figure 25
Figure 26

Figure 27

Figure 28

Figure 29

LIST OF FIGURES (cont’d)

. Sensitivity of the Cost of Capital in an Excess-Credit Position

according to the Length of Deferral ..................................... 125
. Costs of Capital in Korea by Major Industry ................................ 126
. Costs of Capital in Korea for Manufacturing I ............................... 127
. Costs of Capital in Korea for Manufacturing H .............................. 128
. Costs of Capital in the U.S. by Major Industry ............................... 129
. Costs of Capital in the U.S. for Manufacturing I ............................. 130
. Costs of Capital in the U.S. for Manufacturing H ............................ 131
. Costs of Capital for the U.S. Subsidiary by Major Industry ............... 132
. Costs of Capital for the U.S. Subsidiary in Manufacturing I ................ 133
. Costs of Capital for the U.S. Subsidiary in Manufacturing 11 ............... 134

. Costs of Capital for the U.S. Subsidiary

with Tax Holidays by Major Industry .................................... 135
. Costs of Capital for the U.S. Subsidiary

with Tax Holidays in Manufacturing I ..................................... 136
. Costs of Capital for the U.S. Subsidiary

with Tax Holidays in Manufacturing 11 ................................... 137

CHAPTER 1

INTRODUCTION

The Korean economy is very sensitive to changes in the international economic
environment, due to its heavy dependence on the international economy. As a result
of the establishment of World Trade Organization (WTO), the world economy has
been more liberalized and international economic integration has been accelerated.

The recent ﬁnancial crisis in Korea makes ﬂotation of Korea’s currency, the
won, sharply depreciated against the dollar. The Korean government has borrowed
emergency ﬁrnds from International Monetary Fund (IMF) to cope with this crisis.

For rapid stabilization of the economy under the IMF auspices, deregulation and the
openings of capital markets will be accelerated in Korea. The ceiling on aggregate
ownership of foreign investment in the Korean equity market will be increased from 26
percent in 1996 to 55 percent by the end of 1998.

The removal of restrictions on international capital flows and the globalization
of ﬁnancial markets have focused attention on the role of Foreign Direct Investment
(FDI) by multinational corporations. The analysis of FDI in Korea is closely related to

tax policy. Tax policy influences investment decisions through its effect on the cost of

capital and the returns to different activities. The tax system also affects the
investment decisions of multinational ﬁrms, through a complicated interaction of
home- and host-country taxation. Since the cost of capital plays a primary role in
determining the investment decisions of ﬁrms, and since taxation is one of the most
important factors in measuring the cost of capital, it is important to understand and
compare how the tax systems may distort the investment decisions of ﬁrms in the two
countries.

This paper develops a transnational cost-of-capital model, taking into account
various ﬁnancial methods and international taxation. It also consider all tax beneﬁts
prevailing in Korea, with a particular emphasis on tax holidays, which are one of the
popular incentives to foreign investments. By comparing the costs of capital among
Korean ﬁrms operating in Korea, U.S. ﬁrms operating in the US, and U.S.
subsidiaries with or without tax holidays in Korea, I verify which country of these two
countries provide greater tax incentives and whether tax incentives in Korea have an
important role in lowering the U.S. subsidiaries’ cost of capital. While estimating the
transnational cost of capital, this paper examines the empirical effects of the cost of

capital on foreign direct investment by using econometric methods.

1-1. Literature Review for the Domestic Cost-of-Capital Model

In the neoclassical theory of investment, Jorgenson (1963) presents a theory of
investment behavior analyzing the effect of taxation on the user cost of capital. The
cost of capital deﬁned by Jorgenson (1963) is the user cost of capital in production,

equal to the Shadow price or implicit rental of one unit of capital service. In other

words, the user cost of capital is the opportunity cost to a ﬁrm, as a consequence of
owning a unit of capital. If there is no tax, the cost of capital is equal to the real rate
of return plus the rate of depreciation. Taxes make the cost of capital more
complicated to calculate.

Hall and Jorgenson (1967) study the effects of changes in tax policy on
investment for major tax reform, using the cost-of-capital model. They ﬁnd that the
effects on investment of the adoption of accelerated depreciation methods in 1954
were very signiﬁcant. The effects of the reduction of lifetimes used for calculating
depreciation on equipment and machinery in 1962 were also signiﬁcant. Finally, the
effects on investment behavior of the investment tax credit (ITC) of 1962 were very
substantial.

Auerbach (1979, 1983), and King and Fullerton (1984) have attempted to
incorporate personal capital income taxes and the ﬁrm’s ﬁnancial behavior into the
cost-of-capital model. Auerbach (1979) adopts the wealth—maximization approach to
derive the cost of capital, and reviews the investment and ﬁnancial behavior of ﬁrms in
the presence of differential taxation of dividends and capital gains. He shows that the
cost of capital does not depend on either the dividend payout rate or the rate of
dividend taxation, on the assumption that there is no new share issues. Auerbach
(1983) also studies the effects of taxation on the incentive to invest via the cost of
capital, taking into account the role of inﬂation, adjustment costs of capital stock,
uncertainty related to the expected rate of return and the effects of bankruptcy, and
tax-law asymmetries with respect to capital gains and losses. His discussions give a

very insightful guide for future studies. However, Auerbach’s wealth-maximization

approach ignores the role of new share issues to ﬁnance a new investment. His
“trapped-equity” model cannot explain share repurchases. Bagwell and Shoven (1989)
indicate that share repurchases have been extremely large in some years (especially in
1987). They also argue that the cost of capital ﬁnanced by retained earnings must be
adjusted upwards if more efﬁcient tax means of cash transmission1 are allowed.

King and Fullerton use the cost-of-capital model to compute effective tax rates
(ETR) for four countries: the United Kingdom, Sweden, West Germany, and the
United States. They derive the cost of capital, by equating the present discounted
value of the proﬁts of the project with the cost of the project. In their approach, the
gross marginal rate of return (MRR) is constant over time, and it is the same as the
cost of capital. King and Fullerton examine the variations in ETRs for different assets,
sources of ﬁnance, and ownership categories in each country. They ﬁnd that the
weighted-average ETR for 81 hypothetical projects is the lowest in the United
Kingdom, followed by Sweden, the United States, and Germany. This is because the
United Kingdom allows loo-percent expensing for machinery, and because the United
Kingdom has the highest share of machinery in the capital stock. The Swedish
government provides 30-percent expensing for machinery, and an 11-percent
investment tax credit (ITC). The effective tax rate depends on the special incentive
policy and the weights of the capital stock.

King and Fullerton also ﬁnd that the ETR is strikingly low for debt-ﬁnanced

investments, when compared to other sources of ﬁnance, in all four countries. This

 

I Annual cash distributions, such as share repurchases and cash via merge and acquisition, are subject
to small tax burdens. See Bagwell and Shoven (1989).

‘14

b the.

reﬂects the fact that interest payments are deducted from the corporate tax base. In
addition, they ﬁnd that the rate of inﬂation affects the ETRS in four countries, through
the real value of depreciation allowances, nominal interest rate, and the nominal value
of inventories.2

Tajika andSYui (1988) compare the effects of taxation on the cost of capital
and the effective tax rate between manufacturing in the United States and Japan. Their
analysis incorporates tax-free reserves,3 which is the typical incentive policy in Japan.
They ﬁnd that under constant prices the effective tax rate in the U. S. manufacturing is
lower than that in comparable Japanese manufacturing from 1972 to 1984. This result
can be explained as follows. The effects of the tax-free reserves, which are only an
active incentive policy in Japan, are relatively small when compared to the effects of
the ITC and the Accelerated Cost Recovery System (ACRS) in the U.S.

However, unlike the results under constant prices, Tajika and Yui Show that,
under actual prices, the difference between the costs of capital in the United States and
Japan ﬂuctuated over time. The effects of tax incentives on the cost of capital are
oﬂ‘set by the high rate of return on equity and the high interest rates in the United
States. Therefore, the cost of capital in the United States is sometimes higher and
sometimes lower than that in Japan. They also ﬁnd that the price changes in Japan

seem to be more sensitive to external shocks (e. 3., oil shocks).

 

2 The effects of inﬂation on the ETRs operate in different directions and with different scale in each
country. See more details in King and Fullerton, pp. 273 -275.

3 Corporations can defer their tax liabilities on the amount of certain reserves until they are added
back to their taxable income. The tax-free reserves will certainly reduce the corporation tax burden
by putting off the liabilities.

1-2. Literature Review for the Cost-of-Capital Model Including Risk

All of the analyses discussed so far have assumed a certainty framework.
However, we may have different results if we consider uncertainty regarding the
ﬂuctuation of the ﬁrm’s output and the price of capital goods. Fullerton and Gordon
(1983) try to derive improved ETRs by using more disaggregated and extensive data,
and by putting risk into the cost-of-capital framework. In the model of Fullerton and
Gordon, the corporate tax provides a form of insurance to ﬁrms, because it shares in
both the income and the losses of the ﬁrm. They simulate the effects of integration of
the corporate tax with personal taxes. They ﬁnd that the tax distortion caused by
raising labor income taxes, in order to restore government revenue, would be greater
than the tax efficiency gains through corporate tax integration. This is because the
government would no longer share the risk under the method of corporate tax
integration.

However, Bulow and Summers (1984) point to a serious ﬂaw in the augment
of Fullerton and Gordon. Bulow and Summers argue that it is very important to
distinguish between the two types of risk, which are income risk and capital risk, in
the calculation of the cost of capital. With a ﬁill-loss-offsetting tax system, the
government shares equally in both the proﬁts and losses of the company. The cost of
bearing income risk is thus fully deducted in a full loss offset tax system, and no
additional distortions are introduced for income-risky investments. However, Bulow
and Summers also point out that tax depreciation allowances in the United States are
based on the acquisition price of the capital goods, and capital gains and losses are not

indexed for inﬂation. This means that depreciation allowances and capital gains taxes

do not ﬂuctuate with unexpected changes in the replacement value of capital. The
implication is that the tax system does not deduct the full cost of bearing capital risk.
Therefore, an increase in capital risk can be thought of as increasing economic
depreciation, thereby raises the cost of capital.

The ETR for a capital-risky investment may be determined by replacing the
rate of economic depreciation with the rate of economic depreciation plus a risk
premium. Bernheim and Shoven (1989) take risk premia into account, and compare
the costs of capital in the United States and Japan for autos and plants. Bernheim and
Shoven make an explicit distinction between income risk and capital risk. They also
emphasize the fact that the fraction of earnings that is paid out as dividends is a much
higher fraction out of long-terrn earnings (50 percent) than of transitory earnings (2
percent), and they incorporate this into the cost-of-capital formula. They ﬁnd that the
higher cost of capital in the U.S. was attributed to the higher expected rate of inﬂation
and the higher required rate of return on equity. Their empirical results Show that risk
premia are a very decisive component of the cost of capital.

Most of the authors to estimate the cost of capital considering risks have used
the eamings—price ratio as a proxy of the expected rate of return on the market
portfolio. But there are some problems with using the earnings-price ratio to describe
the ﬁrm’s required return. One problem is that the earnings-price ratio is equal to the
current required rate of return only if the required rate of return is constant over time.
However, the required rate of return in the equity market may ﬂuctuate. A second
problem is that the observed earnings-price ratios reﬂect the stock market’s

expectation of ﬁiture corporate growth, rather than the ﬁrm’s required return on

equity. A third problem is that there is little or no evidence on the risk premium that
equity investors require to hold a risky asset (see Poterba 1991). Fourth, because both
Bulow and Summers, and Bernheim and Shoven divide assets only into a risky asset
and a risk-free asset, the risk premium, if it exists, is the same across structures,
machinery, or even inventories. This ignores the fact that the scale of risk is different
across assets (such as machinery, equipment, structures, and inventory). A ﬁnal
problem is the assumptions of the Capital Asset Pricing Model (CAPM) (see Sharpe
(1970)) which summarizes the ﬁnancial market’s required rate of return on securities
of different riskiness. Under the assumptions of the CAPM, the required rate of return
on all securities follows the returns on the capital-market line.

Obviously, many factors affect the cost of capital including tax incentives,
interest rate, and risk. However, many studies on the cost of capital have used only
some of these factors, because it is difficult to incorporate all of these factors into an
analytical framework. Many researchers underestimate or overestimate the real cost of
capital, by ignoring industry-speciﬁc incentives and accounting rules. The Korean
government frequently provides speciﬁc industries with special tax incentives. The
analysis on different tax bases between countries is helpﬁrl to measure the cost of
capital correctly. In my study, I incorporate various tax incentives, ﬁnancial methods,
and different accounting rules between countries into an analytical framework.
However, I ignore risk factor, because it is difficult to get data on the expected rate of

return for U.S. subsidiaries listed on the Korean Stock Exchange (KSE).

1-3. Literature Review for the Transnational Cost-of-Capital Model

The analyses mentioned above are limited to domestic investment and taxation.
Since the taxation of foreign source income depends on the international tax systems
that commonly exist to reduce double taxation, the analysis of international tax
systems is important in calculating the transnational cost of capital. Some countries
(e.g., France and the Netherlands) practice the territorial system, under which the
home country does not tax its own businesses on income generated outside its borders.
Other countries, such as the United States, Canada, and Japan, use the global or
residence system. These countries impose taxes on investment income if their
residents earned abroad, but allow a credit for taxes paid to the foreign country. The
United States allows a credit up to 34 percent. This credit is given in order to avoid
double taxation. The United States also practices deferral of income tax of
subsidiaries of U.S. corporations until their foreign-source incomes are repatriated, but
branches of U.S. corporations are taxed on accrual (i.e., there is no deferral for
branches).

If the home country’s tax is greater than the foreign tax credit, the subsidiary
has to pay the difference between the home country tax and the foreign tax credit only
when its proﬁts are repatriated. At this time, we say that the subsidiary is in a deﬁcient
foreign-tax-credit position. On the contrary, if the home country’s tax is less than the
foreign tax credit, the subsidiary pays nothing to the home country, and the home
country does not reﬁJnd foreign taxes in the excess of the domestic tax rate. We say

that the subsidiary is in a excess foreign-tax-credit position at this time.

The choice between the global and the territorial system affects tax revenues
and the decision of ﬁrm’s investment. Suppose that companies for the U.S. and
France operate in Korea, and that the corporate tax rate is 35 percent in both the U.S.
and France, and 30 percent in Korea. A French company under territorial system pays
a 30-percent tax on its proﬁts to the Korean government, and does not pay tax to the
home country. In contrast, the U.S. company under global system pays a 30-percent
tax on its proﬁts to the Korean government, and also pays 5-percent tax on its proﬁts
to the U.S. government. Because the U.S. company pays more tax for its foreign-
source income, the U.S. company may be at a cost disadvantage. The global system
may distort the company’s decision of investment more than the territorial system
does. On the other hand, the French government does not collect any tax from
remitted proﬁts, but the U.S. raises tax revenue by 5 percent of remitted proﬁts. Thus,
a country that is a capital exporter may collect more revenue under the global system
than the territorial system.

Alworth (1988) extends the cost-of-capital model for investments made by
multinationals. He speciﬁes the tax parameters to take account of different methods of
double-taxation relief, and reviews the ﬁnancial policies of the UK. multinational ﬁrm
in a King-Fullerton framework. He ﬁnds that UK. subsidiaries operating in Germany
have an excess-credit position, which means that they fully use the foreign tax credit
and they face a higher effective corporate tax rate than that faced by domestic ﬁrms in
the UK. This is due to the tax rate in Germany, which is the highest among the
European countries. The ETRs of UK. subsidiaries in Germany were largely

inﬂuenced by the German tax system. Alworth also ﬁnds that the overall ETRs are

10

positively correlated with the rate of inﬂation. This result is very consistent with that
of King-Fullerton (1984).

Yun (1992) provides a framework for the analysis of the impact of taxation of
income from foreign capital on a small capital-importing country, Korea, and estimates
the ETRS of foreign investments, incorporating the various tax incentives prevailing in
Korea. His estimated results Show that the ETR on foreign investment is negative,
due to the generous tax treatments (e.g., accelerated depreciation for tax purpose and
tax holiday) of foreign investments. But Yun’s result on the ETR on foreign capital is
underestimated, because he fails to incorporate the home-country tax on the
repatriation income from foreign subsidiaries in calculating ETR, and he also omits the
double—taxation relief offered in the home country.

Leechor and Mintz (1993) derive the user cost of capital for a subsidiary when
the deferral method is used by the capital-exporting country. They show that both the
home- and host-country tax systems inﬂuence the cost of capital for a subsidiary.
They also ﬁnd that the Hartman (1984) result, implying that the corporate tax of the
home country has no effect on the repatriation of dividends if the subsidiaries ﬁnance
their investments by retained earnings, holds only if the home and host countries have
the same corporate tax base. His model neglects to incorporate equity transfers from
the parents to the subsidiary into the cost-of-capital model, even though the
proportion of investment ﬁnanced by the parent’s new-share issues and intercompany
loans is substantial in transnational investment. This proportion of U.S. subsidiaries
operating in Korea is 37.3 percent for new-share issues, and 8.0 percent for

intercompany loans (which are average values during 1977-1994).

11

Gerard (1993) generalizes King and Fullerton’s marginal effective tax rate
model and extends it to transnational capital income. He argues that King and
Fullerton’s methods might lead to an underestimation of the true ETR, because King
and Fullerton ignore the possibility of non-price rationing phenomena.

An Organization for Economic Cooperation and Development report (OECD,
1991) provided by Devereux and Pearson, allows for international comparisons of the
cost of capital, by considering all of the main features of taxation and ﬁnancial policy
for both domestic and transnational investments. They use the King-Fullerton
techniques to compute the cost of capital of a number of international investments.
They ﬁnd that, at least for European countries, differences in the costs of capital on
international investments are very large. The average cost of capital on direct
investment abroad is the highest in Portugal (9.9%), followed by Japan (8.0%), the
U.S. (7.1%), and France (6.6%).

Jun (1995) calculates the cost of capital for inbound and outbound direct
investment in 11 major investing countries, with a particular attention to U.S. ﬁrms.
He ﬁnds that U.S. ﬁrms operating in major foreign markets, on average, face about a
20-percent higher cost of capital for the foreign investment than that for the domestic
investment in the United States. Therefore, U.S. ﬁrms operating in foreign markets
are at a cost disadvantage, comparing with local ﬁrms in foreign markets.

OECD (1991) and Jun (1995) use a S-percent hypothetical rate of interest.
Furthermore, they assume that both the home and the host country have the same tax

base. These may mislead the calculations of the cost of capital.

12

The existing studies are important advances in our understanding of the effects
of taxation, and they provide an analytical guide for future discussion of tax reform.
However, most of the studies relate to highly aggregated data, and most omit the
home-country tax system from the analysis. In these studies, the analysis of the effects
of taxes on FDI has been limited to a large open economy, and has not been applied to
a developing country. Most authors focus on either the calculation of the cost of
capital or on an econometric study. Unlike these authors, I carry out two analyses in a
consistent framework. These are the theoretical analysis (which is on the effects of tax
policy and ﬁnance methods on the cost of capital) and empirical analysis via an
econometric study (which is on the effect of tax policy on foreign investment).

In my study, I use a wealth-maximization approach (Auerbach (1979)) to
derive the cost of capital. However, Auerbach ignores the difference of tax bases,
which is incurred from various sources of ﬁnancing. This is because Auerbach use a
weighted average of the required rate of returns. Furthermore, he do not distinguish
the difference between the costs of capital for retained earnings and new-share issues.
To overcome the problems of Auerbach’s approach, I assume4 that savers provide
funds to ﬁrms that perform investments, and that the savers face the same after-tax
rate of return. Even under these assumptions, ﬁrms face the different required rate of
return for different ﬁnancing methods and industries, because I use actual interest rates
across various industries and ﬁrms are subject to different tax burdens in various

ﬁnancing methods.

 

4 This assumption comes from King and Fullerton (1984). They use a ﬁxed interest rate ( 5 percent
in their calculation), regardless of various industries. Instead of a ﬁxed rate, I use actual interest rates
for various industries. For more detail discussions on King and Fullerton’s assumption, see pp.34-3 7.

13

The transnational cost-of-capital model I set up is similar to the model used in
the OECD report (1991). The OECD study is limited to the manufacturing sector,
and ignores the industry-speciﬁc incentives and accounting rule. I extend the OECD
model to 21 industries and attempt to incorporate industry-speciﬁc incentives and
accounting rules (which are tax holidays and different tax bases between the home and
host countries). On the other hands, because the OECD report presents cost-of-
capital calculations only under a weighted average of the required rates of return for
retentions, new equity, and debt, it fails to analyze the difference among different tax
bases, which come from different sources of ﬁnancing. We cannot determine which
ﬁnancing method is at a cost advantage, from the OECD’s cost-of-capital calculation.
On the contrary with the OECD report, I calculate the costs of capital separately for
each kind of ﬁnancing, then presents a weighted average of costs of capital.

In Chapter 2, I will begin by evaluating the position of FDI in Korea, and then
analyze the national tax policies toward foreign capital in Korea. Speciﬁcally, I will
review the tax incentives for foreign capital in Korea.

In Chapter 3, I incorporate international taxation, the exchange rate, and the
various tax incentives that prevail in Korea, into the cost-of-capital model. Most
studies assume that both the home and host countries have the same tax base. But I
develop the cost-of-capital model, allowing for the tax bases to be different between
the countries. In addition to using more recent and extensive data, I will calculate the
cost of capital for more disaggregated industries.

The most popular tax incentive for foreign capital in Korea is the tax holiday.

Under the tax-holiday provision in Korea, the corporate tax rate, u, is zero for the

14

 

ﬁrst 5 years, and it is 0.511 for the subsequent 3 years. The required rate of return and
the repatriation tax rate on dividends, which depend on a varying corporate tax rate ,
also vary over time. Thus the tax holiday and international taxation make the cost-of-
capital model more complicated. In Chapter 4, I extend the cost-of-capital model by
incorporating time—variant tax parameters and international taxation. I discuss and
present the sources of data, for Korean ﬁrms, U.S. domestic ﬁrms, and U.S.
subsidiaries in Chapter 5

In Chapter 6, I examine whether the tax policy on FDI currently applied in
Korea is generous, by comparing my estimated results of the costs of capital and
effective tax rates between domestic ﬁrms and U.S. subsidiaries operating in Korea.

Finally, I use econometric methods to study the empirical effects of taxation on
FDI. In this study, FDI serves as a dependent variable, and the cost of capital is one of
the independent variables. I incorporate various nontax factors in the econometric
study, because nontax factors, such as the wage level, political risks, and exchange
risks, have important effects on affecting FDI in other empirical studies. The goal of
this study is to be very helpful in telling us how successful the tax policy has been in
stimulating foreign capital investment. Table 1 shows the difference between my

model and those of other authors.

15

 

Table 1. Difference Between My Model and Those of Other Authors

ma- ‘4“. “to- ..c . . -

 

 

 

 

 

 

 

 

Auerbach King and Chou and Wu OECD Seo
(1979) Fullerton (1994) (1991) (1998)
(1984)
1. Domestic Investment
l-l. Difference among Tax NO NO YES NO YES
Bases from Different
Sources of Financing
1.2. Difference among the NO YES NO YES YES
Requrred Rates of Return
to different Sources of
Financing
1-3. Special Incentives NO N O NO NO YES
across Industries
l-4. Disaggregation 3 27 l 21
1-5. Data Hypothetical Actual Data Hypothetical Actual
Assumption Assumption Data
2. Transnational
Investment
2-1.Transnational Sources NO NO NO A Weighted Different
of Financing Average Rate Rates
2-2. Tax Deferral NO NO NO NO YES
2-3. Different Tax Base NO NO NO NO YES

 

 

l6

 

CHAPTER 2

FOREIGN DIRECT INVESTMENT IN KOREA

2-1. Overview of the Tax Policy and the Position of F D1 in Korea

Since the launching of the First Five Year Plan in 1962, Korea has experienced
rapid growth, due to the emphasis on exports and structural adjustment. High gross-
national-saving ratios (39.3 percent of GDP in 1988) and investment ratios have
accommodated its performance. Foreign capital has also played an important role in
ﬁnancing its economic development. In 1973, foreign capital inﬂows were 8 percent of
GDP in Korea.

As Table 2 indicates, most of foreign capital at the early stage of economic
development was either government borrowing or government-guaranteed commercial
loans. Foreign Direct Investment (FDI) accounted for only a small fraction of capital
inﬂows from abroad. The fraction of FDI in GDP was below 1 percent in Korea,
except for 1973. This can be explained as follows. First, in the ﬁrst 20 years after F DI
into Korea began with the implementation of the Foreign Capital Inducement Act
(F CIA) in 1961, the Korean government generally restricted FDI, while providing tax
incentives for export-oriented investments and those deemed essential for the

development of the Korean economy. Second, the government wanted to carry out

17

investment projects directly, in order to supply the severe deﬁciency of social overhead
capital, and to improve industrial competition by accumulating technologies. Finally,
the Korean government feared that foreign ﬁrms would have dominated the Korean
economy unless domestic ﬁrms had competitive power over their foreign rivals.
Political and social factors, in addition to economic factors, played great roles
in changing the prime goal of government policy. For the ﬁrst time, Korea suffered a
negative growth rate of -4.8 percent in 1980, and FDI dropped to $256 million in 1980
from $473 million in 1979. To overcome this crisis, the government focused on a
strong economic stabilization plan. Korea’s tax policy toward trade and FDI has
gradually liberalized since 1980. In 1984, the Korean government amended FCIA
extensively, from a positive system to a negative system, in order to attract FDI. A
positive system is the policy that allows FDI only in certain investment sectors, while a
negative system is the policy that all investments not listed in restricted ﬁelds of FDI
are qualiﬁed for approval. An adoption of the negative system increased the ratio4 of
liberalization in manufacturing sector from 80% to 86% in 1984 (Korea ’5
International Direct Investment 1995). After the implementation of the FCIA
amendments in 1984, total FDI inﬂows during 1984-1988 more than doubled, relative
to 1979-1983. The dramatic increase in FDI during the latter half of the 19805 is
attributed both to the country’s booming domestic economy, which is related to the

1988 Seoul Olympics, and to improved market access for foreign enterprises.

 

4 The ratio of liberalization is deﬁned as the ratio of unrestricted ﬁelds of FDI to total ﬁelds of FDI.

18

Table 2. Foreign Capital Inﬂows and Major Economic Indicators

(“Constant millions of U.S. dollars” based on 1990 price, and “percent” )

 

 

 

Year GDP at Gross Gross Public Commercial Foreign1 FDI as a
Market Investment Saving Loans Loans FDI Capital as a "/0 of GDP
Price Ratio Ratio (A) (B) (c) % of GDP

1965 68278 15.1 7.4 114 798 137 1.54 0.20
1966 75455 21.9 12.0 1460 2200 280 5.22 0.37
1967 83707 22.3 11.6 1906 2230 198 5.18 0.24
1968 96223 26.2 15.3 1143 4377 310 6.06 0.32
1969 106538 29.2 19.1 2003 5909 187 7.60 0.18
1970 107069 27.0 17.5 1436 4582 824 6.39 0.77
1971 103819 25.4 15.5 3330 3791 473 7.31 0.46
1972 96472 21.8 15.8 3057 3075 575 6.95 0.60
1973 108640 25.6 23.6 3331 3810 1579 8.03 1.45
1974 115669 31.0 20.5 2452 3841 1032 6.34 0.89
1975 102814 28.6 18.1 2421 4066 315 6.62 0.31
1976 113551 26.5 24.2 2983 3510 356 6.03 0.31
1977 123185 28.3 27.5 2263 4416 363 5.72 0.29
1978 146310 32.5 29.9 2375 5561 294 5.62 0.20
1979 156914 35.8 28.5 2643 3830 473 4.43 0.30
1980 124339 31.9 23.2 2961 2740 256 4.79 0.21
1981 118427 29.9 22.9 2826 2085 254 4.36 0.21
1982 118147 28.9 24.4 2919 1428 202 3.85 0.17
1983 123770 29.4 27.6 2213 1441 181 3.10 0.15
1984 127477 30.6 29.9 2014 1214 273 1.17 0.21
1985 127978 30.3 29.8 1389 1308 319 2.36 0.25
1986 140877 29.2 33.7 1141 2101 619 2.74 0.44
1987 168292 30.0 37.3 1369 1923 773 2.42 0.46
1988 210652 31.1 39.3 1031 1144 1035 1.52 0.49
1989 244129 33.8 36.2 522 945 892 0.97 0.37
1990 253674 37.1 35.9 418 30 895 0.53 0.35
1991 267185 39.1 36.1 390 0 1069 0.55 0.40
1992 263644 36.8 34.9 416 128 688 0.47 0.26
1993 269628 35.2 35.2 340 0 851 0.44 0.32

 

 

 

 

 

 

 

 

 

1 Foreign capital is composed of public loans, commercial loans, and FDI. Therefore,

the % of foreign capital is (A+B+C)/GDP
Source: World Bank, World Table, various issues.
Bank of Korea, Economic Statistic Yearbook, various issues.

19

 

To liberalize further, the Korean government implemented the automatic
approval system, which exempted certain categories of F DI projects from government
screening before approval. More industries have been opened to foreign investment,
and various restrictions have been eliminated. By August, 1995, approximately 90.7
percent of all industry sectors (including 98.3 percent of manufacturing business)5 had
been removed from the negative list (see more details in Korea ’s International Direct
Investment 1995).

The new government, which was launched in 1993, made a special effort to
create a hospitable environment for FDI under “a Flag of Globalization”, by
establishing a ﬁve-year FDI liberalization Plan in November, 1995 (implemented in
1996). This plan introduced a new incentive system. Currently, eligible foreign
investors can choose either loo-percent special depreciation (which is a type of
accelerated depreciation method) or a tax holiday. Tax incentives for foreign investors
in Korea include a ﬁve-year tax holiday for corporate income, dividends, and royalties,
as well as special depreciation, immediate expensing, and a tax-free investment
reserve. As government restrictions on F DI are mitigated, the ﬂow of foreign capital
investment may become more responsive to tax incentives available in Korean

markets, and analysis of the effects of taxation on FDI may be more important.

 

5 The liberalization ratio calculated from the asset values of Korean industries should be a more
accurate measure. However, the data about the asset values for entire Korean industries is difficult to
get. I think that the weighted average of the asset values has no distinctive variation with the
reported value in the above. because the number of business in Korea are extremely large.

20

2-2. The Structure of F D1 in Korea

Table 3 shows the sectoral distribution of FDI in Korea. In the 19603, FDI in
Korea was concentrated in the light manufacturing and export industries. At that time,
foreign investors were discouraged from investing in import-substituting industries.
FDI in Korea has increased signiﬁcantly since the middle 19805. The annual average
of FDI during the period 1993-1995 is quadrupled relative to that of FDI during the
period 1979-1985. This may be due to the deregulation and increased market demand
in Korea. FDI inﬂows in the 19605 and 19705 were concentrated in a few
manufacturing sectors, since FDI in most non-manufacturing industries was restricted.
During the period 1962-1978, FDI into the manufacturing sector accounted for more
than 80 percent of total inward FDI. Among the manufacturing sectors, more than 40
percent of total FDI in Korea was accounted for by textiles and clothing, electricity
and electronics, and chemicals. Since 1980, the portions of chemicals, medicine,
machinery, electricity and electronics, and transport equipment have increased or
remained high, while those of textiles and clothing, and metals have decreased. These
changes seem to reﬂect the structural changes of Korean industry in this period, from
labor-intensive industries to capital- and technology-intensive sectors.

In the service sector, the hotel industry attracted substantial FDI in the ﬁrst half
of 19805. This was due to the preparation for the 1988 Seoul Olympics. The portion
of the service sector in total FDI increased with the increase of FDI in hotel industry.
In the 19905, service-sector restrictions on foreign investors have lessened, and the

relative importance of this sector has risen. Accordingly, FDI into the service sector

21

increased signiﬁcantly, so as to take 57.9 percent of the total F DI during the period
1993-1995.

Table 4 shows that Japan and the United States are the largest-investing
countries in Korea. The composition of FDI in Korea by investing country has
changed during the period 1962-1995. Although the U.S., Japan, and the Netherlands
remain the largest investors in Korea, recent investments from European countries

have increased, up to 25 percent of total FDI in Korea during the period 1994-1995.

22

Table 3. Composition of FDI in Korea by Sector

(“Constant millions of U.S. dollars” based on 1990 price, and “percent”)

 

 

 

 

Annual Annual Annual Annual Annual Annual
Industry Average of Average of Average of Average of Average of Average of
1962-1965 1966-1972 1973-1978 1979-1985 1986-1992 1993-1995
Agriculture, Fishery, Forestry 0.0 0.0 12.0 1.0 4.3 0.2
Mining 0.0 0.0 1.7 0.0 0.7 0.0
The share of Manufacturing (‘70) 100.0 84.7 81.9 68.9 66.1 42.1
Manufacturing 73.2 282.0 539.6 194.4 563.5 464.4
Food 0.0 1.8 1.8 13.1 40.2 18.9
Textile & Clothing 5.6 31.7 154.8 1.7 12.9 17.2
Wood, pulp & paper products 0.0 0.0 1.4 3.9 5.3 22.8
Chemicals 33.8 77.7 81.0 52.8 116.5 136.0
Chemical fertilizer 0.0 0.0 10.2 2.0 0.0 0.3
Medicine 0.0 0.0 4.3 10.5 33.1 32.1
Petroleum 28.2 0.0 38.6 4.8 58.9 18.9
Non-metallic 0.0 26.1 6.6 3.7 15.2 20.0
Metals 0.0 13.8 38.3 10.0 9.7 7.9
Machinery 0.0 9.7 45.5 14.2 62.3 51.4
Electricity 8 Electronics 0.0 71.1 95.8 46.9 122.5 84.7
Transportation equipment 0.0 34.7 13.7 28.7 78.3 46.7
Other manufacturing 5.6 15.5 47.7 2.1 8.6 7.5
The share of Service (%) 0.0 15.3 16.1 30.7 33.4 57.8
Service 0.0 50.9 105.9 86.7 284.5 637.1
Financing 8 Insurance 0.0 7.0 18.7 36.9 86.4 202.9
Construction 0.0 19.5 13.7 8.3 2.0 5.0
Electric, gas 0.0 21.5 3.5 0.0 0.0 6.7
Transportation 8 Storage 0.0 1.3 3.3 6.0 1.3 3.1
Wholesale 8. Retail trade 0.0 0.0 0.0 2.9 2.8 57.1
Hotels 0.0 0.0 66.8 23.7 156.0 156.8
Restaurant 0.0 0.0 0.0 0.0 1.9 12.8
Trading 0.0 0.0 0.0 0.0 21.9 77.4
Other Services 0.0 1.6 0.0 8.9 12.2 115.4
Total 73.2 332.9 659.2 282.1 853.0 1101.8 '

 

 

 

 

 

 

 

Source: Ministry of Finance and Economy, Fiscal and Financial Statistics, various issues.
The Korea Development Bank & Ministry of Finance (1993), Foreign Capital
Inducement in Korea for 3 Oyears.

23

 

Table 4. Composition of FDI in Korea by Investing Country

(Constant millions of U.S. dollars based on 1990 price)

 

 

 

Year U.S. Japan Hong Kong Netherlands Germany UK France Others Total 96 US 96 Japan
1962-1965 68 90 0 0 0 0 0 135 293 23.1 30.8
1966 260 20 0 0 0 0 0 0 280 92.9 7.1
1967 180 18 0 0 0 0 0 0 198 90.9 9.1
1968 180 33 0 0 16 0 0 82 310 57.9 10.5
1969 86 72 0 0 0 0 0 29 187 46.2 38.5
1970 587 162 0 62 0 0 0 12 824 71.2 19.7
1971 187 198 0 11 0 0 0 77 473 39.5 41.9
1972 330 1 98 0 1 9 0 0 0 28 575 57.4 34.4
1973 99 1430 0 0 8 0 8 33 1579 6.3 90.6
1974 127 834 0 0 6 0 0 64 1032 12.3 80.9
1975 56 162 5 5 5 0 15 66 315 17.7 51.6
1976 117 205 4 8 4 13 0 4 356 32.9 57.6
1977 43 189 4 64 7 11 11 36 363 11.8 52.0
1978 44 131 12 52 0 12 9 35 294 14.9 44.6
1979 56 303 10 0 17 17 2 68 473 11.8 64.1
1980 139 63 2 4 12 2 2 33 256 54.2 24.4
1981 119 82 8 2 12 2 3 27 254 46.7 32.2
1982 100 36 23 5 9 13 3 13 202 49.6 17.8
1983 81 47 6 4 3 0 3 36 181 45.1 26.2
1984 92 129 14 3 3 4 0 28 273 33.7 47.2
1985 164 94 9 1 9 16 4 20 319 51.5 29.4
1986 209 346 10 4 6 14 1 27 619 33.8 56.0
1987 126 463 51 35 9 6 6 78 773 16.3 59.9
1988 271 513 22 46 35 15 34 100 1035 26.2 49.6
1989 268 438 14 44 33 14 32 48 892 30.0 49.1
1990 269 362 19 19 80 37 18 91 895 30.1 40.4
1991 238 185 5 387 93 53 38 70 1069 22.3 17.3
1992 229 149 8 61 57 11 43 129 688 33.4 21.7
1993 278 233 61 107 29 58 33 52 851 32.7 27.4
1994 241 331 33 52 46 46 43 226 1019 23.6 32.5
1995 477 309 43 126 33 33 27 388 1437 33.2 21.5
Total F Dl 5720 7826 364 1 121 534 377 335 2034 1831 1 31.2 42.7

 

 

 

 

 

 

 

 

 

 

 

 

 

Source: Ministry of Finance and Economy, Fiscal and Financial Statistics, various issues.
The Korea Development Bank & Ministry of Finance (1993), Foreign Capital
Inducement in Korea for 30 years.

24

2-3. Tax Incentive Provisions in Korea

According to Korean tax law, foreign corporations that have a permanent
establishment in Korea are treated as domestic corporations. Therefore, a wholly-
owned foreign subsidiary is eligible for all the beneﬁts given under the Tax Exemption
and Reduction Control Law (TERCL). In addition, FCIA provides various tax
beneﬁts for foreign-invested enterprises. In the next several sections, we discuss major
tax incentives for domestic- and foreign-invested enterprises, including investment tax
credits, special depreciation, immediate expensing, tax-free reserves, income deduction
from new share issues, and tax holidays. We can ﬁnd the various tax-law provisions

from various issues of Korean Taxation.

2-3-1. Investment Tax Credits

The Investment tax credits (ITC) were ﬁrst implemented in 1968. At that time,
heavy and chemical industries were eligible for 6-percent special tax credits on
investments in machinery and equipment. The Korean government raised the tax
credit for domestic-made assets to 10 percent in 1971, for the purpose of facilitating
capital accumulation. Under the title of “special tax treatment for key industries” in
the TERCL, a tax reform in 1974 uniﬁed and rearranged all major incentives. This
new system gave three optional sets of incentives: tax holidays, investment tax credit,
and special depreciation--to eligible ﬁrms in key industries. Korean tax policy during
the 19705 was designed to promote exports and to facilitate capital accumulation in
heavy and chemical industries. Preferential tax treatments were given to these

industries for sectoral development.

Due to the drastic changes in the economic and political environment in the
early 19805, the Korean government began pursuing a policy of structural adjustment
and liberalization. The most important tax reform in the early 19805 took place in
1982. The 1982 reform reduced the beneﬁciary list of the heavy and chemical
industries, and eligibility for investment tax credits was limited to the machinery and
electronic industries. Also, the rate of the credit for foreign-made assets was reduced
from 8 percent to 6 percent, and it was halved to 3 percent in 1983. A tax reform in
1986 broadened the beneﬁciary list of the ITC and raised the credit rate to 10 percent
for domestic-made assets. Furthermore, the list of beneﬁciary industries was extended
to all manufacturing and mining from key industries in 1989.

Korea began to lower the tax rate and broaden the tax base in the early 19905.
Technology-intensive industries and the small and medium-size enterprises (SMEs)
have priority for tax incentives. In 1992, the ITC was limited to domestic-made
assets. The rate of ITC was reduced to 7 percent in 1994. Since 1995, the ITC has
been given for SMEs, and other special purposes, such as development of technology
and manpower, investment in facilities for research and experimental use, and starting

a business using new technology.

2-3-2. Income Deduction from New Share Issues
The Korean government allows an income deduction when a ﬁrm issues new
shares. In 1970, eligible ﬁrms were allowed to deduct 18 percent of the amounts of

new shares each year for 3 years. The present value of the income deduction from a

dollar’s value of new-share issues , i”, is the following:

26

 

L
iN=j
O

1'}?

e“"‘iRdt = -;(1—e"’L),

(l)

where p is the ﬁrm’s discount rate, iR is the rate of deduction from income tax base,

and L is the term of the deduction. For example, if a ﬁrm issues a million dollars of

new shares in 1994, it is allowed to deduct $80,000 from its income for 2 years.

When we assume that the ﬁrm’s discount rate is 8 percent, the present value of the

income deduction from new-share issues is $148,000:

(The value of new-share issue5)* (i N )

= (10,000,000)*[(0.08)*(1-O.852)/(0.08)] = 148,000.

The changes over time in the term and rate of the income deduction are summarized in

Table 5.

Table 5. Income Deduction from New Share Issues

 

Year

1970

1975

1978

1980

1981

1983

1984

1986

1991

1994

 

Term of
deduction
(years)
Rate of
deduction

(percent)

 

3

18

 

3

15

 

3

l6

2

24

 

2

 

 

2

18

 

12

 

12

 

2 (3)‘

10 (12):

8 (10)2

 

 

' The terms of deduction for manufacturing.
2 The rate of deduction for small and medium corporation listed on Korean Stock Exchange.

Source: Ministry of Finance, Korean Taxation, various issues

27

 

2-3-3. Special Depreciation

A special depreciation rate is an accelerated rate, thus is in addition to ordinary
depreciation. That is, a special depreciation deduction makes the speed of
depreciation accelerate for tax purposes, and then makes the value of depreciation

allowances increase. The present value of the special depreciation allowance on one

dollar of investment under the declining method, Z ’ , is the following:

 

r 1 T
Z5 = [0 (1 +y)5’e"‘“*’>5 *Pl‘ds — ( ”)5

‘(1+n6’+p’ (2)

where 7 is the special depreciation rate allowed in the tax law, 67 is the depreciation

rate in the tax law, and p is the ﬁrm’s discount rate.

Let’s suppose that a ﬁrm purchases a million dollars of capital asset, and the

100-percent special depreciation (y = 1) is introduced, which makes the speed of tax

depreciation double. When bf and p are given to 10 percent, the present value of the
income deduction from special depreciation allowance is calculated as follows:
$(1,000,000*2*O.1)/(2*O.1+O.1) = $666,667.

For domestic ﬁrms, tax incentives under a 1972 Presidential emergency decree
also included special depreciation allowances ranging from 40 to 80 percent, for
investments using machinery and equipment in selected industries: shipbuilding, steel
and iron, chemical fertilizer, synthetic ﬁber, automobiles, machinery, straw pulp, food
processing, petrochemicals, electronic equipment, electrical machinery and equipment,

construction, and some mining industries. The rate of special depreciation was varied

28

to each industry, ranging from 80 percent to 40 percent according to the scale of the
ﬁrm and the degree of satisfying the criteria. A tax reform in 1981 which focused on a
policy of structural adjustment, limited the beneﬁts of special depreciation to six
industries (shipbuilding, naphtha cracking plants, steel, machinery, electronic, and
aircraft). The rate was 100 percent for these industries. The 1986 tax reform
decreased to the range of 40-80 percent from lOO-percent special depreciation for the
six industries.

In 1992, the special rate was lowered to the range of 20-50 percent from the
range of 40-80 percent. This is because the Korean government began to lower the
tax rate and broaden the tax base in the early 19905. A special depreciation charge
may be calculated for machinery and equipment used directly in mining,
manufacturing, and the electricity and gas industries. The amount of ordinary
depreciation is increased by 20%, 30% or 50%, provided that the ﬁxed assets have
been used for an average of 12 hours or more each day during the business year. The
incentive of the special depreciation was repealed in 1996.

Before 1984, a special depreciation allowance was only allowed for domestic
ﬁrms. Since 1984, for the purpose of attracting FDI, the Korean government has
provided a lOO-percent special depreciation as a tax incentive for foreign capital

investment. It is still valid for FDI in 1996.

2-3-4. Immediate Expensing
The tax reform in 1986 allowed the heavy and chemical industries to choose a

10-percent ITC or 50-percent expensing. Given the weight of immediate expensing,

29

b, the present value of depreciation allowances on one dollar of investment, for an

asset on which some expensing is permitted, Z E , is the following:

ZE =b+(1—b)Z/', (3)
where Z ’2 is the value of depreciation on the part that is not expensed, and j = s is for

the straight-line method, and j = d is for the declining-balance method.

From the equation, if b = 1, the entire investment is immediately expensed (i.e., there is

no tax due on investment)

2-3-5. Tax-Free Reserves

Tax-free reserves, such as those for investment, export loss, overseas market
development, research and development (R&D), and small-scale ﬁrms, are an
important method to mitigate the tax burden by deferring taxation for a certain length
of time in Korea. The deferral rate and terms of income deduction are varied by the
nature of investment and by industry, ranging from 3 percent to 10 percent. With the
introduction of the tax-free investment reserves in 1982, the ﬁrms in key industries can
reserve up to 5 percent of the book value of machinery and equipment investments as
a taxable income deduction. In 1993, the tax-free R&D reserve was changed to allow
a ﬁrm to reserve up to 3 percent (4 percent in certain cases) of annual sales turnover.
The most popular method in Korea was to defer taxation for 4 years (3 years in 1994),
to be paid back to the government in 3 equal installments for the next 3 years, with no

interest charged on the deferral. The beneﬁts from the tax-free investment reserve are

30

the same as those from interest-free debt, given a nominal discount rate p. We can

derive the present value of the tax-free investment reserve from the following.

m+n—l m+l rn+n-l

1 m
R =1R[ jerk/t ——(je“'a’t + je‘“a’t ...... + jig-“din, (4)
0 n 0 0 0
where m is the number of years over which taxes are deferred, and n is the number of

years over which the deferral taxes are paid back.

The ﬁrst term of the parenthesis in equation (4) stands for the present value of total
deferred income without payments. The second term of the parenthesis is the present

value of total payment made. When we rearrange equation (4), we get the following:

1 .
R = —e"°"'(1 + e”’ ........ + e‘”("“) — ne’pw'”) (5)

2-3-6. Tax Holiday

A tax holiday is a temporary tax exemption for corporate income and dividends. Many

developing countries6 , including Korea, have introduced tax holidays to attract

foreign capital.
According to the F CIA, which was implemented in 1961, the tax holiday in

Korea stated that foreign corporations satisfying the criteria of FCIA were entitled to a

31

complete exemption from corporate tax for the ﬁrst 5 years, and a 50-percent
exemption for the subsequent 3 years after the registration of the business. In addition
to the corporate tax exemption, foreign corporations enjoyed exemptions from various
taxes, such as the acquisition tax, property tax, and global land tax, during the
holidays.

The F CIA7 was revised in 1984 to encourage FDI more. The tax-holiday
provisions were only for those investment projects which improved the balance of
payments, employed advanced technology, and were in the export-free zones8 . Under
the FCIA in 1984, the tax-holiday provisions allowed for the exemption of all
corporate income and dividends for any ﬁve consecutive years within 10 taxable years
after the registration of the enterprise. From 1991, all corporate and dividend incomes
were exempted from the tax base for any initial 3 years, and a 50-percent tax reduction
was allowed in two subsequent years within 10 years after the registration of the
corporation. In 1995, the tax-holiday provision was made more generous by enlarging
the exemption terms to a ﬁill exemption in the initial 5 years, and a 50-perent
exemption in the next 3 years, within 10 years after the registration of the corporation.
The tax-holiday provisions during the transition follow the same tax law that is applied

to the ﬁrst investment.

 

6 They are Brazil, China, Ecuador, Malaysia, and Thailand, etc. See terms and conditions in Boadway
and Shah (1996).

7 The revision of FCIA in 1984 provided two types of tax incentive to qualiﬁed foreign enterprises.
One is lOO-perccnt special depreciation, which makes the speed of depreciation double. The other is
the tax holiday. Qualiﬁed enterprises were allowed to choose one of these two incentives.

8 Tax holidays is generally given for foreign companies which employ high technology. However,
companies for the purpose of export. located in Iksan and Masan cities, are also eligible for tax
holidays.

32

CHAPTER 3

THE DESCRIPTION OF THE MODEL

I will analyze U.S. direct investment into Korea through a cost-of-capital
approach. According to the deﬁnition in the benchmark survey by the U.S.
Department of Commerce, Direct Investment (D1) is the ownership or control by one
U.S. person of 10 percent or more of the voting securities at an incorporated foreign
' business enterprise, or an equivalent interest in an unincorporated foreign enterprise.
Person is broadly deﬁned to include any individual, branch, or corporation. Portfolio
investment that does not give management control, is distinguished from direct
investment by a multinational ﬁrm that controls and manages production
establishments located in at least two countries through FDI.

I derive the equations and discuss the methodology that is used for the
calculation of the domestic and transnational cost of capital in this study. The
transnational cost-of-capital model I set up is close to the model used in the OECD
report (1991). Because the OECD study is limited to the manufacturing sector and
omits various investment incentives, I extend the OECD model to nonmanufacturing

sectors and attempt to incorporate the tax incentives prevailing in Korea and the tax

33

treaty parameters between the U.S. and Korea. I also develop the transnational cost-
of-capital model under the assumption of the different tax bases between the
countries. In addition, I include the time-variant tax parameters due to the tax holiday,
as well as exchange rate and international tax provisions in the neoclassical model. I
will compare the costs of capital and effective tax rates between the U.S. domestic
ﬁrms and Korean domestic ﬁrms, between U.S. subsidiaries without tax holidays and
U.S. subsidiaries with tax holidays in Korea. Then I determine whether tax incentives
provided by the Korean government to encourage foreign direct investment (FDI) are

effective.

34

3-1. Post-Tax Rate of Return

I follow the methodology of King and Fullerton (1984) to deﬁne the post-tax
rate of return (S ), which is exactly the post-personal-tax rate of return. I also use a
wealth-maximization approach to derive the pre-tax rate of return (F ), which is equal
to the net-of-depreciation cost of capital, that is F : C — 6 , where C is the cost of
capital and 6 is the economic depreciation rate. The pre-tax rate of return, which is
equal to pre-personal-tax rate of return or post-corporate-tax rate of return, is the
terminology used by King and Fullerton. From King and Fullerton, the tax wedge,
W, measures the extent to which the tax policy affects the marginal rate of return

from holding the asset. The tax wedge is given by:

WzF—S (6)

The effective tax rate (ETR) measures the tax wedge at the margin for a given type of

capital asset. Therefore, ETR is a useful measure of the distorting effects of taxation

on investments. The effective tax rate, which is a “tax-inclusive”9 measure, is deﬁned

by:

E’ = —— (7)

 

S — F
9 An alternative “tax-exclusive measure” in King and Fullerton (1984) is deﬁned by E ' = ——-—.

S

35

Table 6 summarizes different notations between King and Fullerton and Seo.

Table 6. Different Notations Between King and Fullerton and Seo

 

 

 

King and Fullerton Seo
Cost of Capital MRR C
Pre-Tax Rate of Return p = MRR — 5 F = C — 6
Required Rate of Return 6 H

 

 

 

The arbitrage assumptions that have been used in the literature are key
distinguishing features of alternative effective tax calculations. King and Fullerton
(1984) calculate ETRS under two different arbitrage assumptions. They are the so-
called ﬁxed-F and ﬁxed-r cases. However, I believe that the ﬁxed-r is more realistic
assumption. In next pages, I will discuss three arbitrage assumptions, and argue that
the ﬁxed-r case is superior to the ﬁxed-F case. The ﬁxed-F case assumes that the pre-
tax rate of return, P, which is the net-of-depreciation cost of capital, is the same on all
projects ( 10 percent in their calculation). Given the pre-tax rate of return, we can
work backward and compute the ﬁrm’s required rate of return. Under the ﬁxed-F
case, all arbitrage occurs at the ﬁrm level, so that the required rate of return to the ﬁrm
is the same for retained earnings, debt, and new shares, regardless of the fact that the

ﬁrm faces different tax liabilities for different forms of ﬁnance. Therefore the required

rate of return to the ﬁrm, rf , is given by

36

 

 

 

f: p 2.1— :
r l-t 1( it) l—md (8)

where u is the corporate tax rate, p is the nominal after-tax rate of return on equity

to existing shareholders, is is the effective tax rate on capital gains applied to the
accrual base, (,0 is the nominal after-tax rate of return on equity to new shareholders,

m" is the average marginal tax rate on dividends, and i is the nominal interest rate
(the nominal interest rate is equal to the real interest rate (r) plus the inﬂation rate
(7r )). Given the various tax regimes, we work backward from the required rate of
return to calculate the after-tax rates of return for the householders, such as
p, i, and (p .

In the ﬁxed-r case, all arbitrage occurs at the household level, so that the after-
tax rates of return on various assets are the same, which means that i (1 — m’) = ,0 = (a.
The constant real rate of return (r) is given to 5 percent in King and Fullerton’s
calculation. When the savers (including bondholders and shareholders) provide funds
to ﬁrms that carry out investment with nominal interest rate of i, the savers face the
same rate of return on their saving as ﬁrm make proﬁts. However, the savers in
different tax brackets face the different after-tax returns. Given the real rate of return
to householders, we can calculate the different required rates of return for the various

sources of ﬁnancing.

37

Another approach is used by Boadway, Bruce, and Mintz (1984). They use
the actual cost of debt and equity on the market to determine the ﬁrm’s required rate
of return. The cost of debt is determined by the observed interest rate, and the cost of
equity is calculated using earnings-price ratios from stock market data, appropriately
corrected for inﬂation. Because stock-market data do not allow the rate of return on
retained earnings to be distinguished from that on new shares, the after-tax rate of
return on investment ﬁnanced by retained earnings is assumed to be the same as that
on investment new shares in the Boadway-Bruce-Mintz approach.

The ﬁxed-F assumption that the pre-tax rate of return and the ﬁrm’s required
rate of return are constant for all projects is unrealistic, because the pre-tax rates of
return are varying to various investment projects, and because the ﬁrm’s required rate
of return is different across different sources of ﬁnancing. On the other hand, it would
be difficult for me to implement the Boadway-Bruce-Mintz approach, because it is
very difficult to get market data on equity, since only a few U.S. subsidiaries are listed
on Korean stock market. Because I can observe the actual interest rates across
various industries in the U.S. and Korea, and because the ﬁxed-r assumption is rather
more realistic than the ﬁxed-F assumption, the arbitrage condition used in my
transnational cost-of- capital model is the adjusted ﬁxed-r assumption. While King
and Fullerton use a S-percent interest rate for all industries in their ﬁxed-r assumption,
I use the market-observed interest rates, which are different across all industries.

For the personal tax rate on savers, I use the average marginal tax rate on
interest income, which is an “average term," even though different savers may actually

face different tax rates when the ﬁrm distributes its earnings, either to shareholders as

38

dividends and share-value appreciation, or to bondholders in the form of interest. The
net real return after tax is affected by the rate of inﬂation. Hence the real net rate of

return is given by

Szﬂ—Mﬂ—n (%

where i = r + 71' (i.e., the nominal interest rate is equal to the real interest rate plus

the inﬂation rate), and in" is the average marginal tax rate on interest income.
Consider a multinational U.S. ﬁrm that operates a wholly-owned subsidiary in
Korea. The post-tax rate of return on a transnational investment project depends on
taxation by both the home and the host countries. Because the U.S. uses the global
system and allows a credit to relieve double taxation of foreign-source income, the
withholding tax on interest income is deductible in the home country. When we
assume that U.S. residents provide ﬁinds to a U.S. parent ﬁrm that ﬁnances all

investments of its subsidiary, the post-tax rate of return (or the real net rate of return)

for transnational investment, S T , is deﬁned by

ST =[1-max(m’,a)')]i' —7r (10)

where (0' is the withholding tax rate on interest income determined by tax treaty

between the U.S. and Korea, i' is the interest rate in Korea and it is the inﬂation rate

in the United States. According to the tax treaty between Korea and the U.S., the

39

withholding tax rate on remitted interest income is 12 percent. Because the average
marginal tax rate on interest income in capital-exporting countries such as the U.S. and
Japan are generally greater than the withholding tax rate on interest income in capital

importing-countries, we can rewrite equation (10) as the following:

ST:(1—m')i°—7r (11)

3-2. Transnational Cost of Capital

It is useful to derive a simple user cost of capital, before we derive a
complicated transnational cost of capital. The basic user cost of capital is introduced
by Jorgenson (1963), and developed by Auerbach (1979) and King and Fullerton
(1984). For a simple user cost of capital, I assume that there is no tax, and that a ﬁrm
produces output using a single capital input, and that the marginal rate of return to
capital is decreasing. The wealth of the ﬁrm’s owners is the discounted sum of its cash

ﬂows. The objective of a ﬁrm is to maximize the wealth of its owners. That is:

w = je-~(p,F(1<,) -q,I,)dt. (12)

where r is the ﬁrm’s discount rate, q, is the price of capital goods, p, is the output

price, andF is the concave production function. To solve the objective function, we

need the constraint for capital investment. I, is gross investment and 6K, is the

40

 

amount of the current capital stock that depreciates, provided that capital depreciates

at the rate 5 > 0. Therefore, (K, is the amount of replacement investment that must

be undertaken, in order to maintain the capital stock at its current level. The change of
capital stock is equal to the ﬁrm’s gross investment, less replacement investment. That

is:

K, =I,—§K, (13)
For the optimization problem, we set up the Hamiltonian from equation (12), that is:

H = e""1p.F<K.>—q.1. +241, —6K.>1. (14)

t

Gross investment I, is chosen to maximize H at each point in time. From TI : 0,
(7

we get the following:

i=q, (15)

d .
Since I, — (SK, 2 27(K, ), the necessary condition for the path of capital stock 15

d . .
@— + —(2e'") = 0. When we evaluate the expressron, we get the followrng:

aKdr

41

e‘"(p,FK —6/l)+e'"()i—r/i):0 (16)

Let’s differentiate equation (15) with respect to time t, then insert this result and

equation (15) into equation (16). This gives us e‘" (p,FK — 5%) + e‘" (q, — rq,) = 0,

 

. F.
or p,FK — bit], 2 —q, +rq,. Finally, we can solve for the cost of capital ( p, A ).
) F
C,=4’7£=(r—n+5), (17)
( I

where C, is the cost of capital on a dollar value of capital good, F K is the marginal

product of capital, and it is the inﬂation rate which is equal to :4.
!

Our next objective is to evaluate the impact of income taxation and tax policies
on the cost of capital. To begin, I assume that the only tax is the corporate income tax

( u ), and that the government provides an investment tax credit (k) and depreciation
allowances as investment incentives. The ﬁrm’s cash ﬂow, p,F (K , ), is taxed at the
rate of u, and the ﬁrm beneﬁts from these tax incentives by the amounts of

(k + uZ )q, I , , where Z is the present value of depreciation allowances per dollar of

investment. The ﬁrm’s objective function under income taxation and investment

incentives is given by:

42

w =le-"<<1— u>nF(K.)—-q,1.+<k + uZ>q.I.>dz <18)
0

For the optimization problem, we set up the Hamiltonian, that is:

H, : e‘"[(1 — u)p,F(1<,) — (1 — k — uZ)q,I, + 1(1, - 5K, )1. (19)

 

. 6H d
From the necessary conditions, such as c, = 0 and 7 +-(/le‘") = O, we get the
a] 0K d!

following:

[(l—u)p,F,, -(1—k -—uZ)&],]e’" +[(1—k —uZ)rq, +(1-k—-uZ)q,]e'" = 0 (20)

erK

 

When we solve equation (20) for , we get the familiar cost-of-capital formula.

I

_p,FK_(1—k—uZ) _
C,— ‘I: —_——(1-—-u) (r n+6) (21)

 

For the transnational cost of capital, I assume that shareholders live inﬁnitely,
and that there is no uncertainty. I also assume that there are only two factors of
production (labor and capital), that markets are competitive, and that the marginal rate
of return to capital is decreasing. The objective of a multinational U.S. ﬁrm, say a

parent ﬁrm, is to maximize its market value, equal to the stock value of its current

43

 

Aw___g v.

shareholders. Capital-market equilibrium for the parent ﬁrm should satisfy the

following conditions when the shareholder earns his required rate of return p. V,,1 is
the market value of the parent ﬁrm’s equity at the end of period t, including the value

of new-share equity sold during period t, V,” . Thus the capital gains received by

owners of the preexisting shares is equal to V,_, — V, — V,” .

PVI =(1-mdlDt+(1-tg)(Vr.1-Vt *VtN), (22)

where p is the nominal after-tax rate of return on equity to existing shareholders, t g is

the effective tax rate on capital gains applied to the accrual base, m" is the average

marginal tax rate on dividends, D, is the dividend payments of the parent ﬁrm during
period t, V, is the market value of the parent ﬁrm’s equity at the beginning of period t.

Equation (22)10 states that the shareholder’s after-tax return from holding equity at the
beginning of period t is equal to the sum of post-tax dividends and post-tax capital
gains. The shareholder’s required rate of return is viewed as the opportunity rate of
return aﬁer personal tax.

In most countries, including the US, capital gains are only taxed on sale, or
realization, and there is no adjustment of the cost basis for inﬂation. The taxes on
capital gains are therefore deferred until capital gains are realized. This creates an
incentive for savers to postpone the realization of capital gains, and also lowers the

effective accrual-equivalence rate of tax on capital gains. Because we do not have

 

‘0 For a derivation of this equation, see Auerbach (1983).

44

information about the time at which the shareholder realizes its capital gains or how
large a portion of capital gains the shareholder will realize, we need to make some
assumptions that will allow us to convert the statutory tax rate on capital gains into an

effective tax rate. We use the approach of King and Fullerton (1984) and get the

following result:

in (1+p)
: g 11
3 rl+p ’ (23)

where )t is constant proportion of accumulated accrued capital gains realized by

investors, and us the statutory tax rate on capital gams. Solvrng equatron (22) under

the assumption that the value of the ﬁrm is bounded,

limV,e'" :2 012 , (24)

t-sao

 

” A is taken as 10 percent for calculation. This reflects an average lag of ten years between accrual
and realization. Investors in period one realize xi of accrued capital gain, and an unrealized capital
gain is l — zl. . In the second period realizations are equal to 2(1 - xi.) , and so on. Given the
nominal discount rate, p , and the statutory tax rate on capital gains. us , the present discounted
value of total tax payments for a unit of accrued capital gain (e.g., the effective tax rate on capital
gains applied to the accrual base), t g , is given by:

 

 

Au 2.0—Mug 2.0—aw, 2(1-t)3u, 1(1-,1)°°u,
11,— 2+ l+p + (1+,0)2 + (1+,0)3 ............. (1130)” .Whenwe

compute the above equation, we get equation (23). See more details in King and Fullerton (1984. p.
23, p. 222).
‘2 When we solve equation (22) for V, , the residual is fl (1 + r)—(:-1)V.

J=I

For simplicity, we can rewrite the discrete time residual into continuous time, (in, e-er' = 0

{—90

45

 

We obtain the following:

 

r
,_ mgd V.” MS, (25)

 

l—md 5=°° —r(s—t)
V,=,_, l,=,e (DS-

8

where the required rate of return ﬁnanced by equity, r = T—Qt—' Equation (25) states
8

that the value of the ﬁrm at time t is discounted sum of its cash ﬂows over an interval

from present time t to inﬁnity, where 6’" is the continuous time discount factor. To
determine the optimal path of the ﬁrm’s investment and the cost of capital, we need
cash-ﬂow constraints and constraints for capital. At ﬁrst, I specify a capital-stock
constraint. I consider two ﬁrms, a parent company operating in the home country and

a foreign subsidiary wholly owned by a parent ﬁrm. For the parent, I, is gross

investment and 6K, is replacement investment, provided that capital depreciates at the
rate 5 > 0. I, and 6'K,‘ are gross investment and replacement investment of the

subsidiary respectively, provided that capital depreciates at the rate 6' > 0. The

change of capital stock is deﬁned by

k dK‘ Kt—f-i—Ii (26
‘—dt’ ‘—dt' )

The change of capital stock is equal to the ﬁrm’s gross investment, less replacement

investment. That is:

46

 

K, = I, —- 6K, (27)

K; = I,’ — 6'K,‘ (28)

For cash-ﬂow constraints, I start by equalizing the sources of funds (total revenue)
with the uses of funds (total expenditures). In other words, the following cash-ﬂow
constraints for the parent and the subsidiary must be satisﬁed in equilibrium. In the
following equations, the left-hand sides indicate the uses of ﬁinds of the parent and

subsidiary, and the right-hand sides indicate the sources of funds of the parent and

subsidiary. I: and T,” are the tax liabilities on domestic operation and on repatriated
subsidiary earnings for the parent, respectively. T' is total tax payments of the

t

subsidiary in the host country.

Parent ﬁrm:
(1 — 0)D, + E,V,‘V' + E 1' + 7,’ + ZR + (1 +i)B, +q,I, + W,L,

tt+1

= (1 - 03510: + P1F(K,,L1) + 314+ VIN + E,[1+i(1- (0')]4 (29)

Subsidiary ﬁrm:

(1-0')D,' +[(1+i(1—a)')]l,' + 7,‘ +q,'I,' +W,‘L', +(1+i')B,'

__. ng'(1<°, 1;) + 1,1, + 3,1, + V,“ (30)

I

47

The new notations in equation (29) and (30) are given below. Note that the symbols

with (0) are those related to the subsidiary. In equations (29) and (30),

B, = Borrowing at the beginning of period t by the parent, which will be repaid
at the end of period t with interest at the rate i,

BMI = Borrowing at the end of period t by the parent,

B,‘ = Borrowing at the beginning of period t by the subsidiary from the parent,
which will be repaid one period later with interest at the rate i' ,

l,' = The parent’s loan to the subsidiary, which will be repaid at the end of
period t with interest at the rate i,

o = Withholding tax rate on dividends,

(0' = Withholding tax rate on interest income in the host country,

V,”' = New-share equity of the subsidiary ﬁnanced by the parent,

W,= Wage rate,

L, = Labor input,

q, = Price of capital goods,

p, = Output price,

F = concave production function, and

E, = the spot exchange rate in units of home-country currency per unit of host-

country currency.

48

Here I specify the tax liabilities for both the parent and the subsidiary to
calculate the repatriation tax rate on the subsidiary’s remitted dividends. The
repatriate tax rate on dividends depends on each country’s tax system, and the
methods of double-taxation relief. The changes in tax parameters affect capital ﬂows
and tax revenues in the home country and the host country. I specify the source of tax
revenue in both countries only under a classical system, because major capital-
exporting countries, such as the U.S. and Japan, adopt it. The classical system means
that there is no correction at the shareholder level for taxes paid at the ﬁrm level.
Therefore, dividends are taxed at both corporate level and personal level in a classical
system. Some countries (e. g., the United Kingdom) adopt the imputation system. The
imputation system allows a credit for taxes paid at the ﬁrm level at the rate of c to

reduce double taxation on dividends. Tax liability in each country is

T, = u [If +(o — c)D,‘ for the parent ﬁrm, and (31)
T1. = 11' HT +(0' —c')D,c' for the subsidiary, (32)

where 117 is the home-country taxable base, TIT. is the host-country taxable base, and

c is the rate of imputation.
The following conditions are satisﬁed because the distributed proﬁts are taxed

at the same level as the undistributed corporate proﬁts, and because the rate of

imputation, c, is equal to zero in a classical system:

49

D, = D,‘ = (1 —u)D,G (33)
D,’ = D," = (1—u')D,G' (34)
where D,(D,' ) is dividend payments, D,‘(D,°) is dividend payments gross of any
imputation credit, and D,G(D,G') is dividend payments, grossed up at the corporate

rate. To show how different tax bases between the home and the host countries affect
the tax liability on repatriated subsidiary earnings, I specify the parent and the

subsidiary’s tax bases. Taxable proﬁts of the parent and the subsidiary are as follows:

Hf = p,F(K,,L,)— W,L, —(uZ+k)q,I, —iB, (35)

H,“ = p,'F'(K,',L;) ——W,'L: —(u'Z' +u'R' +k')q,'I,' —i°B,' —il,' — u'iN'IﬂN' (36)

where Z (Z ') is the present value of depreciation allowances, R' is the present value

of the tax-free investment reserve, k(k') is investment tax credit, and i N ' is the

present value of the income deduction for new-share issues.
The tax due on repatriated subsidiary earnings is composed of tax liabilities on

remitted dividends and interest payments of the subsidiary.

T” = T,” + 77' (37)

50

The tax due on repatriated subsidiary earnings depends on the method of double-
taxation relief of the home country and both of the tax bases deﬁned by the home and
the host country’s tax rules. The U.S. uses a credit-with-deferral method for the
subsidiary, and a credit—without-deferral method for the branch. Under a credit-with-
deferral method, the home country allows a credit (at a maximum rate of 34 percent in

U.S.) for the host-country taxes paid, to avoid double taxation.

To calculate 7:0 , the tax liability on remitted dividends, the difference between
the taxable bases of the home and the host country is an important element.
According to the U.S. tax rules, depreciation of the subsidiary’s assets at time of
purchase is based on pro-1981 U.S. rules, and the assets at time of purchase may have
to be denominated in foreign currency, except for some special cases13 .

In addition, U.S. tax rules do not allow tax sparing” and thus the beneﬁts of

tax incentives provided by Korean government are required to be taxed. The taxable

income of the subsidiary according to U.S. tax rules, Hi“. , is given by the following:

II?" -= p:F'(K:.L:)— WJL: — (u'2+k>q:1: -i'B: -11; (38)

By combining equation (36) with equation (38), we can get the following:

 

’3 Assets may be converted to U.S. currency for countries that are continguous to US, or for countries
they are experiencing hyperinﬂation.

‘4 Tax sparing ensures that the home-country taxes are waived or credited on the remittance of
dividends and interest income beneﬁting from the tax-incentive provisions in the host country.
Japanese multinationals have an advantage over U.S. multinationals, because Japan allows a tax-
sparing credit for remitted incomes from 14 developing countries. including Korea (Slemrod and
Timbers, 1991).

51

11:" = nf’ + (u'Z' — "'2 + k' — k + u'R' + u'iN' )q,'I,' (39)

According to U.S. tax rules, the tax liability on remitted dividends of the
subsidiary is calculated as the amount of corporate taxes on the grossed-up value of

the subsidiary’s dividends, less foreign tax credits (FTC) allowed in U.S. rules.
T,” = no? — FTC (40)

Dividends, grossed up at the corporate rate on the U.S. taxable base, include dividends
after corporate tax and corporate income taxes deemed to be paid on dividends in the
host country. In a classical system, dividends are taxed at both the corporate level and
the personal level. Therefore, the gross dividends include the distributed dividends
and corporate taxes on dividends. Tax liability on the subsidiary’s proﬁts in the host
country includes the corporate income taxes which are deemed to be paid on both
dividends and retained eamings. The deemed-paid corporate taxes are taxes on
remitted dividends at the corporate level that the subsidiary will pay or already paid to

the host country. The deemed-paid-corporate taxes on dividends are computed from

the ratio of dividends to the subsidiary’s proﬁts, W, times tax due on the

r_r

subsidiary’s proﬁts, 7;. , according to tax law in the host country.

52

D,’
_) (41)

The withholding taxes and corporate income taxes deemed to be paid on dividends in

the host country are exempt in the home country. Thus, the amount of foreign tax
credit is the following:
D.

nf—FL

FTC = 6'1); + T,'( (42)

where 0' is the withholding tax rate on repatriated dividends in the host country.

From equations (38) to (42), we can calculate the tax due on the remitted dividends of

the subsidiary. That is,

u- U

 

where ud° =u'(1+sy).
In equation (43), u". denotes the deemed-paid corporate tax rate in the host

country, and s is a deviation of tax incentives between two countries, given by

 

'5 By inserting equation (41) and (42) into (40), we can get the following:
um" -u‘ n" , , , .
7:0 =[ t . 7’. -O' ]D , where I: = u H: . For the next step, we divide the ﬁrst
Him --u n:
equation in the parentheses by H?” and insert equation (39), and then rearrange the above

equation, to get equation (43).

53

s = u'Z' —- u'Z + k' —- k + u’R' + u'iN a; , where a; is the fraction ofinvestment

ﬁnanced by new-share issues. 7 is the ratio of capital investments to taxable profits

on
I!

Hus. '
(

 

according to U.S. tax rules, which is y = For simplicity, I assume that y is

constant. The sensitivity analysis of this assumption shows that the cost of capital is
very insensitive to the ratio of capital investments to taxable profits. Therefore, my
results are robust with respect to this assumption. For more details, see Figure 3 in
appendix.

Equation (43) represents the tax due on remitted dividends when the subsidiary
is in a deﬁcient credit position, under which the FTC is less than the home country’s
tax burden on remitted dividends. But when the subsidiary is in an excess credit
position, we can say that the subsidiary ﬁilly uses the FTC, and its tax due on

repatriated dividends is zero. Therefore, we can rewrite equation (43) as

u — ud

 

TD = max( - 0°,0)D: (44)]6

d.
1~u

The equation (44) ignores the fact that the United States allows deferral of the
proﬁts earned abroad by U.S. subsidiaries until their profits are repatriated as

dividends. The withholding tax on remitted dividends is also deferred until the

 

'6 The U.S. use a credit-without-deferral method for the branch of the U.S. corporation. Therefore,
the home country tax burden on remitted dividends for the branch ( Z8 ) is given by:

r,” = unf —u'n,".

54

subsidiary repatriates its proﬁts as a dividends. We can easily incorporate tax deferral
of the proﬁts into equation (44). Because tax is deferred until proﬁts are repatriated,

the beneﬁt of tax deferral is the same as interest-free debt for a certain period. When

the subsidiary’s discount rate ( p') , the length of the deferral (d), and interest rate in

the host country (i') are given, we can calculate the present value of tax deferral per

unit of dividend, i D .

d
I'D =Ie’poti°dt = I (l—e"”.d), (45)
0

 

Tax deferral reduce the home country’s tax liability by the amount of the

present value of tax deferral per unit of dividend, iD . Therefore, the home country’s

tax burden on remitted dividends is changed into the following:

u -— u"

 

7,0 = max[(l —iD)(l d. —a°),0]D,' = a"D,'. (46)
- u

where a" is the effective tax rate on repatriated dividends.

I incorporate tax deferral theoretically into the cost-of capital model. For my
empirical work, however, I assume that subsidiaries remit their proﬁts immediately,
and thereby the present value of tax deferral is zero. This is because we do not have
the information about the time when subsidiaries remit their proﬁts, and because

different subsidiaries have different strategies in remitting their proﬁts. In Chapter 6, I

55

will test the sensitivity of costs of capital to changing the present value of tax deferral,
as the length of deferral varies. I also discuss the eﬁ‘ect of tax deferral on the cost of
caphal

The withholding taxes paid in the host country are exempt from the home

country’s taxation when the subsidiary remits its interest income to the home country.
When the withholding tax rate on remitted interest income, a)’, is less than the
corporate tax rate in the home country, the home country’s tax burden on remitted
interest income is u — a)’ . If the withholding tax rate, a)‘, is greater than the U.S.
corporate tax rate, the effective tax due on remitted interest income in the home
country is zero. Therefore, the tax due on remitted interest income is the following:
77 = max[u — a)’,0]iE,l,'=a)"iE,1,', (47)

where a)" is the effective tax rate on remitted interest income.

When we incorporate all tax parameters in equations (29) and (30), we can get the

following cash-ﬂow constraints.

Parent ﬁrm:

D: : (1-“)[er(KnLr)_”/1Lr]— (1 "k _ Z)qr1r —[1+i(l'—u)]Br

+B

H]

+ V,” — E,V,”' — 5,1,1, + (1 — o" - o“)E,D; +[1+i(1— (0' — w'°)]E,l; (48)

56

Subsidiary ﬁrm:

D: =(1-u')[PIF'(K' L')-W.'LI]-[1-k'-u'(Z'+R')qu1f-[1+i°(1-u')l3,'

t’t

+8;1 —[1+i'(1—u')]l,' +1;

,1 + (1 + u‘iN' )I-",N’ (49)

Modigliani and Miller (1958) indicate that, in the absence of taxation and
bankruptcy, corporate ﬁnancial policy is irrelevant and has no effect on the value of
the ﬁrm. However, the real world is characterized by both taxes and bankruptcy risk,
so that Modigliani and Miller’s theorem (195 8) does not hold.

The analysis by Modigliani and Miller (1963) can be viewed as an analysis of
the choice of the optimal ﬁnancial policy for the ﬁrm. With the introduction of taxes,
Modigliani and Miller (1963) show that a ﬁrm could increase its market value by
increasing its leverage. In the extreme case, a ﬁrm would ﬁnance its investment
entirely by debt, because the cost of capital ﬁnanced by debt is relatively low
compared to the cost of capital ﬁnanced by equity, under the current U.S. tax system.
I do not calculate the optimal ﬁnancial policy for ﬁrms, but calculate different costs of
capital for different ﬁnancial policies. It may seem irrational on part of ﬁrms, which
choose ﬁnancing methods other than those that appear optimal. However, ﬁrms
cannot ﬁnance their investment entirely by debt, because ﬁrms have borrowing
constraints and credit rationing, and because the high share of debt ﬁnancing increases
the possibility of bankruptcy risk. Furthermore, asymmetric information prevents ﬁrms

from choosing the optimal ﬁnancial policy.

57

In order to simplify the transnational cost-of-capital model, King (1977),
Tajika and Yui (1988) and Chou and Wu (1994) assume that a ﬁrm maintains its
optimal ﬁnancial structure, and then the fractions of investment ﬁnanced by borrowing,
new-share issues, and retained earnings are constant. I follow their assumption. They
also use the assumption of static expectations, which implies that B, is equal to 1'3,+1 .
Instead of this assumption, I assume that the ﬁrm maintains its new borrowing and
loans at a steady-state level (Mintz 1993). These assumptions apply both to parents
and to subsidiaries.

Br : qurlr 3 Bx] : (1+7I)Bra (50)

f

where 7: is the rate of inﬂation in the home country.

3' = aﬁq:11., BL] = (1+7r')3,', (51)

I

where 7z' is the rate of inflation in the host country.

a0=l—a,—a2, (52)
where do is the fraction of investment ﬁnanced by retained earnings, a1 is the fraction

of investment ﬁnanced by borrowing, and a2 is the fraction of investment ﬁnanced by

new-share issues for the parent.

a0: =1— al.1— a2: — a3.r’ (53)

58

where 010' is the fraction of investment ﬁnanced by the subsidiary’s retained earnings,
a,’ is the fraction of investment ﬁnanced by the subsidiary’s local borrowing, a; is

the fraction of investment ﬁnanced by the subsidiary’s new-share issues, and a; is the

fraction of investment ﬁnanced by the parent’s loan to the subsidiary.

V)” = 02/1,] t (54)
V?" = 05.61? I: (55)
1? = 03361.7! , 15.1 = (1+7r')l,' (56)
Plugging D: into D, , and inserting these assumptions and tax parameters into the
cash-ﬂow constraints, we get the following:
D: : (1_u)[er(Kr’ Lr)—;V1L:]—(1— A)qur
+ (1-0° -O"){(1-ll')E)[PIF°(KI,LI)-VK°Lil-(1- WWII}, (57)
where A = k + Z —[i(1—u) — 7r]a, + a, (58)
o o o o -o o o 0 0.1V. 1
andA 2k +(Z +R)—[I (1—u)—7r]a,+(1+ur‘ — , ,,)a2
1— 0' — a
+{7r'—i'(1-u')+ ,[i'(1—a)'—a)")—7r']}a3'. (59)

1—a°—0

59

Inserting equation (57), the expression for the dividends of the parent ﬁrm, into the

stock-value equation (25) for the optimization problem, we get the following:

l-md
l—t

8

 

lowe‘" {(1—11)[PF(L,K)—WL]—(1— “11+

(1-0' —0")((1-1")[P'F'(L',K')-W'L']-(1- A')(1'1')}, (60)

where A°=k+Z—[i(1—u)—7r]a,+(1—1 — 9a,. (61)
—m

 

For the global optimization problem, we set up the current-valued

Harrriltonianl7 from equation (60), that is,

H! : (1—u)[pr1:(l’r’Kr)— VVILJ—(l— A°)qr1t
+(1-0° ‘0"){0-U')E[PIF°(LLK:)-Wi'Lilrﬂ - A')E‘Iflf}

+aur(1r _6Kr)+lur.(]t. -6.Kr.) (62)

Gross investment [(1') is chosen to maximize H at each point in time: that is

 

6H 5H
__ :0 63
ﬂ ( )

Zora.

 

‘7 This becomes H = er'H, where H is the general Hamiltonian function. Sec Kamien and
Schwartz (1981).

60

Next, the necessary condition18 for the paths of capital stock is given by

 

 

p— m = — 5K (64)
i. O a,
,1 _ r,u = — 0K. (65)

We can derive the cost of capital from (63), (64) and (65), that is,

AF). _(1‘Af) fI_ f
q, — (l-u) [571,” 1’ (66)

 

Parent ﬁrm:

where Fk is the marginal product of capital and rf is the parent ﬁrm’s required rate of

return.
i

Subsidia ﬁrm' _ [afi—E—Hf']
ry ' q: (l—u') q: E:

p.Fk. _ (1“ Af ) (67)

3

where r]. is the subsidiary’s required rate of return.

14F)

t

 

Af , A? ,rf , and r]. vary to the sources of ﬁnancing. The term is the cost of

 

‘8 The shadow price 1. is equal to ye‘" in the current-valued Hamiltonian. From the second

necessary condition, we get ,1 = _ ﬂ = _ gig" . When we differentiate 2 with respect to time
(K

t, we get the following: )1 : —rpe’" + ne'". By equating — rpe‘" + pe'" to _%{:e-" , we get

equation (64).

61

capital (C,) on a dollar value of capital stock in period t. The pre-tax rate of return in

King-Fullerton methodology is the cost of capital less depreciation rate, that is

pf}
~ —— —- a (68
q. )

I

Without taxes, inﬂation, or relative price changes, the real pre-tax rate of return to the
savers would be equal to the interest rate. The cost-of-capital formulas I have derived
are weighted averages, with the weights given by the ﬁnancial sources of ﬁrnds,
because the required rate of return to the shareholders varies according to the ﬁrm’s

ﬁnancial method. Thus the cost of capital is the following:

C, = 016C? + (INC! + 06in (69)
C' = ag,C,°' + a'C" + a' C3' + a;,C,3' (70)

t 1!! 2!!

C,°, C,l and CE are the cost of capital ﬁnanced by retained earnings, borrowing, and
new-share issues. C,“ , C," , CE' and Cf’ are the costs of capital ﬁnanced by retained

earnings, local borrowing, new share issues, and parent’s loans to the subsidiary,

respectively.

The ﬁrm’s required rate of return is an important factor in determining the cost
of capital. I study here how the source of ﬁJnds affects the cost of capital through the
required rate of return in the host and home countries. Three different ﬁnancial

policies for the parent ﬁrm and eight different policies for the subsidiary are considered

62

under the credit-with-deferral system. This analysis comes from King-Fullerton (1984)
and Alworth (1988). In the case of debt ﬁnancing, the required rate of return, rf , is

equal to i(l — u) in an equilibrium, since interest payments are usually deductible
against the corporate tax base. If the parent ﬁnances investment by retained earnings,

the net yield after effective capital gains tax is (1— tlg)rf , and the opportunity cost of
investment is (l - m’)i . By equating (1— tg)rf to (l —m’ )i, we can get

f _(1-m’)i

(l—tg)

. When we consider an investment project ﬁnanced by new equity in

the parent, the shareholder’s net dividend aﬁer tax is (1 — m" )6’rf , where 9 is the

dividend that a shareholder receives when one dollar of the ﬁrm’s earnings is
distributed. 6 is the additional potential income which the parent would receive if the
subsidiary distributed a units of retained earnings. The value of 6 is unity under the
classical system, chosen by the United States, because there is no correction at the

shareholder level for corporate tax paid. By equating the shareholder’s net dividends

to the investor’s opportunity cost, (1 — m' )r' , we can get the ﬁrm’s required rate of

(l-m’)i
(l—md)

return, rf , which is equal to
In the case of transnational investment through the subsidiary, it is very

important to take taxes on repatriated incomes into account for determining the

required rate of return in the subsidiary. When the subsidiary borrows ﬁrnds from the

parent to ﬁnance its investment, the effective tax on remitted interest income is a key

variable. For example, if the parent were to ﬁnance its loans to the subsidiary by debt,

63

(l-a)" )rf. is the net rate of return on repatriated interest income, and the opportunity
cost of lending funds to the subsidiary is (1— u)i. By equating (l-w")rf° to (l — u)i ,

f. _ (1—u)i

_ (1“ w") . The parent would require a

we can get the required rate of return, r

yield such as times the required rate of return on domestic investment,

____1._
U-w")
where a)" 2: max[u — a)’,0].

Consider now the appropriate required rate of return when the subsidiary

ﬁnances its investment by parent’s purchases of new shares from the subsidiary. The

net rate of return after effective tax on repatriated dividends is (1 — a" )r f. , where

u—ud

0' ° = max[

 

- 0' ,0]. By equating this to the domestic required rate of return,

we can determine the required rate of return for an investment ﬁnanced by the parent’s
purchase of new shares. The required rate of return on an investment ﬁnanced by the
subsidiary’s retained earnings or local borrowing, is set to be the same as that for
domestic investment ﬁnanced by parent’s retained earnings or local borrowing.

Table 7 summarizes the ﬁrm’s required rate of return for different ﬁnancial policies of

domestic and transnational investment.

64

Table 7. Required Rate of Return under Different Financing Policies

 

Financial Policy

The Required Rate of Return

 

 

Domestic investment
. Financed by parent debt
. Financed by parent retained earnings

Nu—d

3. Financed by parent new-share issues

Transnational investment
4. Financed by subsidiary new—share issues
provided from parent retained earnings

5. Financed by subsidiary new-share issues
provided from parent debt

6. Financed by subsidiary new-share issues
provided from parent new equity

7. Financed by subsidiary loan borrowed
from parent retained earnings

8. Financed by subsidiary loan borrowed
from parent debt

9. Financed by subsidiary loan borrowed
from parent new share issues

10. Financed by subsidiary retained earnings

ll. Financed by subsidiary local borrowing

 

((l-u)
(l—m')i
(l—z,>
(l—m')i
(I—m">

 

(l-m')i
(1—o“)<1-r,>
(l—u)i
(1-0")

(l—m')i
(1-a”)(1-m")

(l-m')i
(I—co”)<1-t,)
(l-u)r'

(l—m”)
(l-m')i
(l-w”)(l—md)
(l—w°)i'
(l-t;)
(l—u°)r'°

 

 

 

 

 

 

 

SOurce: King-Fullerton (1984) for domestic investment.

I extend Alworth’s (1988) work for transnational investment.

65

 

CHAPTER 4

THE MODEL WITH TAX HOLIDAYS

Mintz (1990) incorporates the tax-holiday provisions into the cost-of-capital
model. He fails to incorporate the international taxation in calculating the cost of
capital, because his analysis is limited to domestic investment.

Mintz and Tsipoulos (1994) extend the earlier Mintz model by incorporating
both the host-country tax and the home-country tax on the repatriation income from
foreign subsidiaries. Even though equity transfers from the parents to the subsidiary
and intercompany loan are very popular methods in ﬁnancing the subsidiary’s
investment, their model only focuses on investment ﬁnanced by retained earnings.
Furthermore, Mintz (1990) and Mintz and Tsipoulos (1994) underestimate the beneﬁt
of tax holidays, because their model does not fully capture the tax savings from tax
holidays.

In section 4-1, I incorporate the tax-holiday provisions and international
taxation into the cost-of-capital model, taking into account various ﬁnancial policies.
To capture the beneﬁt of tax holidays correctly, I extend the cost-of-capital model in

section 4-2.

66

4-1. Transnational Cost-of-Capital Model with Tax Holidays

As discussed in section 2-3-6, the tax holiday is available to foreign
corporations, for the purpose of attracting foreign capital into Korea. The tax holiday
has four effects on the cost of capital. One is the reduction of the cost of capital,
through a temporary tax exemption. A proﬁt-making ﬁrm enjoys the tax savings from
the tax holiday. However, a ﬁrm in a loss position does not pay corporate tax, and
cannot use the beneﬁt of a temporary tax exemption. Therefore, I assume that the
U.S. subsidiary has positive proﬁts on its investment.19 The second effect is that the
tax holiday reduces the present value of depreciation allowances and other tax
incentives signiﬁcantly. As a result, the beneﬁt of the tax holiday decreases. Third,
the tax holiday has an effect on the required rate of return. The effect on the required
rate of return varies, depending on the source of ﬁnancing. Its effect can be either
positive or negative. Finally, a tax reduction in the host country decreases the
amounts of the foreign tax credit of the subsidiary, thereby raises the home-country tax
burden. I discuss more detailed explanations for these four effects as follows.

Because the corporate tax rate during the tax-holiday provisions varies, there is
variation over time in the value of depreciation allowances, the ﬁrm’s required rate of
return, and the tax rates on remitted income. In order to derive the cost of capital
under the tax holiday, we need to specify the time-variant parameters during the tax
holiday. The corporate tax rate during the tax holiday is zero for the ﬁrst holiday

term, and 0.5 u for the second holiday term in Korea. When )11 and h2 are the ﬁrst and

the second holiday terms respectively, the statutory tax rate in Korea is the following:

67

OforOSt<h1
u' = 0511' for h1 s t <h1 +112 (71)

t

u' fort 2 h1 +h2

When a tax holiday is introduced, the tax savings from the depreciation allowances
are partially offset and varied over time. Depreciation allowances play no role as a tax
incentive when the asset that the subsidiary purchased is short-lived capital that ﬁilly
depreciates during the tax holidays and does not generate income after the tax
holidays. But when the ﬁrm purchases long-lived capital during the tax holidays and it
generates income after the tax holidays, the ﬁrm faces a diminishing tax saving from
depreciation allowances after the holidays. Because the tax saving from depreciation
allowances is signiﬁcantly reduced during the tax holiday, postponing of the
investment will increase ﬁiture saving from depreciation allowances. If the ﬁrm can
defer the deprecation allowances until after the tax holiday, the tax holiday becomes a
more generous tax incentive. Some developing countries20 permit depreciation
allowances to be deferred until after the tax holiday, but there is no deferral of

depreciation allowances in Korea. When an investment is performed at time 0, the

present value of depreciation allowances, Z; , is equal to

 

19The cost of capital for tax loss ﬁrms is well developed in Boadway and Shah (1996).
20The ﬁrms operating in Cote d’Ivoire and Malaysia can defer the depreciation allowances until after

the tax holiday.

68

3:". . o - 0+ 7'. IZhl+h2 . ' _ ‘+ T. :2“)
I rlté're‘p ‘5 ”619+!“ 11,6720” ‘5 )‘ds+ h!
3: S: "

. o _ 0+ 7"
1 6T e (p ‘5 )‘ds
3:0 "2

1 . . r‘ 1 . r‘ +
= "53:?[05’ e'(” “5 ”l3 'p—.+—67'[0.5u'e’(p ”'5 ”1:: h’
1 . r"
0 —( +6 )3 co
- p. + 6T. [1’ e p ]hl+h1
T' aT' . ,- a" . To
= u' —.——6—T' + 0.511. 7.9"” +5 )h‘ + 0.5u' ‘76—“) +6 Xhﬁh’) (72)
p + p + p +

For a long-lived asset, the present value of depreciation allowances at time 0 is
reduced but not eliminated by a tax holiday. More numerical speciﬁcations for Z , are

discussed with the cost of capital at each point of time.

The tax holidays also affect the repatriation taxes on dividends and interest
income. The subsidiary ﬁrm is not exempt from withholding taxes on remitted income
even during tax holidays in Korea. When we assume that the taxable bases in both
countries are the same, the rate of repatriation tax varies with the statutory tax rates in
the host country. Equation (73) represents the time-variant effective tax rates on

remitted dividends.

69

f . ’

01' = max[u —0',O] for O _<. t <hl

-O.5 '
0’" =<a°° = max[-u—u-,--o",0] forh St<h +11, r (73)
2 1—0.5u ‘ ‘ ~

.. u — u' ,
0 =max[ , —0‘ ,O]fort2h,+h2
\ '— ll J

 

 

 

The effective repatriation tax on interest income is max[u — w' ,0]. Because the

corporate tax rate in U.S. is always greater than the withholding tax rate legislated by

tax treaty between the U.S. and Korea, the repatriation tax rate on interest income is

set to u — w' .

The tax-holiday provisions make the ﬁrm’s required rate of return vary. When
the investment of the subsidiary is ﬁnanced by local borrowing, the required rate of
return is increased. Because the corporate tax rate during the ﬁrst holiday term is

zero, the interest-expense deduction cannot be used to reduce taxable income. In

equation (73), a," is the highest value, followed by of and a". The higher tax rate

on remitted income leads to the higher required rate of return. The ﬁrm’s required
rate of return on a project ﬁnanced by the subsidiary’s new-share issues is the highest
in the ﬁrst holiday term, because it depends on the highest tax rate on repatriated
earnings in the ﬁrst holiday term. Table 8 indicates the required rates of return during

the holiday. The remaining required rates of return are the same as those in table 7.

70

Table 8. Required Rate of Return under Tax Holidays

 

 

 

L

 

 

 

Financial Policy The ﬁrst The second
holiday term holiday term
4. Financed by subsidiary new equity provided (1-"1')i (1 -M')i
(1-01")(1-!g) (1-02")(1-Ig)
from parent retained earnings
5. Financed by subsidiary new equity provided (1 — “)1 (1 - ")1
(1—0',”) (1—02”)
from parent debt
6. Financed by subsidiary new equity provided (1- In") (1 -rn')1'
(1-a,“)(1—d) (i-ofxi—m")

from parent new equity

1 l. Financed by subsidiary local borrowing

 

 

(1 — 0.5101'

 

 

Source: Author’s calculation

71

 

Consider a multinational U.S. ﬁrm that operates a wholly-owned subsidiary in

Korea. Tax holidays are provided toward the U.S. subsidiary. Given time-variant

parameters such as u a a) and r, , we can solve the following problem under the

(’I’I

same cash-flow constraints and capital constraints as those in chapter 3.

 

 

 

V = 11:7 136"“ {(1—u)1pF(L.K)—WL1-(1—A°>ql+

(1-0°'01..)([P°F'(L',K')-W'L°]—(1‘-A,°)<I'1')}d’

1__ d
+ 1:: K’e {(l—u)[pF(L,K)—WL]—(l—A°)q[+
(1-0°—0'2")([P°F°(L°,K°)-W'L°]-(1-Ai')(1'1°)}d1

1— d so
+ 1-7, Le‘"{(1—u)1pF(LK)-WL1—<1—A°>4I+
(1-0'-0")([P'F°(L°.K°)-W'L']-(1-A')(I'1°)} (74)

where A" = k +Z—[i(l—u)—7r]a, «him, (75)

The cost of capital derived from equation (74) is varying with time. After setting
current Hamiltonian as in section 3-2, we get the solution from the necessary and the
sufﬁcient conditions, such as equations (63), (64), and (65) in section 3-2. By
rearranging these two conditions, we get the cost of capital of the U.S. subsidiary for

new investment at the time t.

72

. l—A' ' E Z'
C.”=<——:—><6'—5’%-—i+rf>+ ‘

l—u, (10 E0 ' 1—u,’

 

(76)

The price of capital and the exchange rate are constant over time, because my study is
limited to the case of certainty. The other tax incentives, such as the investment tax
credit, tax-free reserve, and income tax deduction for new-share issues, are no longer
offered during tax holidays. Equation (76) indicates that the cost of capital for new
investment varies at each point of time during tax holidays.

The cost of capital during tax holidays is composed of two parts. The ﬁrst

portion is quite similar to the cost-of-capital formula in chapter 3. A,’ in the ﬁrst

portion contains the information on time-variant parameters, such as depreciation
allowances and tax rates on remitted income. When the investment is performed

during the ﬁrst holiday, the cost of capital is

 

 

. ° E
C,“ =(1—A,’)(6' —1ﬁi———.—°+r,f)+z,', (77)
0 £0
0 ' O o. O O 1 O O -.
where A, 15 equal to Z, —[1 —7r ]a, +(l— , ")az —{7r —1
1—0 —0,
1 - O O
+ . .. [1(1-ll)-7r ]a,}, (73)
1—0 —0’,

and the present value of depreciation allowances during the ﬁrst holiday term, Z ,' , is

equal to

73

F") r. 5:51”: 3:00

. o _ .., _ . o _ o T. _ . o _ 0+ 1'. _
Iu,6re(p ‘5 X”)ds+ Irl,6Te(p+5 X")+ In 5Te1p’5 X")
3:! 32h s=h,+h2

_ 11°67. + 0.511'6T.
— p' +6T. p' +57.

 

. o o I"
(e-(p +57 )(hl-t) + e-(p +5 Xh‘h‘O)‘ (79)

The second part of equation (77) represents the loss of annual depreciation
allowances from holding a dollar of capital during tax holidays. Mintz (1990)

interpreted this as a tax depreciation penalty of investing in an asset during the tax
holiday. When the subsidiary performs investments during the ﬁrst holiday term, Z, is

equal to 0.5u'67' (e’("""roxh"’) + e""'*"r.x"'"‘*"’) . Because the loss of annual
depreciation allowances decreases with the value of t , the postponing of the
investment would increase ﬁiture tax savings.

When the investment is performed during the second holiday term, the cost of

capital is given by:

 

 

 

o 1"A. q. E Z.
ch = f §'___O___ 0 + f _+_ __t ’ 8O
' (l-O.5u')( (1:, EO r2) 1—0.5u° ( )
O ' O u. C O O 1 C
where A, rs equal to Z, —[r (l—O.5u )—rr ]a1 +(1— , “)az
1—0' —02
O o. O 1 - I O
—{zr _, (1-0.5u )+, . ..1z(1—u)—4 1a.}, (81)

and the present value of depreciation allowances during the ﬁrst holiday term is equal

10

74

s:hl+h2 3:“) r.
. O - .+ r. __ . . _ .+ _.
111,67?” ‘5 X’ "(15+ I14 57c” ‘5 ””615
3th 3:h*h2

0.51723”
2 ———e

P-+5T’ WW “Hi-()- (82)

When the subsidiary performs investments during the second holiday term, the

change in the present value of depreciation allowances, Z,’ , is equal to
0517'5T°e*‘0'*5’°*’~"h-” .

After the tax holiday, Z,’ is constant and equal to Z', and thus the change in the
present value of depreciation allowances is zero. A,’ is also constant and equal to

A' (see equation (59)). Thus the cost of capital is constant over time after the tax

holidays. The constant cost of capital is given by

l—A' " E,
_( )[5-_f.le__

'11—; q; E. ”1 ‘8”

Because the constant cost of capital implies that the capital stocks are constant over
time, an investment after the tax holidays is the replacement investment to maintain the

current capital stocks.

75

4-2. Effective Transnational Cost of Capital with Tax Holidays

In section 4-1, I develop the cost-of-capital model by incorporating
international taxation into Mintz’s cost-of-capital model (1990). However, the cost of
capital in chapter 4-1 tends to underestimate the real cost of capital during tax
holidays, because the cost of capital developed in section 4-1 does not capture the
entire tax savings from tax holidays. The cost of capital derived from dynamic
optimization is a long-run equilibrium concept. The tax holiday is a temporary
exemption from corporate tax. The cost of capital during the tax holiday in section 4-
1 does not capture the temporary effect correctly, resulting from tax holidays. Given
the tax holiday toward U.S. subsidiaries, I will derive the real cost of capital for new
investment at time 0.

Let’s consider that the Korean government provides a 5-year tax holiday to the
U.S. subsidiary. When the U.S. subsidiary performs an investment at time 0, the cost

of capital at time 0 derived in section 4-1 is given by:

 

. 19°F» ' E 1
C: = =(1—Ag><6°-3‘3———.—°-+r1’)+25 (84)
‘10 0 Lo

As long as the capital assets invested at time 0 generate income afterward, we have to
take into account the future beneﬁts of the tax holiday. The cost of capital at time 0,
however, considers only the tax-holiday beneﬁt at time 0. It tends to underestimate
the beneﬁts of the tax holiday, because it ignores the future saving from the tax holiday

on capital invested at time 0. The capital invested at time 0 will be partly depreciated

76

during the next periods, but this capital will still generate income and enjoy the tax-

holiday beneﬁts after time 0. Thus we have to consider appropriate future tax savings

from tax holidays.

The cost of capital

 

 

 

 

 

3..
b-r
......................... d
K
E
l
5 time t

------- the cost of capital when we do not consider future tax savings

from tax holidays
—— the effective cost of capital when we consider future tax savings
from tax holidays

Figure 1. Cost of Capital for a New Investment under Tax Holidays

Figure 1 depicts the cost of capital for a new investment at each point of time, under

the assumption that a ﬁrm performs investment every year. The cost-of-capital line,

bd , is decreasing during a tax holiday, because the annual loss of depreciation

allowances is decreasing over time. The effective-cost-of-capital line, a , is

77

increasing during a tax holiday. Because the future savings from tax holidays may

decrease with time, the effective cost of capital is increasing during a tax holiday. The

difference between bd and 3 represents the present value of future savings from a
tax holiday at a point of time. I only calculate the cost of capital for a new investment
at time 0, namely point e. The cost of capital for a new investment at each time,
during 1971-1994 is used, for the comparison of costs of capital with or without tax

holidays.

Capital stock (K: )

 

 

r
I

5 time t

------- The path of capital stock when we do not consider future tax savings
from tax holidays

—- The path of capital stock when we consider future tax savings
from tax holidays

Figure 2. Capital Accumulation under Tax Holidays

78

Figure 2 shows the optimal path of capital stock. When we incorporate ﬁiture tax
savings from tax holidays into the cost-of-capital model, the new capital stocks are
adjusted upwards. The lower cost of capital leads to higher capital stocks, because the
production ﬁJnction is concave. When we consider future tax savings from the tax
holiday, the difference between A and B is the additional capital stock made at time 0.
Because the cost of capital is relatively low during tax holidays, the U.S. subsidiary
has an incentive to invest more during tax holidays. However, the loss of annual
depreciation allowances and the higher required rate of return21 during tax holidays
overwhelm future savings from tax holidays, so that the level of capital stock would be
decreased even during tax holidays.

To derive the cost of capital at time 0 correctly under the tax holiday, we

incorporate the present value of the tax holiday (0") into the cost-of-capital model by

using the speciﬁed time—variant parameters. The cost of capital under the tax holiday,

considering ﬁJture savings at time 0, Cf , namely the effective cost of capital, is given

by

C,” = C5" — Q” (85)

The present value of the tax holiday is the present value of the ﬁrst holiday term (0"?)

plus the second holiday term (52"; ). Since we are interested in the cost of capital at

2' The required rate of return ﬁnanced by the subsidiary’s new-share issues and local debt are higher
during tax holidays than is the required rate of return ﬁnanced by these methods at other times.

79

time 0, we set t = 0 for the next equation. The present value of the tax holiday is

given by:
0 h, 0 o hi+hz o 0
oh = je-U’ +5 "(C' - C,')ds + [W *“‘(C' — c; )d5 (86)
0 hi

When we arrange equation (84), we can rewrite it as follows.

 

(l _ e_(po+§oxhl,h1))Co _

_ 1_ -(p’+6')hl C.
p'+5' p'+5'( e )5”

 

 

(e-(puanh _e—(,C,-.(5-)(h.+h,))C;2 , (87)

where C ' is the subsidiary ﬁrm’s cost of capital excluding all tax incentives, and C , is
the subsidiary ﬁrm’s cost of capital during the tax-holiday periods, where the time
variable sis 171 or 71,. From equation (84), C ' — C ,' represents the difference between

the cost of capital without the tax holiday and the cost of capital with the tax holiday.

The present value of the tax holiday is viewed as the present value of the cost

reduction, which is the sum of C — C 5' during the tax holiday. When we assume the

continuous time discount factor, e“"°“‘", the beneﬁts of the tax holiday fall in value at
the rate of economic depreciation (6') and at the rate of discount, p' .
In specifying C ' , C 1., , and C ,3, , we do not need to consider the present value

of depreciation allowances and other tax incentives, because the cost of capital at time

80

0 already captures them. The subsidiary ﬁrm’s cost of capital excluding all tax

incentives, C ', is speciﬁed as the follows.

(1-0') .1; 1:,

 

 

 

C. 2 (l—u‘) (5 q; ’E’r')’ (88)
where O' = -[i'(l- u')—7r']a,' +(l +— 1_0.1_0" )a;
+{Jr'—i'(1—u°)+1_o_,l_0,,[i'(1-—a)'—a)")-—7r']}a3° . (89)

We get 9' by setting 2', k', and i”. to 0 into equation (58) in chapter 3.
When we incorporate time-variant tax parameters under tax holidays into the cost—of-

capital model, we can get the subsidiary ﬁrm’s costs of capital during tax holidays,

 

C; and cg,
o o o it. E o
C... =(1—®.>(6 ﬂigfw. >, (90)
h 9' — —"—— ' ”I. ' (91)
w ere '_-(l—u)az —(1—u)0(3
_ ; " E
. — (1 G.) (60 -@__—9_+r20), (92)

h‘ — (1—0.5u') qc', E0

81

l

where O; 2 (1+ O.5u°i~° — .. )a; + {77' —i°(l — 0.5u')

7
b

———.’—..1i'(1—w°—w">—zr']}a; (93)
1—0 —0'2

9], and 9; come from the cost-of-capital formula. The subscripts l and 2 indicate

the ﬁrst and the second holiday terms, respectively. By using the speciﬁed parameters

in each period, we can get the present values of the tax holiday. When we incorporate
all these values, such as C ' ,C 7., , C11,, and C ,ﬁ' , into equation (85), we obtain the

effective cost of capital under the tax holiday. The estimated results of the effective

cost of capital with the tax holiday are presented in Chapter 6.

82

CHAPTER 5

SOURCES OF DATA

5-1. Data Sources for Korean Firms
5-1-1. Corporate Tax Rates
Korea has progressive corporate tax rates, and the statutory corporate tax
rates have frequently varied over time. Before 1983, the statutory corporate tax rate
depended on categories of corporations and the amount of corporate income.
Domestic corporations were classiﬁed for tax purposes primarily as proﬁt or nonproﬁt
corporations. Proﬁt corporations were divided into listed or unlisted corporations on
the Korean Stock Exchange (KSE). Different progressive tax rates were levied on
corporate income in each of the categories of corporations. With the 1982 tax reform
(implemented in 1983), the classiﬁcation of proﬁt corporations as listed or unlisted
corporations on the KSE was eliminated.
The 1990 tax reform (implemented in 1991) simpliﬁed the structure of the
corporate tax system by eliminating the classiﬁcation of corporations. The statutory
tax rate was 34 percent on corporate income exceeding W100 million, otherwise it is

34 percent of corporate income below W100 million in 1991.

83

In addition to the basic corporate income tax, there are resident (or inhabitant)
and defense tax, which are surcharges applied on other incomes including the
corporate income. Since 1975, a resident surcharge was 7.5 percent of corporate
income tax liability, and a defense surcharge was 20 percent of corporate income
exceeding W500 million, otherwise it is 25 percent of corporate income below W500
million. The resident surtax was raised to 10 percent in 1996. The defense tax was
abolished in 1991. The maximum corporate tax rate including surtaxes reached 30.8
percent in 1996. The changes of statutory corporate tax rates over time are
summarized in Table 9.

A foreign corporation which has a permanent establishment in Korea is subject
to the same tax rate and the same taxation method as a domestic corporation. On the
other hand, a foreign corporation without a permanent establishment is subject only to

stipulated withholding taxes on its Korean source income.

5-1-2. Personal Tax Rates

We need three parameters associated with personal taxation to derive the ETR.
These are the tax rates on dividends, interest income, and capital gains. Different
withholding tax rates on dividend income are applied for resident shareholders and
nonresident shareholders. Dividend income was taxed at a ﬂat 20-percent rate from
1972 to 1974. Since the global income tax system was introduced in 1975, a taxpayer
has faced a marginal tax rate on his aggregated income, which is called “the base of
global income.” But a substantial part of dividend and interest income have been still

taxed under a ﬂat withholding tax. From 1991, the dividend income of domestic

84

shareholders received from listed companies on the KSE has been subject to a 20-
percent withholding tax, plus a resident surtax of 7.5 percent of the 20-percent
withholding tax. The withholding tax is 25 percent for dividends from unlisted
companies. These taxes apply regardless of whether the dividends are received by
majority or minority shareholders. Such dividends to majority resident shareholders
are also subject to global income tax, at progressive rates. In the case of dividends
from unlisted companies, even minority shareholders are subject to global income tax
as well as withholding tax. For nonresident individuals and corporations, the
withholding tax on their dividend incomes depends on the tax treaty between the
relevant countries. The tax treaty between Korea and the U.S. indicates that a
corporation which has at least lO-percent ownership is taxed at a lO-percent rate of
withholding tax, including the surtaxes. Otherwise, corporations are subject to a 15-
percent withholding tax on dividends.

For resident individuals and corporations, all interest income is subject to
withholding tax, plus a resident surtax. In addition, such income is also subject to
global income tax. But there exist various exemptions from global income taxation for
minority-interest income. For nonresidents, the withholding tax rate is described in tax
treaties. According to the tax treaty with the U.S., the withholding tax rate on interest
income is 10 percent.

The total withholding tax on interest income and dividends in Korea, from
1982 to 1990, was 16.75 percent, including the resident, defense, and education
surtaxes, which is calculated as 10(l+0.075+0.1+0.5). Yun (1988) estimated the

average marginal tax rates on dividend income and interest income under the base of

85

Table 9. Corporate Tax Rates in Korea

 

 

 

 

(unit: %)
Profit Corporations Non-Proﬁt Surtaxes
Corporations
Year Corporations Not-listed Top Rate of
Listed on KSE Corporations Resident Tax Defense Tax Corporation Tax
1972 27 40 35 0 0 27.00
1973 27 40 35 7.5 0 29.03
1974 27 40 35 7.5 0 29.03
1975 27 40 27 7.5 25 35.78
1976 27 40 27 7.5 25 35.78
1977 27 40 27 7.5 25 35.78
1978 27 4O 27 7.5 25 35.78
1979 27 40 27 7.5 25 35.78
1980 27 40 27 7.5 25 35.78
1981 33 40 27 7.5 25 43.73
1982 33 38 27 7.5 25 43.73
1 983 30 33 27 7.5 25 39.75
1984 30 33 27 7.5 25 39.75
1985 30 33 27 7.5 25 39.75
1986 30 33 27 7.5 25 39.75
1987 30 33 27 7.5 25 39.75
1988 30 33 27 7.5 25 39.75
1989 30 33 27 7.5 25 39.75
1990 30 33 27 7.5 25 39.75
1991 34 34 34 7.5 0 36.55
1992 34 34 34 7.5 0 36.55
1993 34 34 34 7.5 0 36.55
1994 32 32 32 7.5 0 34.40
1995 30 30 30 7.5 0 32.25
1996 28 28 28 10 0 30.80

 

 

 

 

 

 

 

 

Source: Korea Development Institute (1996), Facts and Figures on the Public Sector In Korea.
Ministry of Finance, Korean Taxation, various issues.

86

1.1V.
WI

Ld.‘ i

i
9.0
V

inA,
.lik‘

global income. These are 49.62 percent for dividend income and 57.27 percent for
interest income. According to Yun (1988), in 1986, some 43.8 percent of dividends
was subject to ﬂat withholding tax, and the rest was taxed on the base of global
income. Some 98.9 of interest income was taxed under flat withholding tax, and only
1.1 percent was taxed on the base of global income. Thus the effective tax rates on
dividend and interest income are weighted averages of the marginal income tax and
withholding tax rate. The effective tax rate on dividend and interest income are
calculated as [(effective withholding tax rate)(the ratio of taxable income subject to
withholding tax) + (average marginal tax rate)(the ratio of taxable income subject to
global income tax)]. For example, the effective tax rates on dividend and interest

income are computed as follows:
m"'=(16.65)*(0.438)+(49.62)*(o.562)=31.15%

m"=(16.65)*(o.989) + (57.27)*(0.011)=17.20%
I summarize the changes over time in the tax rates on dividend and interest income in
Table 10. Capital gains on the sales of shares by domestic-majority shareholders are
taxed as ordinary income, but those of listed companies on the KSE are generally
exempt from taxation. I use a O-percent rate on capital gains to derive the cost of
capital in Korea. Capital gains on the sale or transfer of shares for nonresidents are
subject to a withholding tax of 10 percent on sales value, or 25 percent of net capital
gains22 . For U.S. ﬁrms, I use the 25-percent tax rate as a statutory tax rate on capital

gains, because most U.S. ﬁrms do not list their stock on the KSE.

 

22 For simplicity, I assume that every taxpayer chooses 25 percent of net capital gains.

87

Table 10. Personal Tax Rates in Korea

 

 

 

 

 

 

 

 

 

 

 

 

(unit: %)
Year Dividend Interest Resident Defense Education Total Total Effective Effective
Withholding Withholding Surtax Sunax Surtax Withholding Withholding Dividend Interest
Tax Tax Tax for Tax for Tax Tax
Interest Dividends

1971 15 5 0 0 5.00 5.00 15.00 5.00
1972 20 5 O 0 20.00 5.00 20.00 5.00
1973 20 5 7.5 0 0 21.50 5.38 21.50 5.38
1974 20 5 7.5 0 0 21.50 5.38 21.50 5.38
1975 25/5 5 7.5 10 0 5.88 5.38 25.04 6.44
1976 25/5 5 7.5 10 O 5.88 5.88 25.04 6.44
1977 25/5 5 7.5 10 0 5.88 5.88 25.04 6.44
1978 25/5 5 7.5 10 0 5.88 5.88 25.04 6.44
1979 25/5 5 7.5 10 0 5.88 5.88 25.04 6.44
1980 25/10 5 7.5 10 O 5.88 5.88 28.34 6.44
1981 25/10 10 7.5 10 0 5.88 5.88 28.34 12.25
1982 25/10 10 7.5 10 50 11.75 11.75 31.15 17.20
1983 25/10 10 7.5 10 50 16.75 16.75 31.15 17.20
1984 25/10 10 7.5 10 50 16.75 16.75 31.15 17.20
1985 257'10 10 7.5 10 50 16.75 16.75 31.15 17.20
1986 25/10 10 7.5 10 50 16.75 16.75 31.15 17.20
1987 25./'10 10 7.5 10 50 16.75 16.75 31.15 17.20
1988 25/10 10 7.5 10 50 16.75 16.75 31.15 17.20
1989 25/10 10 7.5 10 50 16.75 16.75 31.15 17.20
1990 25/10 10 7.5 10 50 16.75 16.75 31.15 17.20
1991 25/20 20 7.5 0 0 21.50 21.50 33.82 21.89
1992 25/20 20 7.5 0 0 21.50 21.50 33.82 21.89
1993 25/20 20 7.5 0 0 21.50 21.50 33.82 21.89
1994 25/20 20 7.5 0 0 21.50 21.50 33.82 21.89
1995 25/20 20 7.5 0 0 21.50 21.50 33.82 21.89
1996 15 15 10 0 0 16.50 16.50 31.01 16.95

 

Source: Korea Development Institute (1996), Facts and Figures on the Public Sector In Korea.
Ministry of Finance, Korean Taxation, various issues.
Note: Dividend tax rate is (Majority shareholder/minority shareholder) from 1974 to 1979.

Since 1980, dividend tax rate is (unlisted company/listed company).

88

 

5-1-3. Depreciation Allowances

For the deprecation method, ﬁrms in Korea can choose between straight-line
and declining balance. The Korean tax law indicates that either the straight-line or the
declining-balance method is to be used for tangible ﬁxed assets other than plant and
buildings, and the straight—line method is to be used for plant, buildings, and intangible
assets. We assume that all ﬁrms choose the declining-balance method for machinery
and equipment, because it is more beneﬁcial than the straight-line method when the
ﬁrm purchases a relatively long-lived asset.

The present value of depreciation allowances depends on the depreciation rate
in the tax law and on the ﬁrm’s discount rate. Given the ratio of residual value to
acquisition value and tax lifetime in the tax law, the depreciation rate of the declining-

balance method can be calculated as follows.

residual value 1

6T. = 1— . . . L 94
(acqursrtron value) ( )

 

Before 1995, the ratio of residual value (salvage value) to acquisition value was set to
10 percent in Korean tax law. From 1995, the salvage value was set at 5 percent for

the declining-balance method and zero for the straight-line method. In the case of

declining balance, the present value of depreciation allowances, Z 8, is the following:

89

5T
p+67

 

Z, = j: ziaTe-‘5’*P>’ds = (95)

where T is the tax lifetime allowed in the tax law, and 6T is the depreciation rate in

the tax law.

In the case of the straight-line depreciation method, the investment capital is

depreciated for tax purposes by 1/ T per unit in each year, where T is a tax lifetime.

The present value of straight-line depreciation allowances, Z S , is given by

_ pf

) (96)

l—e

7' 1 _
Z, :10 (7)e “’ds =( pf

5-1-4. Parameters Associated With Price and Financial Data
For the price of capital goods, the Bank of Korea publishes data on the
composition of ﬁxed domestic capital formation in both current and constant price

levels. We can ﬁnd the implicit price deﬂator from ﬁxed capital formation data,

expressed in current and constant prices, by assuming that q, = q, — q,_1 . In addition,
the inﬂation rate is needed to calculate the cost of capital. I use the percentage change

of the consumer price index as a measure of inﬂation.
The proportions of investment ﬁnanced by retained earnings, borrowings, and
neW-share issues, are obtained from ﬁJnd-ﬂow statements in Financial Statement

Analysis, published by the Bank of Korea. The proportion of investment ﬁnanced by

90

retained earnings is estimated by the changes in retained earnings, depreciation
allowances, and other allowances. The proportion of investment ﬁnanced by
borrowings is estimated by the changes in external ﬁnancing, which is the sum of long-
terrn debt, short-term debt, bonds payable, accounts payable, and others liabilities.
Finally, the proportion of investment ﬁnanced by new-share issues is estimated by the
changes in capital stock. I assume that the ﬁrm maintains its optimal ﬁnancial
structure. Therefore, the proportions of investment ﬁnanced by retained earnings,
borrowings, and new-share issues are ﬁxed at the optimal level. I use the average
proportions of investment funds during 1971-1994. For the comparability of the
Korean classiﬁcation industry with the U.S. classiﬁcation of industry, I unify the
categories for the ﬁnancial source of investment.

For the interest rate, I use the average interest rate of total borrowing23 for
each industry in Korea. This interest rate is also obtained from fund-ﬂow statement in
Financial Statement Analysis. In addition, for the exchange rate, I use the market
exchange rate, which refers to period averages, which is from International Financial

Statistics.

 

 

23'I'Otal borrowing is estimated by the sum of long-term debt and short-tenn debt plus bonds payable,
less accounts payable.

91

5-2. Data Sources for U.S. Parent Firms
5-2-1. Corporate Tax Rates in the United States.

I assume that all corporate ﬁrms are subject to the top statutory tax rate at the
federal level (11,). I use the average tax rate for state and local taxation of corporate

income ( as) as a measure of the state and local corporate tax rate. Because state

corporate taxes are deductible from corporate income for the purpose of the federal

corporate tax, the combined corporate tax rate is estimated by the following.

it = u, +115 — urns (97)

The top corporate tax rates were 48 percent during 1971-1978, 46 percent during
1979-1986, 34 percent during 1987-1992, and 35 percent after 1993. The average tax
rate for state and local taxation during 1971-1986 is obtained from Jorgenson and Yun
(1991). Because they ﬁnd that the average state and local rate has a steadily rising
trend, and because they estimate the top corporate tax rate by 1986, I use 9 percent
during 1987-1990, and 9.5 percent during 1991-1993 for the average state and local

rate. The corporate tax rates are summarized in Table 11.

5—2-2. Personal Tax Rates
There are three parameters associated with personal taxation. The rates of

taxes on dividend income, interest income, and capital gains have been based on the

92

Table 11. Tax and Price Parameters for the United States

 

 

 

(unit 0/o)
Year Top Average Average Effective Tax "/6 Change of % Change o/o
Corporate Marginal Marginal Tax Rate on Price in of Price in Change
Tax Rate Tax Rate on Rate on Capital Gains Machinery and Structures of CPI
Interest Dividend Equipment
1971 50.9 28.4 47.5 14.0 3.1 7.2 4.4
1972 51.1 28.4 47.5 14.0 1.3 6.7 3.2
1973 51.0 28.4 47.5 14.0 1.7 7.3 6.2
1974 51.2 28.4 47.5 14.0 7.1 11.9 11.0
1975 51.4 28.4 47.5 14.0 13.1 10.7 9.1
1976 51.5 28.4 47.5 14.0 5.8 4.7 5.8
1977 51.6 28.4 47.5 14.0 5.6 7.5 6.5
1978 51.4 28.4 47.5 14.0 5.3 9.4 7.6
1979 49.6 28.4 47.5 14.0 6.1 10.5 11.3
1980 49.9 28.4 47.5 14.0 8.2 8.4 13.5
1981 50.2 25.8 39.6 11.8 6.8 12.2 10.3
1982 50.8 25.8 39.6 11.8 4.0 7.5 6.2
1983 50.7 25.8 39.6 11.8 0.4 -2.8 3.2
1984 50.5 25.8 39.6 11.8 -0.7 0.8 4.3
1985 50.7 25.9 31.9 10.9 0.2 2.4 3.6
1986 50.8 24.6 29.7 10.6 2.2 1.9 1.9
1987 38.3 24.5 29.6 13.8 1.1 1.6 3.6
1988 38.3 22.2 25.0 14.9 1.5 4.6 4.1
1989 38.3 22.7 25.2 14.3 2.0 3.7 4.8
1990 38.3 22.6 24.9 13.9 1.7 3.2 5.4
1991 38.6 22.9 25.7 13.5 1.6 2.4 4.2
1992 38.6 21.5 25.0 13.2 0.2 -l.2 3.0
1993 39.6 21.8 27.0 13.5 .01 3.2 3.0
1994 39.6 21.8 27.0 13.5 -1.1 3.3 2.5

 

 

 

 

 

 

 

 

Sources: See the text.

93

 

estimates reported by King and Fullerton (1984), Fullerton and Karayannis (1993), and
Feenberg (unpublished data). These estimates are outcomes of the TAXSIM model“.
5-2-3. Depreciation Allowances and Investment Tax Credits

Under U.S. tax rules, three cost-recovery methods have been employed for tax
purposes: straight-line, sum-of-year’s-digits, and declining-balance method. For
machinery and equipment, including public utility structures, the sum-of—years digits
method was used during 1954-1980, and the double-declining-balance method has
been used for assets in 3-, 5-, 7—, or lO-year classes since 1986. The straight-line
method has been used in structures since 1987. Gravelle (1994) summarizes the
depreciation methods from 1953 to the present for each asset. In the case of the
double-declining-balance method, the investment capital is depreciated for tax

purposes by 2 / T . The present value of the depreciation allowances is:

 

(98)

For the sum-of-year’s-digits method, the tax deduction declines linearly over the tax

lifetime. The present value of depreciation allowances is:

— 2
Z ijgQ—sle’p‘ds-z —-,-—[1——1,—-(l—e""7)] (99)
0 f‘ p T pf

 

3" See more details in Fullerton and Karayannis (1993. p. 344).

94

Table 12. Depreciation Method Rules in the United States.

 

 

 

Asset type Year Depreciation method
Equipment (including public utility structures)
1953 Straight line
1954-80 Sum-of-years digits
1981 Double-declining balance
1982-86 150% declining balance

Structures

1987-present

1953
1954-1969

1970-1980

1981-1986

1987-present

 

 

Double-declining balance

Straight-line

Sum-of-years digits

Nonresidential structures—
150% declining balance;

Residential structures—
Sum-of-years digits

175% declining balance

Straight line

 

Source: Gravelle (1994, p. 298)

95

 

We need two parameters to calculate the present value of depreciation

allowances. These are the tax lifetime, T, and the ﬁrm’s discount rate, p. Data on

the lifetimes of various ﬁxed assets over time are available from Jorgenson and
Sullivan (1981), and Gravelle (1994). I estimate the average interest rate of total
borrowings for the ﬁrm’s discount rate and interest rate. The average interest rate of
total borrowings is obtained by dividing the amount of interest payments and discount
fees paid by all corporations in each industry by total debt, except accounts payable
and other liabilities. The data for the estimation of the average interest rate are from
balance and income statement of returns of active corporations in Corporation Income
Tax Return, which is published by the Internal Revenue Service.

Hulten and Wykoff (1981) estimated the rates of economic depreciation for 34
categories of depreciable non-residential assets and one residential asset. I use the
weighted average of the rates of economic depreciation on 35 ﬁxed assets in each
industry. For the investment tax credit (ITC), I use also the weighted average rate of
ITC on various ﬁxed assets in each industry.

Data on capital stock weights are derived from the study of Jorgenson and
Sullivan (1981). By aggregating the proportions in their unpublished data, I obtain the
rate of economic depreciation, ITC, and tax lifetimes in the 29 industries, for 3 types

of asset in each industry.

96

5-2-4. Parameters Associated with Price and Financial Data

For the price of capital goods and the rate of inﬂation, I obtain the data from
the Survey of Current Business, published by the Bureau of Economic Analysis
(BEA). In addition, I estimate the proportion of investment ﬁnanced by retained
earnings, borrowings, and new share issues by using the data of balance sheets from

Corporation Income Tax Return, which is published by Internal Revenue Service.

5-3. Data Sources for Transnational Investment

Data on the ﬁnancial structure of U.S. investment in Korea are available from
U.S. Direct InvestmentAbroad, published by U.S. Department of Commerce. In
addition, I get the ratio of investment to total taxable base by using the data from U.S.
Direct Investment Abroad (Benchmark Survey 19 7 7, 1982, 1989). Because there
exist many missing data, I am forced to assume that all manufacturing industries have
the same ratio of investment to total taxable base, and the same ﬁnancial structure.

The U.S. subsidiary has 8 sources of ﬁnance for investments in Korea, as
shown in Table 7. Because there are restrictions on the ﬁnancial data for U.S.
subsidiaries, I assume that the parent ﬁrm wholly owns the subsidiary, and that there is
no local borrowing. The required rate of return for the subsidiary’s investment
ﬁnanced by new shares is the weighted average of the transnational investment
ﬁnanced by the parent’s retained earnings, debt, and new-share issues. The weights
are assumed to be the same as those found in domestic investment. These assumptions

enable us to calculate the U.S. subsidiary’s cost of capital on various sources of

ﬁnance.

97

ca
as:
“l
the

fec

can

0ft;

inﬂu

Chapter 6

Calculations of Effective Tax Rate and Cost of Capital

I calculate the effective tax rates (ETRS) at constant prices for different types
of assets and industries in Korea and the United States. I also estimate the cost of
capital at actual prices for the different methods of ﬁnancing and different types of
assets, and for 21 industries in Korea, the U.S., and for U.S. subsidiaries with or
without the tax holiday. Because my main concerns are the effect of tax incentives and
the required rates of return on investments in various industries and countries, we

focus on the estimation of the cost of capital.

6-1. Comparison of ETRs in Korea and the United States.

We have some problems with estimating ETRS under actual prices. ETRs are
very sensitive to the investor’s required rate of return and various price changes, as we
can clearly see in the formulas of the cost of capital. Moreover, these statistics have
been so volatile in the estimation period that we cannot distinguish between the effects
of taxation and various price changes from ETRS. If ETRs at actual prices are more

inﬂuenced by capital good prices and nominal interest rate, than by tax factors, it

98

W0

Ant

of:

ind

of

var

of

po

CO

116‘

CO

Ht

111

would be very difﬁcult to compare the tax policies between the two countries.
Another problem we encounter in estimating ETRS under actual prices is the difﬁculty
of determining the post-tax rate of return. Because a saver invests in various assets or
industries, the effective rate of return for the saver is the weighted average of the rates
of return prevailing in different industries. Even though different rates of return in
various industries are given, we scarcely determine the weighted average of the rates
of return for the saver because it is hard to get the weights data, which represent
portions of investment to industries. To overcome these problems, it is appropriate to
compare the ETRs at constant prices. I assume that all goods prices increase at 5
percent per year and interest rates are 10 percent.

The following ﬁgures allow a comparison between the tax policies of the two
countries. Firstly, Figure 4 shows that the Korean tax policies give preferential
treatment to machinery and equipment compared to structures. Figure 4 indicates that
the difference between the ETRS for machinery and structures was very high for 1973-
1981. This difference reﬂects the preferred tax treatment for key industries in Korea
during 1970s. Secondly, both Figures 4 and 5 exhibit the slow-rising trends of ETRS.
This can be explained by the elimination of tax preferences for key industries and the
curtailment of tax incentives over time. Thirdly, Figure 5 shows that the ETR in
transportation and public utilities is lower than the ETRS for any other sectors. This is
due to the high depreciation rate allowed in the tax law, and the relatively low
structure ratio in transportation and public utilities. The relatively high ETR of
construction in Korea is explained by the high share of structures in total assets,

because the tax beneﬁt for investing in structures in Korea is very small.

99

 

(ID

Figures 6 and 7 indicate the ETRs by assets and major industries in the United
States. The decline in the period of 1975-1980 for machinery and transportation
equipment is explained by the increase of the rate of ITC. A heavy decline of the
ETRS in 1981 is due to the adoption of double-declining-balance method of
depreciation accounting and the increase of the rate of ITC. The increase of tax
lifetimes and the abolition of ITC by the Tax Reform Act (TRA) in 1986 caused an
increase in the ETR afterwards. But the ETRs for machinery and equipment and for
structures since 1986 are still slightly low, compared to those that prevailed in the
period of 1971-1980. This can be explained by the corporate tax rate reduction in
1986. Unlike what we have in Korea, the ETRS for machinery and transportation are
higher than those for structures. Even though investments in structures do not qualify
for ITC (except for public utilities) and face a relatively low present value of
depreciation allowances, the rates of economic depreciation for structures are
strikingly low, compared to those for machinery and transportation equipment. The
effect of the low rates of economic depreciation for structures dominates the
disadvantage in tax position, thereby leads the low ETRs. The wholesale and retail
trade industry had the lowest ETR during the period of 1971-1985, because the share
of structures in wholesale and retail trade was high.

The comparison of ETRS in both countries shows that the ETRs in machinery
and transportation equipment in Korea were lower than those in the U. S. until 1980,
but the situation was reversed from 1981 to 1985. It also lets us know that the ETRS
in both countries tend to be stable after 1985. While the Korean government provided

generous tax incentives for key industries during the 1970s, the U. S. government

100

off:

198

(an:

the 1

rant

the l

offered benevolent incentives to promote overall investment during the ﬁrst half of the
19808.

The comparison of coefficients of variation in both countries enables us to
analyze the degree of tax discrimination across industries or assets. Figure 8 shows
that the special incentives for key industries in the 1970s increased the coefficient of
variation in Korea, and the introduction of double-declining—balance depreciation and
the increase of the rate of ITC during the period of 1981-1986 raised the coefficient of
variation in the U. S. But TRA 86 brought the coefficient of variation back down in

the U.S.

101

1441 I.:F.::.C.r ..: 77.223.022.35"

Effective Tax Rate

Ewe..." A. @2896 ﬂax 5:8 5 R25» aw. >83

 

c.ecoo .
325
938
988 u
9.38 ,
9.58
525
Secs .

Paco

 

 

,
/la\..l..l..\ _

 

 

cbooo

E:

q 11 q a A q in q — a q A A A a q .

$3 3.3 3.3 6.3 7:: 53 3mm .23 3%
<3...

Ill 2.23:3. 25 92639: 3 38:52:35
Isl jasmeozezo: min—=2: m: Zea—.3815

lxl 9:82.48 3 ZuzﬁmnElé

_

.3.

u

38

102

Effective Tax Rate

Ewe—.5 m. H2835 Hex 5:8 m: nae—.2. E. 33.8. =5:an.

 

 

 

1

 

bi

 

 

 

$3

33

b
1

Sum Gag 3.3 53 5mm Emu Emu Ewe

<9...
133% +335
lelzgemmoeaam lenoamﬁaomo:

Inlcssgo 23 Dog: Hag—o
IT magnom .

IOI Hanna—C0338: 25 page gang

p
U

52

Bow

103

1‘1 $343453": .3 :.W. zm-Scnwucnclzm

Effective Tax Rate

Ems—6 m. wanna: Hex 5:8 E :5 gm. aw. >82

 

obooc -

0.158

  

 

Page .

page i

Page .

 

104

cacao .

 

oboco -

98cc .

 

 

 

Cocccc .— d — q q 1 q q J u d u J4 a d A q q u 4 . q
3: 33 Sam 3.: .30 $3 38 Sam 53 .08 $2 $8

<2...

Ill 383.54% gm minivans" 5 CM. Enhancer—mum Iol ﬁgmconmao: may—€89: 5 cm. Zebﬁeoalsm
le Madonna 5 CM. inseam—0815

ill TC—Jxmﬁﬂm

Effective Tax Rate

mam—=6 4. H3825 Hun 5.8m :— :5 gm. 5% 3&8. =5:an

 

 

 

 

 

 

 

 

0.88 _ a A u _ 1 q a _ 4 {at d . q a a
.0 80mm: 3.3 Sum 33 3.3 _ m Emu im 33 Sue :5— .03
.o.~ooo

 

 

lol >m1o£38 Ba 335m

Ibl Zgggenam

lxl ﬁre—ago 25 H88: .730

IT madmoam

<2...

III 355m
le 025538
+ agave—Sac: 85 35:0 9:58

105

I‘filxllururu-I :- (nnuuo--A\- huh ?Av~ 9.05

Coefﬁcient of Variation

ans..." a. 032.83.: 3.19:2. i. EEG :— 28 ch. 2:. 7.2. .2

 

 

106

 

 

cacao
cases
.225 -
cuooc .
ill/IT} .
58o ﬂ]
C.Avccc '4 lq q _ u q u q q _ q u u - 4 u q q u . Jq d
3.: $3 3.3 2.3 3.3 .3: Saw Sam ES .23 5.: ES
<9...

IOI Goa—down... on. 559:2. m: :5 Cm. III 082.83.: 2. <u..§.o: E Wong

5-2. Comparison of Costs of Capital at Actual Prices.

We now compare the costs of capital in Korean domestic ﬁrms, U. S. domestic
ﬁrms, and U.S. subsidiaries with or without the tax holiday, at actual prices and
interest rates that existed from 1971 to 1994. The cost of capital in Figure 12 is a
weighted average of the costs of capital in all industries, with the weights given by the
capital stocks. Figure 12 gives the following observations. First, while the costs of
capital in Korean and U.S. domestic ﬁrms have similar trends, the cost of capital in
U.S. subsidiaries with or without the tax holiday has very different trends. Second, the
cost of capital in the U.S. is lower than that in Korea, except for a few years. The low
cost of capital in the U.S. is due partly to the low required rate of return. The low real
interest rate in the U.S. leads to the low required rate of return.

In addition, the tax incentives in the U.S. during the period 1981-1986
increased the difference of the cost of capital between the two countries. In some
cases, the lower cost of capital in Korea than that in U.S., developed due to relatively
high inﬂation in Korea (the real interest rate was even negative due to high inﬂation).
Third, the price changes in Korea seem to be more rapid, and more susceptible to
external shocks than those in the U.S. Figure 9 shows that the oil shocks each in 1973
and 1979 were ﬁilly transmitted to the price index in Korea. Such oil shocks increased
the price level at ﬁrst, and then raised interest rates afterwards. Figures 9 and 10
indicate that the increased price level caused higher interest rates after some time lags.
The price changes affect the cost of capital as follows. At ﬁrst, the increased price
level reduces the cost of capital by decreasing the real interest rates. Next, the

increased price level increases the nominal interest rates, and thereby raises the cost of

107

capital. The drastic ups and downs of the cost of capital in Korea show that the ﬁrst
effect is transmitted to the price level, and the second to the interest rate. In Figure
12, the lowest cost of capital in Korea was in 1974 and the cost of capital was
drastically increased in the following year.

In comparing the costs of capital between the U.S. subsidiaries with or without
the tax holiday, one interesting feature is that the difference between the costs of
capital in U.S. subsidiaries with or without the tax holiday ﬂuctuated over time. As I
discussed in chapter 5, the tax holiday has four effects on the cost of capital. One is
the reduction of the cost of capital through the tax cut. The second effect is that the
tax holiday reduces the present value of depreciation allowances and other incentives
signiﬁcantly. Third, the tax holiday has an effect on the required rate of return.
Fourth, the tax holiday affects the subsidiary’s taxable base. Even if the U.S.
subsidiary is exempted from corporate tax in Korea, it cannot avoid the home
country’s corporate tax burden.

We can also make the following observations from Figure 12. We can ﬁnd
that the cost of capital for the U.S. subsidiary with the tax holiday until 1981 was
greater than that of the U.S. subsidiary without the tax holiday. This can be explained
as follows. Tax incentives, including ITC and special depreciation allowances, were
great in Korea until 1981. Tax reforms in 1981 and 1982 which focused on a
structural adjustment limited or eliminated various tax incentives. Because the tax
savings from depreciation allowances and other generous incentives overwhelmed the
beneﬁt of the tax holiday before the tax reforms, the adoption of the tax holiday raised

the cost of capital. Since the tax reforms in 1981 and 1982, the difference between the

108

costs of capital in U.S. subsidiaries with or without the tax holiday has ﬂuctuated over
time. We can say that the tax reforms in 1981 and 1982 raised the tax savings ﬁom
the tax holiday. Since 1982, the cost of capital for the Korean ﬁrm has been greater
than that for the U.S. subsidiary with or without tax holiday. The reduction of tax
incentives raised the cost of capital for the Korean ﬁrm. The U.S. subsidiary also faces
the diminished tax incentives, but the subsidiary ﬁnances its investment with a
relatively low interest rate, and thereby faces a lower cost of capital than Korean ﬁrm.
Instead of eliminating tax incentives, TRA 86 in the U.S. reduces corporate tax rate
signiﬁcantly. This reduces the home-country tax burden, and the U.S. subsidiary still
has a cost advantage over the Korean ﬁrms.

Another feature is that the cost of capital for the U.S. subsidiaries is very
sensitive to the exchange rate. Figures 11 and12 show that the depreciation of the
dollar during 1980-1986 increased the cost of capital for the U.S. subsidiary, and that
the appreciation of the dollar during 1987-1989 (which was due to the ﬁrst trade
surplus in Korea) decreased the cost of capital. When the U.S. subsidiary remits its
income to the home country, the appreciation of the dollar makes the remitted income
more proﬁtable. The exchange rate causes the cost of capital for the U.S. subsidiary
to ﬂuctuate more than that for the U.S. domestic ﬁrm.

Figures 13-15 shows that the costs of capital in Korean domestic ﬁrms, U.S.
domestic firms, and U. S. subsidiaries with or without the tax holiday vary signiﬁcantly
to different ﬁnancing methods. Figure 13 shows that the cost of capital ﬁnanced by
retained earnings for the U.S. subsidiary with the tax holiday is the lowest, while it is

the highest for the Korean ﬁrms. This can be explained as follows. Because the cost

109

of capital ﬁnanced by retained earnings is independent of the exchange rate and the
repatriation tax, the cost of capital for the U.S. subsidiary depends only on the Korean
tax system. On the other hand, the U.S. subsidiary and the Korean ﬁrm face a
different tax rates on interest income. While the interest income for the U.S.

subsidiary is subject to a withholding tax of 12 percent, by tax treaty between the U.S.
and Korea, the interest income for the Korean ﬁrm is subject to an effective marginal
tax rate25 . Since 1981, the effective marginal tax rate for the Korean ﬁrm is greater
than the withholding tax rate by tax treaty for the U.S. subsidiary. The low
withholding tax rate leads to a low required rate of return, and the low required rate of
return decreases the cost of capital since 1981.

Figure 13 also shows that the cost of capital ﬁnanced by retained earnings
under the tax holiday is lower than that without the tax holiday. This can be explained
as follows. When the investment of the U.S. subsidiary is ﬁnanced by the subsidiary’s
retained earnings, the required rates of return in both cases are of the same level (the
withholding tax rate on interest income is not exempt in Korea, even during the tax
holiday). Since 1981, the tax saving from exemption of the corporate tax is greater
than the loss of present value of depreciation allowances and ITC.

Figure 14 indicates that the cost of capital for the U.S. subsidiary under the tax
holiday is much lower than that of the U.S. subsidiaries without the tax holiday, except
in some cases. When the investment of the U.S. subsidiaries is ﬁnanced by a loan from

the parent company, the cost of capital depends on the repatriation tax rate. The

 

2’ The effective tax rate on interest income in Korea was 6.44% during 1975-1980, 12.25% during
1980, 17.20% during 1981-1990, 21.890/0 during 1991-1995. See more details in Table 5-2.

110

repatriation tax on interest income is invariable for U.S. subsidiaries, with or without
the tax holiday. Because the corporate tax rate in the U.S. is greater than the
withholding tax rate on repatriated interest income, the eﬁ‘ective tax rate on repatriated
interest income is the U.S. corporate tax (u) in U.S. subsidiaries, either with or
without tax holiday.

The strikingly low costs of capital ﬁnanced by retained earnings and an
intercompany loan during 1981-1986, for the U. S. subsidiary with the tax holiday, can
be explained by the high interest rates during these periods, which increase the tax
savings from the tax holiday.

As Figure 15 indicates, another interesting result is that the cost of capital
ﬁnanced by new shares in Korea is very high, in spite of the existence of the income
deduction from new shares. The main reason is the extremely low tax rate on interest
income in Korea. Because the tax rate on capital gains is zero and the tax rate on
interest income is low in Korea, the opportunity cost of new shares is considerable.

When the investment of the subsidiary comes from the subsidiary’s new shares,
the repatriation tax rate on dividends varies. The repatriation tax rate under the tax
holiday is greater than that without the tax holiday, because the home country levies
the ﬁill tax burden on U.S. subsidiary at the rate of u during the holiday. Therefore,
the required rate of return is greater under the tax holiday. The increased required rate
of return, the loss of present value of depreciation allowances and ITC, and the
increased home-country tax base overwhelm the corporate tax reduction from the tax
holiday. Because the foreign tax credit is zero during the ﬁrst holiday period, the

home-country tax burden grows larger during the tax holiday. Thus, the cost of

111

IE

Cap
C01]
IEpa
i”co;
and 1

baa:

capital ﬁnanced by new share issues with the tax holiday is greater than that without
the tax holiday. The tax holiday increases the cost of capital ﬁnanced by new-share
issues.

Basically, if the U.S. subsidiary in Korea ﬁnances its investment entirely by
retained earnings or local borrowing, the U.S. subsidiary does not pay tax to the home
country, thus its tax burden only depends on the Korean tax system. However, if the
U.S. subsidiary ﬁnances its investment by new shares, the subsidiary’s tax burden
depends on both countries’ tax systems. In the case of new-shares ﬁnancing, if Korean
tax rates are high relative to U.S. tax rates, we should certainly expect for Korean tax
rates to have an effect on FDI. However, if Korean tax rates are low relative to U.S.
tax rates, the U.S. system of foreign tax credits should lead to a situation in which the
ultimate tax rate for a U.S. ﬁrm with a Korean subsidiary should not depend on the
Korean tax rate. However, there is one qualiﬁcation to this, and this is deferral. Since
deferral is an advantage to the ﬁrm, the Korean tax rate should still make a difference.
Therefore, we can say that the Korean tax system plays an important role in
determining multinational corporation’s tax burden for all kinds of ﬁnancing.

I have omitted two important issues in estimating the transnational cost of
capital. Firstly, the U. S. practices deferral of income of controlled foreign
corporations. The proﬁts earned in Korea by the U.S. subsidiary are not taxed until
repatriated to the U.S. as dividends. If the subsidiary defers the repatriation of its
income for a substantial period, the deferral would decrease the required rate of return
and the home-country’s tax base. The transnational cost of capital is overestimated,

because the cost-of-capital model does not consider the beneﬁts of tax deferral. Under

112

on}:
the

~13“

the tax holiday, the beneﬁt of tax deferral will be greater because the subsidiaries can
defer more proﬁts. The length of tax deferral affects only the subsidiary’s investment
ﬁnanced by new-share issues, because the investment ﬁnanced by retained earnings or
an intercompany loan is independent of repatriation tax on dividends. I assume that
the U.S. subsidiary remits its income immediately in the cost-of-capital model, because
it is hard to know the length of deferral for the subsidiary. To test my hypothesis, I
analyze the sensitivity of the cost of capital according to the length of deferral. In the
case of a U. S. subsidiary in a deﬁcient-credit position (in which case foreign tax
payments are less than the foreign tax credit), Figure 16 shows that the cost of capital
for the U.S. subsidiary’s new-share issues in a deﬁcient-credit position is sensitive to
the length of deferral. The cost of capital ﬁnanced by new-share issues decreases with
the length of tax deferral. This can be explained as follows. Because the U.S.
subsidiary in a deﬁcient-credit position can defer both the home-country’s repatriation
tax and the host-country’s withholding tax on dividends, the beneﬁts of tax deferral
increase with the length of deferral.

In contrast with the cost of capital in a deﬁcient-credit position, the cost of
capital in an excess-credit position is insensitive to the length of deferral. Because the
U.S. subsidiary in an excess-credit position (in which case foreign tax payments are
greater than the foreign tax credit) has no tax due to the home country, it can defer
only the host-country’s withholding tax on dividends. From Figure 18, we ﬁnd that
the beneﬁt of withholding-tax deferral is small. Because most subsidiaries in Korea

have been in an excess-credit position since Economic Recovery Tax Act (ERTA) 81

113

Kr

in the U.S., the assumption that the U.S. subsidiary remits its income immediately is
not strong.

Secondly, for the purpose of calculating foreign tax credits, U.S. ﬁrms with
subsidiaries in more than one foreign country are allowed to average across the
different subsidiaries, so long as the averaging is done within similar “baskets”.
Furthermore, excess credits are carried back or forward for a certain number of years
without interest under the current U. S. tax law. These excess credits give some value
to the U.S. multinational. These two features could have an effect on the effectiveness
of Korean tax policy, since the ultimate degree of taxation on a particular U. S.
subsidiary in Korea may depend on whether the U.S. multinational has subsidiaries in
other countries, as well as the tax rates in those other countries. However, I do not
incorporate these two features into the transnational cost-of-capital model, because it
is very difﬁcult to incorporate these into the model.

Figure 10 represents the weighted average of interest rates of 21 industries in
the two countries. It shows that U. S. ﬁrms have always faced lower interest rates than
the Korean counterparts, and the two countries follow very similar trends. A
comparison of the cost of capital in both countries indicates that the cost of capital is
more sensitive to the interest rate in Korea rather than that in the United States. The
required rate of return is in direct proportion to the interest rate. The interest rate
serves as a discount rate on the present value of depreciation. The high interest rate in
Korea can partly explain the high cost of capital in Korea.

Figures 18-20 indicate the costs of capital in Korea by industry. Because the

Korean government owns a large part of transportation and public utilities and it can

114

borrow

and pub

so high 1

capital it

Other ha]
to a relat
machiner
structure:
Pi
the Unite
Share of 51
industries
cost of ca]
about pric.
Subsidian'e

in KOrea.

borrow ﬁinds at a signiﬁcantly low interest rate, the cost of capital of transportation
and public utilities is relatively low. The ratio of structures to total assets in services is
so high that it produces a relatively high cost of capital. The signiﬁcantly high cost of
capital in petroleum and coal industry in 1984 is due to the high interest rate. On the
other hand, the transportation-equipment industry has faced a low cost of capital, due
to a relatively long and special tax treatments, and to a relatively high share of
machinery and equipment and transportation equipment (which have been preferred to
structures in Korean tax law).

Figure 21 indicates that the cost of capital in services is the lowest overtime in
the United States, because structures are more favored in the U.S. tax law and the
share of structures in services is high. The high cost of capital in chemical and allied
industries is explained by the high interest rate and the low share of structures. The
cost of capital in the U.S. subsidiaries in Figures 24-29 is very sensitive to assumptions
about prices and the exchange rate. Therefore, the cost of capital in the U.S.
subsidiaries has been ﬂuctuating in respond to the changes of price and exchange rate
in Korea.

A comparison of the ETRS between two countries shows that the ETRS in the
U.S. are lower than those in Korea, except in a few cases. While the Korean
government provided generous tax incentives for key industries during 19705, the U.S.
government offered benevolent incentives for overall investment during the ﬁrst half of
the 19803.

Our calculations of costs of capital at actual prices Show that the cost of capital

is very sensitive to price changes, interest rates, and the exchange rate. In addition,

115

the cost of capital varies signiﬁcantly to different ﬁnancing methods. The U. S.
subsidiary and domestic ﬁrm have a cost advantage over the Korean ﬁrm, because they
face lower interest rates and corporate tax rates. One interesting result is that the cost
of capital for the Korean ﬁrm ﬁnanced by new shares is really high, in spite of the
existence of the income tax deduction from new shares. While the beneﬁt of the tax
holiday is really great when the investment is ﬁnanced by retained earnings or an
intercompany loan, the investment ﬁnanced by new shares is penalized under the tax
holiday. Finally, one clear implication from the calculations of costs of capital is that
transnational investment is quite sensitive to both the home- and the host-country
taxes. The tax holiday provided by the host country affects the cost of capital for the
U.S. subsidiary, and credit-with-deferral method offered by the home country affects

the cost of capital for the subsidiary.

116

Consumer Price Index (%)

Ema—d e. .x. 0.5:? on $18 m: =8 Fm. 2:— R26»

 

Nu

No -

_u-

.c .

 

 

 

o
.cao

d

.04—

_equ .eum .euq .eqc _cm_ _emu _omu _cwu .owo _ec_

<o-q

I1 00:95.2. wane can.“ 5 W089 III moans—=2. vane :52 m: =5 C.m.

J

_ecu

_ceu

117

Interest Rate

ﬂuid a... saw—:2. ><2§me 3. .333... 5:8 .= :5 Fm. 2:. Renee

 

oboe

9.8.

9.2..

93o .

c..~c .

9.8

928

A

obac .

9?... ,

obwc

 

 

 

obcc

.3.

q q d — q 4 q d - q u u u u d ~ 4 a A q d _

.93 .93 3.3 .30 .ca. .cmu .03 .03 .30 .8. .03

<2...

to... $355.2. ><o3mo 0.. .228. .388 5 =5 Cm. Ill .5539. ><o~mma a». .228. 5:8 5 .98»

118

!.l.:. e\... he..m..mﬁ C~.~..G monorbzkgw zunﬁ

% Change of the Exchange Rate

Enid Z. .36.... 0.. =5 mung—ﬁe .38

 

.u

.c-

 

.3.
rm 4

Lo .

i—u .

 

...E

q

E:

.03

 

+ q q q a q

.00 .0... .03 .03

u

.03

.23

 

 

 

<9...

Ill .X. 0.5.50 0.. 3.0 98.50% ”0.0

119

Cost of Capi

Eng; 3. 00.....0...m0.. 0.. a... simian .55....«0
0.. 000.0 0.. 0......»—

9.5....

 

9.8.. L

 

 

 

 

d J/ + q a
BMW .33 .omu .omm .omu Bum

 

 

.c.~ccc

doc.

. on“.

 

 

<2...

LII 00m. 0.. 00......— ..0n ...0 .93.... 3....

00m. 0.. 003.0. .0. 2.0 Cm. 000.830 33.

i? 00... 0.. 0.6.8. ..0.. 50 Cm. 93.0.5

lul 00m. 0.. 03...... ..0a :5 Cm. 9......an 2.... .4... 30:00.5

120

Cost of Capital

05...... 5. 00500100.. 0.. 000.0 0.. 000...: 0.0002... 3. ”0.0.02. 003...“...

 

0.38
9.38 .
obocc
.
0.8.. .
9.58 -
— .
5.8.. .
.938 .

.obooc L

.9008 .

 

........

 

 

 

 

.938

<2...

III 000. 0.. 000.3. ..0n ...0 W080: 0.3.

00m. 0.. 000.8. 001.5 Cm. 00.08.... 0......

ix... 000. 0.. 000.8. .0. 90 Cm. 93.0.5

LT 08. o. 00...... .2 =5 0.0. 0500.5 .0... as. :25...

121

Cost of Capi

90......

50...... .0. 005000.000 0.. 000. 0.. 000.8. 300000.. .5. .00... 0.. ....0..00.....00v~

r000

 

90cc: -

9N2...

0.08..

.0
.98....

.9080 1

.902... L

 

.9mcoc

 

... .03 . . .0... .03

 

 

 

<2...

II... 000. 0.. 000.8. ..0. .00 K0800 0.00

000. 0.. 000.8. ..0. ...0 9m. 00.00. 0.0..

2...... 000. 0.. 000.8. ..0. ...0 Cm. $00.08....

[0| 000. 0.. 000.8. ..0. ...0 Cm. 9.5.0.03. 2.... .00.. =0....0<0

122

Cost of Capital

..1.m0..0 .m. 005000.000 0. 000. 0.. 000.8. 0.00002. 0.. 203.0003 .00000

 

. .80..

:52. 1

920.. 1

90...... 1

u
90...... 1

0.000.. m.

9.5.... 1
.0
.98.... 1

 

 

 

......................
.......

I
I
- s
:5:-

 

.. 0‘
a‘flfxv’
.l.

to N 8 q . krfrgf. d M. d . 1 u 1 q u q duo/{1. - J] a q .
0N0 .00”.

.00.. «m .03 .03 . .00. .03 .03 .00.. .000 .00.

 

 

19.5....

<00..

II: 000. 0.. 000.8. ..0~ .00 .9300 0.3.

000. 0.. 000.8. ..0.. .00 9m. 00.00000 0.3.

1...... 000. 0.. 000.8. ..0n .00 cm. 9.00.083~

In! 000. 0.. 000.8. ..00 .00 gm. 9.00.0.5 00.0 .00.. =0....0<0

123

Cost of Capital

0.05.0 .0. 0000.33. 0.. 000. 0.. 000.8. .0 0 0000.00.10.00: 000.000 0000.....00
.0 ...0 .1000... 0.. 50.0.0.0.

 

 

 

 

 

 

 

 

 

 

 

...uucc
Aywcoc 1.”... ......................... n4. .................. n“ u. ..v. .. .............................
93....
...: i u I I i I I H
{11x35§"’
0.0000 . 11.101.253.10. 1...!
if.5womiiiic..;¢1..$2.wm.1¢$...t...e$...¢(..w.m.::...3. .2913..."$5.113... 5

...—MOO 1. .... o 3:..33. 5. 1.3....
...—...... 1
.....u.... 1
c.ogc .— q d W q q d d d

c . N 0 . u 0 1. 0 0 .0

.1000... 0. 00.0.0.0.

.1011 000. 0.. 000.8. 0.000000 .... ...0 00.005.00.00 008.00.. 0.0.0.000
..1.... 000. 0.. 000.8. 0.000000 .... ....0 0000.00.00 88.00.0003. .1000
ix... 000. 0.. 000.8. 0.000000 .... ...0 00.00.05... 20.010003 .3000
III 20.00.00 ><0800 0.. 000.0 0.. 000.8.

‘124

Cost of capital

0.00.0 .... 0000......3. 0.. 000. 0.. 000.8. .0 00 0000001002... 000.000 0000.....00

.0 ...0 .1000... 0.. ..0..0....0.

 

 

 

 

 

 

 

 

 

 

 

 

 

90......
0.0000 I I I I n I I u u a
T 0 0 T OIIIIIIOIIIIIIQIIII? + A.
c. Ncg 4 WM b. at V0 vb Vb lb. 0. 95 .
93.... 1. ............................
9.0.... 1
9.5.... 1
90...... . . J . . _ . . .
c . N u a u 0 .0 0 0 .0
.1000... 0.. 02.0.0.2. 0010....

I0... 000. 0.. 000.8. 0.00002. .... .00 00.00.080.00 ”0.0.02. 0000.000
1*: 000. 0.. 000.8. 0.00002. .... ...0 0000.08.00 5.0.8.0003. .1000
., ., 000. 0.. 000.8. 0.00002. .... ...0 0000.08.10... 20070003 .00000
1.0.... $0.052. ><0800 0.. 000.0 0.. 000.8.

.125

Cost of Capital

0.0.00 .0. 000.0 0.. 000...: .0 .0300 ...~ 2.0.0.. .0000...~

 

93....

90...... 1

 

.......... 1

9........ -
.0

-... 5.... L

...
......
.

 

 

 

. a d a u 4 q a —

... .3”. /\ u ...... . . .00. .00. .33 .00.. .00.. .cc. .0...

 

 

 

 

....N......

000..

IQI 0 .0300 It... 70.0.00
...... zwﬂcmmonclam ......0..... 00.52.6030...
is! 2.6.80.0 00.. 06.0.. .800 lol 08000030000 00.. 002.0 C......00

IT. 002.000

126

Cost of Capital

0 .0000 .0. 000. 0.. 000...: .0 .0200 ..0.. 3000.00.00.00.

 

. .....c..

90...... I

90...... 1

90......

 

0.0000

 

 

 

0.0000-
...: .30 .3. .8. .30 .00.

 

 

.00. .00u .00.. .000 .00. .00.

 

 

....N......

IOI 0000 3.0.00 03000.0 000 00.008
......s. £000 000 .0800. .00.:0.00 00.0.8.0
It... 9.0.0.00. 000 >500 03000.0

1+! ”00.00. 000 0.0000 08000.0

<2...

11! 0000.0 70.... 2.008.. .0030: 000 0.00. 0000.0
2...... 0000... 00.0000. 00....0300 000 2:00 03000.0
IQI 00.3.00... 000 000. 03000.0

all 0.000. 0.0... 0.000 03000.0

127

0.0000 0... 000. 0.. 00.0.0. .0 .00—.00 ..00 2000.00.00.00 ..

92.0..

 

90...... 1

 

 

 

 

 

 

q

.00m

. a q u 17 u u

.00.. .000 .00. .000

 

 

0.0000
m
m
:0. 90...... 1
.0
0
C

0.800 .

90...... .

.0... .000 .000 .00.. .000 .00. .000
.9 .....c
000..
IOI 00.000. 7.0.0. ...l... 000000.00 7.0.0. 03000.0

.20... 0.00.000. 000 500.0000 £50500. ....xl 03000030000 0.00.0800.
....T. 0.00.. 3000.00.00.00

7.00.000... 00000. 500200.00
lol 50.2—000.0 000 ”0.0.00 03000.0

128

Cost of Capital

3%...» 3 name 3. 03—59. m: :5 Ch. 3. 33.2. 595.5.

 

 

 

 

 

d u q ‘

Emu $3 Ewe Em:

 

 

 

Iol >m1o:_88. $.2an Ea Ema—i
zgcmwnncaam

is! 230—835 25 W08: 2QO

1+... magnum

<2...

st... 795%
....3... 0035830:
I... 4.3833330: 8:. 3.2.0 9:38

129

Cost of Capital

39:6 MN. 0?: 3. 02:3— 5 =5 Fm. ...... 3335381..“ _

 

Aywucc

cbccc 1

 

 

 

 

:3— 5: 34m. $3 .3.» Em—

 

. q a u q _ A u d 4 d

Emu .23 3mg 3:3 3:: $3

 

 

.obuoo

I0... moon 8:58 30988 23 dwcmono
....3... $02. 2:. 5.362 man—:35 55:83
Ii! Organ»— 83 2:2— 52.:on

.....I ”:32 gm Sumac 30988

<02.

It... ﬂea—o 2:... 5633—. r352. 85 099. ﬂea—o
........x 363. 5.555. magma—5 25 >58 369.08
IOI $93.83 Ea now— 58:08.

l..| mnoao. 98.. 2:. 93mm 50988

130

Enid 3. new. 3. 02:8— 3:: =5 Fm. ...... Susanna-...?“ =

abocc

 

 

cbucc 4

 

 

 

 

 

 

m
m.
C
..m
n
0
C
c.o§ — u q a u 4 q a d d d d d d a A q d u d -
Eq— 53 Sam .34 3.3 Em— 35 33 Eng .23 :5— $3
193ch
<2...
IOI $3an gonn— III m 3518.2. 7.38— 5855 . . 330380. axon? 58.338.
....x ..... E8303: 2:— 98.330 mop—€38. it... ﬁngmuoawno: magma—=2: IOI 33:32.3 2:. ”organ $3988.

IT: 092 32.38815

13]

Cost of Capital

.0.N000

0.0000 .

0.~000 0

0.N000

0.500 ..

988 M.

0.88 0

0.0000 ..

_0
0.0000 .

-988 .

0.300 .

-0.N000 .

 

 

 

50:3 NA. 02... a. 9.0:». 3.. =8 0.0. macaw—$5. 5. 3&2. 3.95.5.

 

 

Iol guacamoncasm + Ooamgoag

.. .l IJ ..v I..- 1-! lildl’lllld 1! 34.0.1]- .i...IIl _..t .1". 1 ,I_ 1

$00 .03 $00 .000 .00.

<2...

€329.35 E:— ”03: H800 if- 02.308

 

|.\sl rl'l'lll

 

 

132

Cost of Capital

0.00.0 00 000. 0.. 00.0.3. ..0.. ...0 0.0. 00.5.0.2... .0 300030.015 .

0.0000

 

0.0000 -

0.0000 1

 

9008

0..000 .

 

 

 

0.88 .. KN A q . ..m 4 d
.0... .00/... M... .000 .03 . 0 .00.
.988 .

 

0.0000

A 4 .1 4 q u u d u d u d

.000 .000 .000 .000 .00. .000

 

 

IOI ...000 00.200 0000008 000 ...000000
2...-.. (<000 2.0 .0302 30.00.00 0.03.800
...xl 0.00—0.00. 20 >500 300008

IT. 000.000 2.0 0.0000 5.00008.

<02.

1:... 0.02:0 7...... >332. 002.000. 2.0 00.2. ...0x...0
.....x... .0002. .00—0.00. 00.0.0200 000 >500 030008
IOI 0030.000. 000 000.. 080003

ll. 0.000. 0.8.. 0.2.0 030003

133

0.0000 00. new. 0.. 9.0.8. ..0.. .00 Pm. 00.83.05. .0 7.000.000.0100 ..

0.0000

 

0.0000
0.0000

0.0000 . .

 

0.000 4

0. .000 -

 

0.0000 4

C.C8C c /’ q 1 1 q , q . q 0 a . q 0 0 u 0 a 0 q
.0... .. 4 000 .03 .000 .00. .000 .000 .000 .000 .00. .000
..

 

Cost of Capital

 

0.0000 .

0.000 1

0.000 .

 

 

0.0000

 

<02.

IOI 00.0.5. 7008. III. 000000.00 7.08. 030008 3000.003. 9800. 00.00.000.00
-..-..x 0.00300. ~00 0.00.330 0.00.0300. It! 08000030000 0000.00.00. lol .00.—000008 ~00 ”0.0.00 080008
IT: 00.0.. 700000008000

134

Cost of Capital

ﬂuid 3. 08. a». 0938. ..S. :5 Fm. gear—~93.

in. HE» :23an 3 3&2. 5::an

 

.°
g
1

 

E

F3
U
8
°
«c

obccc

0. Eco

 

 

 

 

988 .
3.:

d

Ed .93 5.3

IOI chmﬁncnam

5.3 $3 33 Sam Emu Ewe

<aun

...... Ooamnanaoa 2:29.55 Ea ”n8: .3on

d J- d -

$3 $3

lxl mansoom

135

Cost of Capital

ENE.» wa. now. a. 0938— m8. :5 Fm. wear—menu.
i=— Hmu mic—Esau m: awning—lam —

 

 

Aywcoc

      

AV—uoc .

 

Mm , >
Qwhw1§§1
(Q ,Q

Ay~coo L

     

Ayowoo 1

 

‘x 4 , g 1} ”V

W

5A..
988 ‘

~©Q~ ~equ _oqu ~cqq .oqe ~em— ~omu _owu ~omq uome ~9o— ~cow
Ayouoo

 

1 q 1 ﬂ 1%

 

 

 

<0...

lolmooa 3:58 $332.0. Ea Hug—coo III Asa—o 3E. >255. 599.. 2a 092 ﬂea—o
Jpl $03 Ba 5559. Seaman 5.35.8 lxl 56$. gnaw. 3:53am 25 >58 56988
Isl 0:238. 85 >an $395... IOI 33:95. 2:— Con— 38.88

+5.39 Ea Sumac $8953 lnl 928. 98» 98m $3988

136

Ems—d uc. new. on 6938. m2. :5 Ch. mucuamaJ.
in... den mac—Esau E 39533815" =

 

 

 

 

 

 

 

 

 

w
.m.
a
C
(I
0
u
0
C
x , \
. x l‘ l \
C 68° 4 t \_ «kt . IF.\
. 1 u u A u d u 1 d 4 _ q a A ‘ .‘ d 4 d
3.: $3 ' Ed 53 $3 $3 Gmu 5mm _ H . .. 32 G3
.988
5. 30¢
Kn...
I9: 35:5. gag: Illmwcaownoa gas.— wnoqcﬂm Ibl Ema—59% 98%» 982.038
lxl @838— 86 £03330 macmvaoa Ill dam—553:0: what—=9: IOI 53:..ng 25 ”£28 $39.03

IT 032 gangs—9:13

137

CHAPTER 7

DETERMINANTS OF FOREIGN DIRECT INVESTMENT

IN KOREA

The theoretical model derived in Chapter 5 shows that the tax policies of both the
host and the home countries can have a signiﬁcant impact on international capital ﬂows,
because tax policy is the one of the most important factors in measuring the cost of
capital. The study of the effects of taxation on foreign direct investment (FDI) is
important for analyzing capital ﬂows and the behavior of multinational ﬁrms. Most studies
of FDI, however, have focused on case studies of nontax factors in FDI. These studies
include Goldberg (1972), Agarwal (1980), Dunning (1973, 1981), Torrisi (1985) and Tsai
(1991). For more details about the determinants of FDI, see Singh and Jun (1995). They
give a clear review of the literature on the nontax determinants of FDI.

Empirical study focusing on the effect of taxation on FDI was pioneered by
Hartman (1984). Hartman (1984), and Boskin and Gale (1987) analyze the effect of
taxation on FDI, by estimating time-series equations of aggregate FDI in the United
States. Hartman’s regression results suggest that there is a positive relationship between

the after-tax rate of return received by foreign investors in the U.S. and ratio of F D1

138

ﬁnanced by retained earnings to GNP for the US, and a negative relationship between the
FDI-GNP ratio and the relative tax rate on foreigners compared to domestic residents.

Boskin and Gale (1987) reestimate Hartman’s equation using updated tax-rate and
rate-of-return series from F eldstein and Jun (1987). Although their estimated coefficients
on the response of FDI to the rates of return are somewhat lower than Hartman’s results,
they conclude that the results of Hartman are fairly robust.

Slemrod (1990) disaggregates the data on FDI into the United States by seven
major investing countries. He points out that the disincentive to investment caused by the
tax system was implicitly measured by an average tax rate in the previous studies. He
extends and updates a Hartman model of aggregate FDI in the US, by incorporating a
measure of the effective tax rate on new investment in the United States. He also attempts
to estimate the effect of both U.S. and home-country taxation on FDI. He ﬁnds that FDI
from Canada and France (which adopt the territorial system) is more sensitive to U.S. tax
rates than FDI from Japan and the UK. (which adopt the global system). His result
indicates that the investment ﬁnanced by retained earnings is, as proposed, not affected by
the home-country taxation in global system countries. However, his estimation result
reveals that the effect of the home-country taxation on FDI ﬁnanced by equity transfer26 is
not obviously more negative for global system countries than for territorial system

countries.

 

2" In case of FDI ﬁnanced by equity transfer, the country adopting a territorial system is independent of
the home-country taxation, because the proﬁts earned abroad are exempt from the home-country taxation.
Unlike the territorial system country, the global system country depends on the home-country taxation,
because the proﬁts earned abroad are taxed in the home country when they are repatriated as dividends.

139

Grubert and Mutti (1991) analyze the impact of the host-country taxation and
tariffs on the international distribution of real capital stocks, by using 1982 data on a
cross-section of 33 countries. They show that taxes and tariffs play an important role in
the investment decisions by U.S. multinational ﬁrms. They use the average effective tax

rate (t,) in the host country as a tax variable. The tax coefficient on l/te is 0.11, and is

particularly sensitive to the capital stock of the US subsidiary.

Cummins and Hubbard (1995) indicate that FDI data into the US, published by
the Bureau of Economic Analysis, do not distinguish between new capital investment and
acquisitions of existing assets. Because the tax treatments of the two forms of investment
in the United States are substantially different, empirical studies which ignore the
distinction of the two forms of investment fail to estimate the effect of taxation accurately.
Cummins and Hubbard use previously unexplored micro-data on multinational ﬁrm’s
overseas investment (1980-1991) to measure more precisely the effect of taxation on new
direct investment into the United States, and to analyze subsidiaries’ new investment
decisions. The authors consider tax incentives, and ﬁt a neoclassical model with tax
considerations to the data on the U.S. subsidiary’s investment in other countries. Their
results reject a simple speciﬁcation in which taxes do not inﬂuence investment. The
estimated tax effects are economically important.

Most of studies on the effect of taxation on FDI omit the home-country taxation
from the analysis, and do not incorporate simultaneous nontax determinants of FDI in the
regression model. In addition, most authors, including Slemrod (1991), do not
incorporate into their regression models the details of the international taxation, such as

the credit-with-deferral method and withholding taxes on repatriated dividend and interest

140

ch:

bet

income. The transnational cost of capital for multinational subsidiaries is relevant in
analyzing the effect of international taxation on FDI. In this chapter, I incorporate ETRs
of both the home and host countries, as well as various nontax factors into the regression
model. Furthermore, I incorporate international tax system on repatriated earnings into
the regression model, by using the transnational cost of capital for the U.S. subsidiary. On
the other hands, I disaggregate total FDI in Korea by manufacturing and services, and by
major capital-exporting countries, the United States and Japan. Our interest in this
chapter is to investigate the links between tax policy and investment decisions, and
between the cost of capital and investment decisions. By using the effective tax rates and
the costs of capital which I derived in Chapter 6, I attempt to estimate whether the cost of
capital or effective tax rate in either Korea or the U.S. have a signiﬁcant effect on the level
of FDI in Korea.

Most economic variables that exhibit strong trends, such as GDP, FDI, and the
wage level, are not stationary. This nonstationarity, which has a unit root problem, makes
economic interpretation difficult (see more details in Greene (1997)). When we linearly
combine the variables that have a unit-root problem and this linear combination is
stationary, we say that the variables are cointegrated with each other. However,
cointegration is not a serious problem in time series with a small number of observations,
because cointegration implies that there is a long-run relationship between variables. The
cointegration problem does not matter in our regression model, because we do not have
enough observations to consider cointegration

Before I run my regression model, I check for unit root problems, by using the

Dickey-Fuller test. The test result indicates that most variables have unit-root problems.

141

To circumvent unit root problems, I use the ﬁrst differences of variables. The ﬁrst-
differencing technique reduces or eliminates the unit root problems”. The empirical
results are obtained by ordinary least squares (OLS) technique. While heteroskedasticity
primarily arises in cross-section data, the regression model using time-series data usually
encounters serious autocorrelation problem. Another problem we encounter in estimating
the regression model is the existence of multicollinearity, in which dependent variables are
too highly correlated to allow precise analysis of their individual effect. I have serious
autocorrelation problem in the regression model for the analysis of U.S. direct investment
in Korea. I use the Cochrane-Orcutt method to eliminate this problem. I also ﬁnd the
existence of multicollinearity in my regression models. The most frequently used remedy
is to drop variables suspected of causing multicollinearity. In so doing, I present four or
ﬁve models for explaining the effects of various dependent variables precisely. This

procedure improves the efficiency of OLS technique.

7-1. Empirical model and results

Kim (1994) attempts to estimate the determinants for FDI in Korea. He
incorporates various nontax factors, such as GDP, wage level, the number of strike. The
statutory corporate tax rate is used as a proxy for the Korean tax policy. His result shows

that the Korean tax policy is not sensitive to FDI in Korea.

 

27 Let’s suppose a nonstationary variable, y,. I assume that y, = a + yH + v, , where a is a constant

term and V, is a white noise. By iterating substitution, we get y, = 2(01 + V,) . This equation will be
i=0

eventually be an inﬁnite effect. When I use the ﬁrst difference of y, instead of y, , the ﬁrst difference is

stationary, and has a ﬁnite effect, such as y, — yH 2: a + V, . For more details about unit root, see

Dickey and Fuller (1981) and Engle and Granger (1987).

142

 

Ahn (1994) estimates the effect of the corporate tax rates in the home and host
countries on FDI inﬂows. He ﬁnds that the statutory tax rates in Korea has a signiﬁcant
effect on total FDI inﬂows, but not for the weighted average of the statutory rates in the
U.S. and Japan.

Kim (1994) and Ahn (1994) omit the role of tax incentives and the cost of capital
in determining FDI inﬂows. As Slemrod (1990) indicates, the average tax rate or the
statutory tax rate is not appropriate to capture the effect of capital cost and tax incentives.
Because we are interested in explaining the effects of tax policy and the actual cost of
capital on F D1 in Korea, we will use the effective tax rate at constant prices, and the cost
of capital at actual prices. ETR at constant prices represents only the effect of tax policy
in the host country or the home country. The cost of capital under actual prices is a
comprehensive measure, which represents the effects of rates of return, the ﬁrm’s ﬁnancial
structure, and taxation in both the home and the host countries.

We use data from 1971 to 1994. The speciﬁcation of dollar values is in logs, so
that the coefﬁcients are elasticities. The estimated regression equation is speciﬁed as

follows:

Aln FDI, = ,3, + awn GDP, + ﬁZAETR, + ,6,AC0C, + ,64Aln WAGE,

+ ﬂSATARIFF, + ,66ACREDIT, + ,B7Aln EX, + ADM/[84, + ﬂgASTRIKE, + e, , (98)

where FDI is the arrival of foreign direct investment, GDP is gross domestic product,
ETR is the effective tax rate, CDC is the cost of capital, WAGE is a wage index, TARIFF

is a measure of openness on trade and capital market, CREDIT is credit rating, which is

143

related to political risk and economic stability, STRH<E is the number of industrial
disputes, EX is the relevant exchange rate, and DUM84 indicates the dummy variable,
proxying government regulation.

Equation (98) will estimate the impacts of Korean tax policy and the cost of capital
for all industries, or separately for manufacturing and services, during the period of 1971-
1994. When we estimate the impact of Korean tax policies on total FDI, we use the wage
level in Korea as a wage index, not the wage differential between Korea and capital-
exporting countries. Because many countries invest in Korea, it is difficult to get the
accurate weighted average of the wage differential between Korea and capital-exporting
countries. Equation (98) can be modiﬁed for the estimation of the effect of tax policy and
the cost of capital in capital-exporting countries, namely, the United States and Japan, by
using data related to the countries. For the U.S. and Japan, we use the wage differential
between the two countries, and the relevant exchange rate in the regression model. We
also estimate the effect of taxation on the U.S. subsidiary operating in Korea. The
transnational cost of capital for the U.S. subsidiary captures the effect of both the host-
and the home-country taxation. In addition to ETR and the cost of capital, I incorporate
other nontax variables in regression model. The rationales for the various independent
variables are as follows:

(a) Gross Domestic Product (GDP) is used as a proxy of the market size. This
would reﬂect the so-called “market-size hypothesis,” which assumes a positive relationship
between FDI and the expected sales of foreign subsidiaries in the host country. Wheeler

and Mody (1992) ﬁnd market size to be signiﬁcant.

144

(b) The effective tax rate in Korea (KETR) is a measure that summarizes the
effects of the Korean tax system, including tax incentives toward foreign capital. The
effective tax rate in the U.S. (USETR) summarizes the U.S. tax system. To capture only
the effects of domestic taxation, KETR and USETR are calculated at constant prices.
While KETR is expected to inﬂuence the U.S. direct investment into Korea negatively,
USETR is expected to have a positive effect. When the tax burden in the U.S. increases,
U.S. investors tend to invest more abroad.

(c) The cost of capital in Korea (KCOC) captures both the effects of capital cost
and the Korean tax policy. It is also used to capture the effects of economic stability. We
expect a negative association between KCOC and FDI inﬂows. The cost of capital for the
U.S. subsidiary (KUSCOC) captures both the home- and the host-country taxation
simultaneously. We also expect to have a negative association between KUSCOC and
U.S. direct investment in Korea. As the cost of capital in the U.S. (UCOC) increases,
U.S. ﬁrms would reduce domestic investment and increase foreign investment. Thus,
UCOC is expected to inﬂuence U.S. investment into Korea positively.

(d) The wage rate should be negatively related to FDI inﬂows. That FDI takes
advantage of cheap labor in developing countries is well recognized. Many studies
suggest that there is a positive relation between the magnitude of FDI and wage
differentials between the host and home countries (Agarwal, 1980). While we use the
wage level in Korea as a measure of labor cost in the regression model for total FDI, we
use the wage differential in the regression model for the U.S. direct investment or

Japanese direct investment in Korea. We expect to have a negative relationship between

145

the wage level and FDI in Korea. On the other hand, the wage differential between the
U.S. and Korea should be positively related to FDI in Korea.

(e) The degree of openness on the trade and capital markets is expected to be
negatively related to FDI inﬂows. As the degree of openness on the trade and capital
markets in Korea increases, the subsidiary operating in Korea would increase the supply
for the Korean market and raise the volume of its trade, thereby would increase FDI into
Korea. I use the ratio of taxes on international trade and transactions, such as tariffs and
customs, to total government revenue in Korea, as a proxy for the degree of openness. As
the Korean government is more liberalized, the rates of tariff and customs duties may tend
to decline, and the ratio of taxes on international trade to total revenue of Korean
government drops.

(f) Government regulation of FDI should be negatively related to FDI ﬂows.
Because it is very difﬁcult to index, I use the tax reform in 1984 as a dummy variable. The
tax reform in 1984 adopted the negative system, which was devised to increase FDI. We
expect that the reform would be positively related to FDI inﬂows.

(g) We use two indexes as a proxy of risk. These are an index of country risk, and
the number of industrial disputes. The country risk is an index, which summarizes the
political risk and various economic indicators, including commercial and monetary factors.
The measure of country risk used here is similar to as a credit rating. The E UROMONE Y
publishes an annual series for country risk for various countries. The level of country risk
decreases as the scale of index rises. Therefore, the index would be positively related to

FDI inﬂows. The industrial disputes lead an economy toward instability, and are expected

I46

to be negatively related to FDI inﬂows. The data source is “Yearbook of Labor
Statistics,” published by International Labor Office.

(h) The exchange rate plays an important role in determining the decision of
investment and the value of repatriated proﬁts. A high dollar may lead to a high rate of
U.S. investment into Korea, and a low repatriation rate. Thus, the exchange rate is
expected to be positively related to FDI inﬂows

To improve goodness of ﬁt, I change the current values of FDI, GDP, and the
wage into constant values, on the basis of 1990 price level. Table 13 presents the
estimation results for total FDI in Korea from 1971-1994. The coefficients of KETR and
KCOC are correctly signed, and are not differ signiﬁcantly from zero in each model.
Although the t statistics on KETR does not have signiﬁcant effects on FDI, the magnitude
of tax coefficients on KETR, which are -3.253 in model 1, and -3.231 in model 2, are
substantial. Thus, we can say that the Korean tax policy has an inﬂuence on FDI,
although its effect is small. This story does not works well in case of the cost of capital.
The cost of capital does not have a signiﬁcant effect on FDI. The results in Table 13 show
that GDP has a positive effect on FDI, and rising labor costs and tariffs on FDI have a
negative effect. The coefﬁcients on GDP in all model have a signiﬁcant effect on F DI.
The coefﬁcient on GDP in model 2 is 7.819. This indicates that an 1-percent increase in
the ﬁrst difference of GDP is associated with a 7.819-percent increase of the ﬁrst
difference of FDI. The value of t statistic shows that GDP has a very signiﬁcant effect in
attracting FDI in all models. The coefficients on tariffs and the wage level in Korea have
the expected negative signs, and the dummy and credit variables also have the expected

positive signs. However, these variables do not have signiﬁcant effects on FDI in all

147

Table 13. Estimation Results for Total F D1 in Korea

 

 

 

Model 1 Model 2 Model 3 Model 4 Model 5
Intercept -0.491 -0507 -1.278 -1290 -0.562
(2.010) (2.142) (-2.778) (-2.607) (2.334)
GDP 7.308 7.819 18.800 18.438 8.054
(2.627)" (3.092)” (3.495)“' (3.075)” (3.053)”
TARIFF .0111 -0.092
(1.228) (0.734)
KETR -3.253 -3231
(1.275) (1.294)
KCOC 0737 -0.708 -0.674
(0.356) (0.313) (0.570)
WAGE -2.659 -4.817 -5231 -3.024
(1.197) (1.244) (1.232) (1.347)
STRIKE 0.005 -.0003 0.004 —0.674
(0.517 ) (-0.367) (0.202) (0.250)
DUM84 0.232 0.179 0.256
(1.454) (0.481) (1.555)
CREDIT 0.013 0.022 0.014 21.352
(0.506) (0.471) (0.275) (0.853)
R2 0498 0.490 0.485 0.497 0.450
Durbin-Watson 1.828 1.793 1.976 2.053 1.807
Statistics

 

 

 

Note: KETR and KCOC are the effective tax rate and the cost of capital in Korea, respectively. Credit is
the credit rating in Korea. FDI is measured in millions of dollars, and GDP is measured in billions of
dollars, and WAGE is measured in dollars. The numbers in parenthesis are t statistics.

"* is signiﬁcantly different from zero at the 1% level.

"”" is signiﬁcantly different from zero at the 5% level.

" is signiﬁcantly different from zero at the 10% level.

148

Table 14. Estimated Results for Manufacturing of Total FDI

 

 

 

Model 1 Model 2 Model 3 Model 4
Intercept -O.581 —O.509 -0.498 —0.3 17
(-1.3l6) (-1.123) (-1.101) (-0.755)
GDP 6.846 7.479 5.924 4.029
(1.242) ( 1.299) (1.072) (0.838)
TARIFF 0143 -0.126 -0.126
(-l.222) (-1.078) (-1.087)
KMETR -2.125 -2.041
(0.700) (0.676)
KMCOC -1.790 -1.944
{-1.041) (-1.339)
MWAGE -l.569 -1.087
(0.405) (-0.281)
STRIKE 0.002 0.012 0.002
(0.764) (0.633) (0.130)
DUM84 -0.063 -0.014 —0.047
(-0.191) (-0.041) (0141)
CREDIT 0.019 0.026 0.177
(0.430) (0.057) (0.605)
R2 0.182 0.155 0.163 0.094
Durbin-Watson 1.995 1.999 1.870 2.125
Statistics

 

 

Note: KMWAGE is the wage level of the Korean manufacturing. KMETR and KMCOC are the effective
tax rate and the cost of capital for the Korean manufacturing, respectively. The numbers in parenthesis

are t statistics.

"* is signiﬁcantly different from zero at the 1% level.
** is signiﬁcantly different from zero at the 5% level.
* is signiﬁcantly different from zero at the 10% level.

149

 

V2

3
38‘.

models. In case of tariffs, the wage level, and dummy variable, the magnitudes of these
coefﬁcients are substantial and the signs on the coefficients are correct in all model, thus
we can say that the hypotheses of the degree of openness , wage level and government
regulation are weekly supported. Finally, the credit rating has the expected sign, but is not
statistically signiﬁcant. The number of strike in model 4 has neither the expected sign nor
a signiﬁcant effect on FDI.

We divide total industry into manufacturing and services, and test several
hypotheses. As Table 14 indicates, we ﬁnd that the results for the manufacturing sector
do not deviate from those for total FDI in Korea seriously. Although the effects of GDP
and ETR is not strong as an total FDI, we can still argue that GDP and ETR in Korea
have effects on FDI, because the scales of the coefficients are consistently substantial.

We get the expected signs for most variables, except for the strike and the dummy
variables. The regression models are performed well in the service sector. From models 1
and 3 in Table 15, ETR has a signiﬁcant effect in attracting FDI. The effect of GDP on
FDI in the service sector is very strong. The coefficients on credit rating and dummy
variable have the expected signs, but is not signiﬁcantly different from zero at a 90-percent
level of conﬁdence. The coefficients of credit rating and government regulation in Table
15 show the facts that FDI in the service sector has been more restrictive than that in the
manufacturing sector in Korea, and that the credit rating is likely to be more important in
the service sector. On the other hands, we get the expected signs for tariffs and wage, but

the effects on FDI are not statistically signiﬁcant in the service sector.

150

Table 15. Estimated Results for Service of Total FDI

 

 

 

Model 1 Model 2 Model 3 Model 4
Intercept -0.917 -l.674 -0.973 0.003
(2.010) (3.164) (2.051) (0.017)
GDP 14.286 20.733 14.556
(2.639)" (3.248)". (2.675)"
TARIFF .0095 -0.105
(0704) (1.074)
KSETR -l6.203 -15.597 -10494
(2.119)" (1.990)‘ (2.397)“
KSCOC 2.566
(1.375)
SWAGE -0.268 -l.361
(0.268) (0.314)
STRIKE -0.01 1 0.041
(0.627) (0.196)
DUM84 0.797 0.114 0.330
(0.204) (0.362) (1.217)
CREDIT 0.048 0.035
(0.908) (0.708)
R2 0.502 0.433 0.495 0.368
Durbin-Watson 1.938 2.108 1.973 2.035
Statistics

 

 

Note: KSWAGE is the level of wage for services in Korea, KSETR and KSCOC are the effective tax rate

and the cost of capital for services in Korea, respectively.
The numbers in parenthesis are t statistics.
*** is signiﬁcantly different from zero at the 1% level.
** is signiﬁcantly different from zero at the 5% level.
* is signiﬁcantly different from zero at the 10% level.

151

 

The cost of capital does not have a signiﬁcant effect on FDI in all industries. The
possible explanation is as follows. The cost of capital is a comprehensive measure, which
are considering tax policies, inﬂation, and the rates of return. Although the tax policy has
a signiﬁcant inﬂuence on FDI, other factors can offset the effect of the tax policy. The
number of strikes are not operating well as a measure of risk, because they are correlated
with 86 Seoul Asian Games and 88 Seoul Olympic Games, which have the opposite effect.
Therefore, the coefficient on STRIKE in all models is not sensitive to FDI.

From the analysis, effective tax rates (ETRS) have the expected signs on FDI in all
regression models. It is important to know what tax rate or tax policy is sensitive to the
calculation of ETR. Table 16 summarizes the correlation matrix of ETR. It allows us to
know which factors have a relatively strong effect in calculating the ETR. The average
marginal tax rate on interest income has the strongest effect on the ETRS in both
manufacturing and services. Special depreciation allowances and the investment tax credit
play an important role in reducing the ETR for the manufacturing sector, but not for
services. This is because the special depreciation allowances and investment tax credit are
provided mostly for machinery and equipment, and because the share of machinery and
equipment is considerably higher in manufacturing. The income deduction from new

shares is an important factor to reduce ETR for services.

152

Table 16. Correlation Matrix of ETR

 

 

 

Types of Tax and Tax Incentive KETR KMETR KSETR
Corporate Tax Rate 0.3339 0.2673 0.3749
Average Marginal Tax Rate on Dividends 0.5163 0.5730 0.4997
Average Marginal Tax Tate on Interest Income 0.9814 0.9668 0.9897
Present Value of Depreciation Allowances Including -O.680l

Special Depreciation in Total Industry

Present Value of Depreciation Allowances Including -0.7212

Special Depreciation in Manufacturing

Present Value of Depreciation Allowances Including -0.0068
Special Depreciation in Services

Investment Tax Credit in Total Industry -0.4567

Investment Tax Credit in Manufacturing -0.5365

Investment Tax Credit in Services -0.0074
Present Value of Income Deduction -O.3615 -0.2799 -03973

from New Shares

 

 

 

 

 

Note: KETR is the effective tax rate in Korea.

KMETR is the effective tax rate in Korean manufacturing

KSETR is the effective tax rate in Korean services

153

lI

Suppose that a Korean ﬁrm in the manufacturing sector purchases machinery, and
that its investment is ﬁnanced by retained earnings in 1994. In this case, an l-percent
increase of the corporate tax rate is associated with a 0.310-percent increase in the ﬁrm’s
ETR, and an l-percent increase of the special depreciation allowances decreases the ﬁrm’s
ETR by a 0.589 percent. For a ﬁrm in the service sector, an l-percent increase of the
corporate tax rate is associated with a 0.326-percent increase in the ﬁrm’s ETR.

It is also important to analyze the effect of corporate tax and selective tax
incentives on FDI in Korea. I set up a new regression model to test the hypothesis that
Korean tax incentives and the corporate tax have a signiﬁcant effect on the level of FDI in

Korea. The estimation regression is speciﬁed as follows:

AlnFDI, : bo +b,ACORP, +b2AITC, +b3ASPECIAL, +b4ANS, +63, (99)

where CORP is the corporate tax rate in Korea, ITC is the investment tax credit,
SPECIAL is the present value of special depreciation allowances per dollar, and NS is the
income deduction from new shares. While the selective tax incentives are expected to
have positive effects on FDI, the corporate tax rate is expected to have a negative effect
on FDI. Table 17 shows that all variables have the expected signs, and that both the
corporate tax and the tax incentives in Korea have signiﬁcant effects in attracting FDI into
Korea. In particular, the coefficients on the investment tax credit and the corporate tax

rate are signiﬁcantly different from zero at a 95-percent level of conﬁdence in all models.

154

Table 17. Effects of Corporate Tax Rate and Tax Incentives on FDI

 

 

 

 

 

 

 

Model 1 Model 2 Model 3 Model 4
constant 0.068 0.055 0.068 0.067
(0.947) (0.779) (0.988) (1.023)
Corporate Tax Rate -7 .466 6.899 -7.412 -7.684
(2.591)" (2.449)" (2.708)" (2892)"
Investment Tax Credit 40.598 39.880 29.798
(2.884)" (3.075)” (2.095)"
Special Depreciation 76.525 39.915 46.268
Allowances (2.602)" (1.218) (1.507)
Income Deduction 165.84 108.00 117.48
from New shares (2.771) ' (1.759)‘ (1.965)
R2 0.503 0.514 0.549 0.600
Durbin-Watson Test 1.823 1.963 2.019 1.881

 

155

 

The major capital exporting countries in Korea are the United States and Japan.
For the United States, we calculate the cost of capital across 21 industries, and various
ﬁnancial structures. In addition, we have also estimated the cost of capital for U.S.
subsidiaries in Chapter 6. I obtained the data on FDI in Korea from the Ministry of
Finance and Economy. These data are only on new investments, and do not include the
investment ﬁnanced by retained earnings. Due to the limitation of data availability, we
cannot distinguish the effect of tax policy on FDI between retained earnings and equity
transfers. If we had the disaggregated data on FDI by retained earnings or equity
transfers, we could analyze the effects of the various ﬁnancial methods on FDI, by using
the cost of capital ﬁnanced by retained earnings, an intercomany loan, or new-share issues.
Therefore, we use aggregated FDI data, and test whether the tax policies in both Korea
and the U.S. have a signiﬁcant effect in attracting U.S. direct investment into Korea. The
cost of capital for the U.S. subsidiary is a comprehensive measure, which takes into
account the tax policy and user cost of capital in the home and host countries. We also
estimate what is the most important factor in attracting the U.S. direct investment.
Because the value of the Durbin-Watson (DW)28 statistic is very high in the regression
model for U. S. direct investment in Korea, there is an autocorrelation problem in the
regression model. I use the Cochrane-Orcutt method to eliminate autocorrelation. Table
18 shows that signiﬁcant factors in determining the U.S. direct investment are tariffs, the
wage difference between the U.S. and Korea, and the dummy variable. Especially, the

tariff variable has a very signiﬁcant effect on the U.S. direct investment in three models.

 

28 The most widely used for test of autocorrclation is the Durbin-Watson (DW) test. The value of the DW
enables us to know whether there is an autocorrclation problem. The problem of the DW test is the
existence of a no-conclusion area of autocorrclation.

156

The tariff variable works well in the regression model, as a proxy for the degree of
openness on trade. The coefﬁcient of the dummy variable, proxying government
regulation on FDI is 0.438 in model 2, and is signiﬁcantly different from zero at the 5%
level. The coefﬁcients on the dummy variable in models 3 and 4 are substantial, although
these variables are not signiﬁcantly different from zero. Therefore, we can say that the tax
reform in 1984 has an inﬂuence on FDI. The wage differential between two countries also
has the expected positive sign, and is a signiﬁcant effect on FDI. The exchange rate the
expected positive signs in all models. The exchange rate between the U.S. dollar and the
Korean won was ﬁxed during 1975-1979. The ﬁxed exchange rate for 5 years is an
obstacle to estimate the effect of the exchange on FDI correctly.

Unlike in the earlier results, GDP does not have the expected sign. The market-
size hypothesis does not work for U.S. direct investment in Korea. According to Table
18, both the effective tax rate and the cost of capital in the U.S. have the expected positive
signs on the U.S. direct investment in Korea, and the effective tax rate in Korea also has
the expected negative sign in all models. However, the costs of capital for the U.S.
subsidiary and Korean ﬁrm do not have the expected signs. We conclude that neither U.S.
nor Korean tax policy has a great inﬂuence on FDI. There are a few possible explanations
for the lack of ﬁtness of tax policy and the cost of capital. One is the possibility of transfer
pricing, which is used to reduce tax burden by U.S. multinationals. If U.S. multinational
corporations adjust transfer prices to raise or lower the proﬁts they report to the U.S.
government, the reported after-tax proﬁt rate should be negatively correlated with the tax
rate. Hines and Rice (1994) show that a l-percent increase in tax rates reduces the

reported proﬁtability of U.S. subsidiaries by 2 percent. Therefore, the actual cost of

157

capital for the U.S. subsidiary engaging in transfer pricing would be much lower than the
cost of capital under theoretical models. Secondly, the credit-with-deferral method in the
U.S. allows the U.S. subsidiary to defer taxation. If the U.S. subsidiary defers the
repatriation of its proﬁts indefinitely, it renders the home-country taxation insigniﬁcant.
Thirdly, the U.S. foreign-tax-credit system does not allow for tax sparing. Without a tax-
sparing system, the effects of tax incentives provided by the host country would be
substantially offset by the home-country taxation on repatriated income. Finally, there is a
limit to the availability of data that are by industry and by the source of funds. Table 18
represents the estimated results for U.S. direct investment in Korea.

Japan is the one of the largest investing countries in Korea. Because the
calculation of ETR and the cost of capital in Chapter 6 is only for Korea and the United
States, we estimate the effects of tax incentives and cost of capital in Korea on Japanese
direct investment. The results in Table 19 show that the coefﬁcients on all variables,
except the strike and the dummy variable of Japanese direct investment, have the expected
signs on FDI. In addition to real GDP, the Korean tax policy, and the tariff play
signiﬁcant roles in attracting Japanese investment. Unlike the US, the dummy variable
neither has the expected positive sign, nor being also statistically signiﬁcant. As far as the
exchange rate, its coefﬁcient is positive, as proposed, but not signiﬁcant. As can be seen

in Table 19, the regression models for Japanese direct investment ﬁt and work well.

158

 

Table 18. Estimated Results for U.S. Direct Investment in Korea

 

 

Model] Model 2 Model 3 Model 4 Model 5
Intercept 0040 0.198 0154 0056 0210
(0.687) (0.731) (1.298) (0.649) (1.563)
GDP -5.604
(1.637)
TARIFF 0162 —0.056 0120 0118
(2.633)" (0.946) (2.068)" (2.243)"
KETR 0910 -2.400
(-0.362) (-0.843)
USETR 0.770
(0.473)
KCOC 0.347
(0.239)
USCOC 1.965
(0.690)
KUSCOC 0.488 1.465
(0.382) (1.012)
USKWAGE 1.356 2.868 2.679
(1.126) (2.224)" (1.829)‘
STRIKE 00005 0009 0.006 0.002
(0.001) (0.799) (0.520) (0.200)
EXUS 1.758 0.005
(1.394) (0.004)
DUM84 0.458 0.254 0.171
(2.388)" (1.686) (0.861)
CREDIT 0.008
(0.324)
R?- 0.483 0.635 0.371 0.467 0.563
Durbin-Watson 1.721 2.081 1.744 1.641 1.723
Statistics

 

 

Note: USETR and USCOC are the effective tax rate and the cost of capital in the U.S., respectively.
KUSCOC is the cost of capital for the U.S. subsidiary operating in Korea, and EXU S is the exchange rate,

in terms of a dollar. The numbers in parenthesis are t statistics.
*** is signiﬁcantly different from zero at the 1% level.
** is signiﬁcantly different from zero at the 5% level.
* is signiﬁcantly different from zero at the 10% level.

159

 

Table 19. Estimated results for Japanese Direct Investment in Korea

 

 

 

Model 1 Model 2 Model 3 Model 4
Intercept -l.430 -1.416 -1.564 -1.314
(-2.788) (2.907) (2.713) (2.724)
GDP 18.501 18.318 19.693 19.221
(3.711)” (3.934)” (3.565)” (3.413)”
TARIFF 0162 0152 0150 -0.126
(1.483) (1.784) (-l.468) (1.249)
KETR .10.233 -10.105
(2.524)" (2.643)"
KCOC 0909 0910
(0.430) (0418)
WAGE1 -2.136 -2.338 -3.249 -4.681
(0.535) (0650) (0.752) (1.249)
STRIKE 0.007
(0.377)
EXJ 1.470 1.313
(1.037) (0.849)
DUM84 0043
(0. 138)
CREDIT 0.028 0.024
(0.724) (0.515)
R2 0654 0.653 0.507 0.489
Durbin-Watson 1.953 1.947 2.164 1.982

Statistics

 

 

Note: II also experimented with the wage difference between Japan and Korea, but the wage difference is

insensitive to Japanese investment ﬂows in Korea.

EXJ is the exchange rate in terms of a Japanese yen. The numbers in parenthesis are t statistics.

*“ is signiﬁcantly different from zero at the 1% level.
** is signiﬁcantly different from zero at the 5% level.
* is signiﬁcantly different from zero at the 10% level.

160

 

Models 4 and 5 in Table 19 show that the wage level in Korea have an inﬂuence on FDI,
although it is a small inﬂuence. We can not reject the hypotheses that Japanese direct
investment during the 1970s and 19808 centered on labor-intensive projects, and most
Japanese direct investment tends to exploit cheap labor costs in Korea, by the middle of
1980. The estimation results in Table 19 also show that Japanese direct investments are
also very sensitive to Korean tax policy. This can be explained as follows. Both the U.S.
and Japan adopt the foreign-tax-credit system. The major difference between two capital-
exporting countries is whether tax sparing is allowed. Since Japan allows tax sparing,
Japanese ﬁrms have taken more advantage of the tax savings provided by the Korean
government. While U.S. ﬁrms are insensitive to tax incentives provided by the Korean
government, Japanese ﬁrms are sensitive to the change of tax policy in the Korean
government. Our results conﬁrm that Japanese ﬁrms have take advantage of the tax
saving from tax incentives, relative to U.S. ﬁrms.

In this chapter, I have examined the effect of taxation on FDI in Korea, by
incorporating both the home— and host-country taxation into a regression model. There
are three distinct results in this chapter. First, the results of this empirical approach
support a negative effect of Korean tax policies on total FDI. In addition to the market
size, the wage level, and the tariff have a inﬂuence on FDI in Korea. ETR in Korea, which
is used to estimate the Korean tax policy, has a signiﬁcant effect on FDI. However, the
costs of capital for the U.S. subsidiary, the U.S. parent, and the Korean domestic ﬁrms do
not have effects on FDI in Korea.

Second, while both the home- and the host-country taxation have no signiﬁcant

effect on the U.S. direct investment in Korea, the host-country taxation is negatively

161

CC

mills

related to Japanese direct investment in Korea. This result implies that the international
tax system which a country adopts is an important determinant of F D1 in Korea. Although
the U.S. and Japan adopt the credit-with-deferral system for their subsidiaries, the major
difference between the two countries is whether tax sparing is allowed. Since Japan
allows tax sparing, Japanese subsidiaries enjoy the tax savings from tax incentives
provided by the Korean government. Thus Japanese direct investment is sensitive to the
Korean tax policies. On the contrary to Japan, the U.S. direct investment in Korea is not
sensitive to the Korean tax policies.

Finally, GDP and tariffs in most models are signiﬁcantly related to FDI in Korea.
This estimation result implies that capital—exporting countries, including the U.S. and
Japan, have invested in Korea for the purpose of supplying the Korean market, and
importing raw materials from third countries and exporting their products to third
countries.

The results in this chapter provide an analytical guide for both the home- and host-
country’s tax policies. My estimation results indicate that the Korean tax policy has an
effect on total FDI. The international tax system is also an important factor in determining
F DI. In ﬁiture study, if we incorporate the beneﬁts of tax deferral and transfer pricing into
calculations of the transnational cost of capital, we may get more valuable and accurate
econometric results. On the other hands, my regression models do not explain the tax
effects of the various ﬁnancial methods. The more sophisticated ﬁnancial data and

disaggregated data will allow us to analyze the effect of taxation on FDI correctly.

162

CHAPTER 8

CONCLUSIONS

Most of studies focusing on the cost of capital have omitted the international tax
policies between the host and home country. Although some authors29 incorporate
international taxation into the cost of capital framework, their studies are limited to the
calculation of the transnational cost of capital ﬁnanced by equity capital. This paper
develops the transnational cost-of-capital model for the U.S. subsidiary operating in
Korea, taking into account various transnational ﬁnancial methods and international
taxation. It also considers all tax beneﬁts prevailing in Korea, with a particular emphasis
on tax holidays, which are one of the popular incentives to foreign investments. The
existing studies on the transnational cost of capital assume that both the home and host
countries have the same tax base, and that the exchange rate is determined by the price
level between the two countries, and that the interest rate is equal to a ﬁxed value (such as
5 percent or 10 percent) in all countries. These assumptions can have misleading effects
on the calculation of the cost of capital. Therefore, I incorporate the real data on the
exchange rates, and interest rates, and different tax bases into the cost-of-capital model.

On the other hand, most authors focus on either the calculation of the cost of
capital or on econometric study of the effect of taxation on FDI. Unlike those authors, I

carry out the two analyses in a consistent framework.

163

Our calculations of the cost of capital at actual prices show that the cost of capital
is very sensitive to price changes, interest rates, and the exchange rate. In addition, the
cost of capital varies signiﬁcantly to different ﬁnancing methods. The U.S. subsidiary and
domestic ﬁrm have a cost advantage over the Korean ﬁrm, because they face lower
interest rates and a lower corporate tax rate. One interesting result is that the cost of
capital for the Korean domestic ﬁrm ﬁnanced by new shares is really high, in spite of the
existence of the income tax deduction from new shares.

Most authors conclude that equity ﬁnancing is more costly than any other
ﬁnancing. However, this is not the case for the U.S. subsidiary operating in Korea. The
results show that the cost of capital for the U.S. subsidiary ﬁnanced by new-share issues is
not much higher (and is sometimes even lower) than those ﬁnanced by retained earnings,
local borrowing, and an intercompany loan. This is due to the gap between the prevailing
interest rates of the two countries.

Most authors underestimate the beneﬁt of tax holidays, because their models do
not fully capture the tax saving from tax holidays. I extend the cost-of-capital model to
capture the tax saving from tax holidays correctly. I ﬁnd that the cost of capital under the
tax holiday varies signiﬁcantly to different ﬁnancing methods. While the beneﬁt of the tax
holiday for the investment ﬁnanced by retained earnings or an intercompany loan is very
substantial, the investment ﬁnanced by new shares is penalized under the tax holiday.

Finally, one clear implication from the calculations of the cost of capital is that the
transnational investment is quite sensitive to both the home- and the host-country taxation.

The tax holiday provided by the Korean government affects signiﬁcantly the cost of capital

 

29 See OECD (1991), Mintz (1993), Mintz and Taipoulos (1994). and Jun (1995).

164

for the U.S. subsidiary. When we incorporate the beneﬁt of tax deferral offered by the
United States into the cost-of-capital model, the length of deferral has a substantial effect
on the cost of capital ﬁnanced by the subsidiary’s new shares when the subsidiary is in a
deﬁcient-credit position.

My estimation results in Chapter 7 indicate that the Korean tax policy has an effect
on total FDI, the service sector, and Japanese direct investment, but not on the U.S. direct
investment in Korea. The only signiﬁcant factors in determining the U.S. direct
investment are government regulation on FDI and tariffs. The tariff variable work well in
my regression models, as a proxy of the degree of openness on trade. This estimation
result implies that capital-exporting countries, including the U.S. and Japan, have invested
in Korea for the purpose of supplying the Korean market, and importing raw materials
from third countries and exporting their products to third countries. The estimation
results also imply that the international tax system is also an important factor in
determining FDI.

The results in this paper provide an analytical guide for both the home— and host-
country’s tax policies. In ﬁiture study, if we incorporate transfer pricing and risk into
calculations of the transnational cost of capital, the calculation of the cost of capital will
become a more relevant measure, as a proxy of multinational corporations’
competitiveness. Firm-speciﬁc FDI data would allow us to analyze the effect of taxation

on FDI adequately.

APPENDIX

APPENDIX

Figure 3. Cost of Capital as the Ratio of Investment to
Taxable Income Varies in Each Financing Method

 

 

 

 

3
‘a
8
“5 0.15 «~
iii"
0
0.05 ~r
0.00 a #1 L 4 + , , f,

 

 

1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0
The ratio of investment to taxable income

—0— Subsidiary's cost of capital financed by new-share issues

+Subsidiary's cost of capital ﬁnanced by retained earnings

+Subsidiary's cost of capital ﬁnanced by intercompany loans
+Weighted average of the subsidiary's cost of capital

166

Table 20. Classification of Industry

 

 

 

Industry
1 Agriculture, Forestry, and Fishing
2 Mining
3 Manufacturing
4 Food Related Products and Tabacco
5 Textile Mill, Apparel, Leather, and Other Textiles
6 Wood and Lumber, Including Furniture
7 Paper, Printing, Publishing, and Allied Products
8 Chemical and Allied Products
9 Petroleum and Coal Products
10 Rubber and Plastic Products
11 Stone, clay, and Glass Products
12 Primary Metal
13 Fabricated Metal Products
14 Machinery except Electronics
15 Electrical and Electronic Equipment
16 Transportation Equipment
17 Instruments and Related Products
18 Other Manufacturing
19 Construction
20 Wholesale and Retail Trade
21 Transportation and Public Utilities
22 Services

 

 

167

 

Table 21. Economic Depreciation Rates in the U.S. and Korea by Asset

 

 

 

 

 

 

 

 

 

 

United States Korea
Machinery and Transportation Machinery and Transportation
Equipment Ermipment Structures Equipment Equipment Structures
1 14.20 9.26 4.39 12.09 21.89 2.38
2 15.95 25.33 3.89 16.51 19.17 3.33
3 12.29 25.42 3.27 12.89 20.71 3.42
4 12.20 21.04 3.38 12.87 23.30 3.34
5 12.23 28.43 3.43 11.60 29.18 3.33
6 12.27 21.85 3.32 11.81 20.62 3.45
7 12.27 29.51 3.46 11.37 28.51 3.41
8 12.26 27.24 3.23 13.36 19.19 3.40
9 12.25 19.84 3.36 12.74 24.89 3.44
10 12.36 28.32 3.35 12.42 28.24 3.52
11 12.34 25.17 3.25 12.82 24.73 3.40
12 12.27 22.84 3.24 12.31 14.04 3.46
13 12.35 28.73 3.19 13.27 29.78 3.46
14 12.34 27.66 3.32 14.28 31.26 3.44
15 12.45 28.90 3.33 13.86 31.21 3.42
16 12.38 24.92 2.83 13.96 26.88 3.40
17 12.47 24.52 3.24 15.16 32.63 3.53
18 12.22 28.44 3.32 14.19 28.94 3.44
19 16.62 22.99 2.55 16.72 24.94 2.75
20 12.29 27.54 2.44 13.88 27.61 2.47
21 11.79 25.59 3.45 13.60 16.42 3.04
22 12.99 20.61 2.31 14.86 27.49 2.47

 

168

 

Table 22. Weight of Financing Methods

 

 

 

 

United States Korea U.S. Subsidiary
Retained New Retained New Retained Intercompany New
Earnings Debt Shares Earnings Debt Shares Earnings Loan Shares
1 51.11 41.11 7.78 40.89 56.05 3.06 N/A. N/A. N/A.
2 50.87 44.79 4.33 37.13 52.10 10.77 N/A N/A. N/A.
3 42.68 53.95 3.37 41.21 53.12 5.67 54.67 8.01 37.33
4 42.54 54.42 3.04 32.44 63.16 4.39 54.67 8.01 37.33
5 45.11 50.93 3.95 36.88 57.98 5.14 54.67 8.01 37.33
6 44.28 51.22 4.51 19.28 75.65 5.07 54.67 8.01 37.33
7 47.67 47.61 4.73 35.07 58.65 6.28 54.67 8.01 37.33
8 44.75 51.25 4.00 37.04 49.18 13.79 54.67 8.01 37.33
9 54.43 39.00 6.57 31.51 62.67 5.83 54.67 8.01 37.33
10 45.87 49.59 4.54 39.71 56.15 4.14 54.67 8.01 37.33
11 45.62 47.00 7.39 46.86 48.9 4.24 54.67 8.01 37.33
12 28.87 55.89 15.24 53.07 43.38 3.55 54.67 8.01 37.33
13 43.85 49.58 6.57 32.28 61.19 6.53 54.67 8.01 37.33
14 41.27 49.49 9.25 37.81 54.64 7.56 54.67 8.01 37.33
15 37.01 59.13 3.86 44.15 49.78 6.07 54.67 8.01 37.33
16 32.68 59.27 8.04 32.02 60.69 7.29 54.67 8.01 37.33
17 45.66 50.53 3.81 31.16 61.88 6.96 54.67 8.01 37.33
18 37.13 53.61 9.26 28.18 65.49 6.34 54.67 8.01 37.33
19 48.88 46.07 5.05 17.03 74.68 8.29 63.84 -1.06 37.22
20 33.72 62.14 4.14 21.28 72.78 5.94 63.84 -1.06 37.22
21 35.75 57.37 6.88 72.54 26.26 1.20 63.84 -l.06 37.22
22 38.32 57.24 4.44 47.49 46.85 5.65 63.84 -1.06 37.22

 

 

 

 

 

 

 

 

 

 

169

 

Table 23. Interest Rates in the U.S. by Industry

 

 

 

1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982

1 5.84 5.59 6.24 7.05 6.94 6.88 6.83 6.95 8.06 9.13 10.35 10.22
2 5.97 6.42 7.11 7.59 8.33 7.84 7.35 8.58 9.53 9.78 11.56 11.50
3 6.98 6.98 8.05 9.17 8.59 8.46 8.33 8.98 10.29 12.08 14.37 14.52
4 7.19 7.43 8.16 9.38 8.81 8.67 8.52 9.19 10.82 11.95 12.67 13.22
5 8.31 7.33 9.47 11.50 10.36 9.90 9.43 10.59 13.13 14.12 16.19 15.07
6 6.65 6.32 7.27 9.22 8.23 8.05 7.87 8.55 9.41 10.15 14.94 12.41
7 7.24 6.82 7.65 8.17 8.07 7.89 7.72 8.15 8.57 10.34 11.03 11.89
8 6.55 6.57 7.28 7.47 7.84 7.92 8.00 8.13 9.84 10.89 12.26 14.17
9 7.88 8.53 9.65 9.29 7.38 8.36 9.34 8.80 8.88 10.53 15.97 16.92
10 6.95 6.74 7.48 9.55 9.22 8.70 8.18 8.81 10.69 12.52 15.02 14.87
11 6.98 6.84 7.48 8.30 8.51 8.27 8.02 8.14 9.97 11.25 13.40 14.63
12 6.22 6.50 7.55 8.90 8.53 8.49 8.45 9.68 10.18 12.71 14.66 14.55
13 7.50 6.97 7.91 9.42 8.81 8.45 8.09 9.22 10.69 12.94 14.10 14.26
14 7.42 6.96 7.98 9.20 9.08 8.79 8.50 9.44 10.36 13.15 14.87 14.13
15 6.84 6.71 8.28 10.93 9.71 9.60 9.49 10.64 12.89 15.74 17.45 17.61
16 2.91 2.82 3.50 3.98 4.04 3.70 3.35 3.97 5.06 5.77 6.54 6.43
17 6.58 6.79 7.91 9.91 8.85 8.49 8.13 8.89 11.26 12.50 15.47 15.20
18 8.83 8.38 9.72 12.33 11.92 10.75 9.59 10.00 11.87 15.92 16.31 16.85
19 6.02 5.78 6.30 7.43 7.64 7.22 6.81 7.26 8.74 10.28 11.65 11.54
20 6.68 6.57 8.05 9.87 8.67 8.20 7.73 8.52 10.39 11.77 12.59 11.79
21 5.75 5.84 6.37 6.95 7.13 7.20 7.27 7.83 8.37 9.52 12.60 10.85
22 6.42 6.21 6.74 7.99 7.81 7.49 7.18 8.05 9.13 10.38 11.34 10.99

 

 

 

170

._.""‘"-.'7

kg“. z-..

 

Table 23 (cont’d)

 

 

 

 

 

1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994
1 9.46 9.96 9.23 8.77 7.67 8.02 8.57 8.27 7.52 6.84 6.14 6.05
2 10.09 9.77 9.29 9.57 8.03 8.55 9.43 9.94 9.26 7.80 7.68 8.00
3 12.93 12.91 12.13 11.19 10.28 9.82 11.81 10.99 11.46 9.66 9.11 8.97
4 12.36 13.50 12.74 13.26 10.80 9.27 8.47 9.82 11.92 10.12 8.44 8.80
5 11.88 12.25 11.55 10.31 9.94 13.55 12.14 12.86 12.29 9.89 8.53 9.24
6 10.28 9.59 9.94 10.22 10.99 9.20 10.87 9.88 9.73 9.04 7.31 8.85
7 11.10 11.12 10.48 11.02 9.25 9.95 10.87 9.86 10.58 8.79 8.82 9.02 1
8 12.63 13.29 10.59 11.08 10.13 9.47 11.05 9.58 10.77 9.12 8.60 8.12
9 16.22 13.69 17.12 14.98 12.55 11.82 18.18 15.07 12.65 10.49 9.89 11.08
10 11.65 11.61 8.48 10.94 10.69 10.29 11.46 10.03 11.73 10.00 9.35 10.19
11 11.21 11.55 11.27 5.08 7.64 9.01 11.29 9.44 12.40 9.66 12.60 11.87
12 12.84 13.29 13.75 9.98 9.61 9.76 10.15 10.01 9.96 8.33 9.02 8.34
13 12.27 12.67 11.97 11.89 12.25 12.18 12.20 11.70 10.37 9.11 8.51 8.90
14 12.85 13.10 11.83 11.41 10.56 10.37 10.03 10.97 11.77 9.18 8.82 8.87
15 14.47 14.97 13.69 11.69 10.23 12.34 13.27 12.51 12.49 10.58 9.57 8.22
16 5.81 5.71 4.62 4.44 4.37 3.16 5.44 5.49 5.59 4.70 4.69 4.00
17 14.15 13.16 11.14 12.16 9.15 9.67 11.10 10.62 11.14 9.66 8.90 10.14
18 12.03 12.41 12.01 11.69 10.02 11.75 11.00 10.41 10.30 8.82 8.75 7.99
19 8.40 8.64 8.98 9.07 7.88 8.17 8.45 8.16 8.33 7.35 6.78 6.40
20 9.54 9.88 9.93 9.52 9.34 10.09 10.73 10.78 8.52 8.09 7.17 7.48
21 10.30 10.55 10.41 10.13 9.62 10.17 10.53 10.08 7.18 9.15 8.11 8.53
22 9.64 9.85 9.78 8.91 9.93 9.20 10.20 10.09 9.81 8.60 7.70 7.92

 

 

171

Table 24. Interest Rates in Korea by Industry

 

 

 

1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982

1 11.34 11.34 8.96 5.92 7.02 8.16 8.80 10.18 10.50 12.90 15.27 13.94
2 10.89 10.89 7.98 10.34 7.01 10.27 8.50 7.46 10.42 14.25 14.89 14.40
3 12.07 12.07 8.53 8.82 10.09 10.68 10.58 11.02 2.79 16.34 18.49 15.58
4 14.95 14.95 8.10 9.30 12.87 14.16 15.99 14.78 11.02 18.65 23.31 18.51
5 12.64 12.64 8.72 7.32 9.14 10.48 12.17 14.14 15.06 16.90 19.22 15.14
6 12.83 12.83 6.68 7.85 12.01 9.47 9.98 9.75 11.08 18.49 19.21 13.99
7 14.70 14.70 11.65 11.96 14.10 15.23 15.99 16.33 16.95 20.05 22.64 21.21
8 13.57 13.57 8.79 9.86 11.37 11.59 9.86 13.26 14.51 18.92 22.13 17.14
9 14.17 14.17 9.91 14.96 14.87 16.40 12.55 18.69 20.94 11.82 18.39 21.28
10 13.49 13.49 10.75 10.75 11.86 24.42 14.06 12.68 14.69 16.88 27.56 15.64
11 12.28 12.28 7.29 7.29 10.25 9.59 10.29 9.17 12.30 13.26 16.35 16.40
12 9.70 9.70 6.88 6.88 6.75 6.48 6.21 5.80 8.99 10.17 15.72 11.70
13 14.84 14.84 11.36 11.36 15.24 12.55 13.27 18.06 15.72 18.98 16.24 16.08
14 13.20 13.20 11.49 11.49 14.39 11.06 9.29 8.91 11.74 20.24 15.53 15.06
15 14.78 14.78 12.33 12.33 10.13 12.56 12.45 13.71 14.36 19.13 20.49 17.07
16 11.44 11.44 8.05 8.05 7.92 9.06 7.81 8.74 11.76 14.22 14.37 12.41
17 13.99 13.99 12.97 12.97 7.82 11.16 12.58 13.04 15.55 25.64 18.34 15.64
18 13.48 13.48 13.78 13.78 14.46 11.28 11.97 11.60 15.35 19.55 16.74 15.84
19 12.27 12.27 9.39 9.39 7.52 13.73 9.69 12.48 14.36 14.45 16.91 14.79
20 15.09 15.09 11.07 11.07 11.69 13.32 13.81 14.68 18.46 21.14 21.06 15.73
21 6.11 6.11 4.74 4.74 5.03 4.83 3.54 3.98 4.46 5.74 7.25 5.19
22 12.39 12.39 8.16 8.16 8.01 12.56 8.52 10.76 17.55 31.17 17.00 12.27

 

 

 

172

‘1?" 4...,

Table 24. (cont’d)

 

1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994

 

 

13.92 13.81 12.85 11.61 12.37 12.69 12.04 10.22 13.07 13.74 10.48 11.13
12.71 13.41 10.84 10.99 11.49 12.64 11.59 9.25 9.71 10.88 10.27 10.20
13.51 14.10 12.87 12.23 12.15 13.20 13.02 12.06 12.97 13.85 11.25 11.36
16.46 16.72 16.88 14.96 15.37 16.00 15.25 13.47 14.98 15.69 13.60 12.72
13.95 13.86 13.73 12.44 12.13 12.10 11.68 11.28 12.41 13.26 11.27 11.72
12.92 12.91 15.83 13.97 10.37 11.36 14.08 15.21 15.23 16.66 12.25 13.07
13.87 12.74 13.56 14.32 12.97 13.42 13.49 14.00 12.66 12.79 10.92 12.63
13.55 14.73 15.15 14.88 14.56 14.70 13.54 11.52 11.22 14.16 11.16 11.95
17.19 40.34 12.72 10.75 9.05 9.12 8.05 7.87 9.95 10.35 8.54 8.05
13.30 14.01 13.31 14.52 13.67 14.41 13.18 13.01 13.75 14.24 11.75 12.85
15.02 13.94 13.18 13.96 13.67 14.36 13.50 11.95 12.04 12.64 11.72 12.02
11.98 12.95 10.17 8.39 9.92 10.27 10.49 10.38 12.81 11.68 9.41 9.26
13. 18 16.15 14.06 16.03 13.99 16.12 17.40 12.95 12.25 15.97 14.24 15.28
14.52 13.17 12.42 13.04 11.78 14.72 14.04 15.14 14.16 14.31 11.59 12.21
13.59 12.25 12.64 11.71 12.43 14.40 14.21 12.99 38.72 14.63 11.84 11.21
10.66 13.29 11.31 11.45 11.57 13.03 13.90 12.33 14.14 15.32 11.22 11.02
12.96 9.80 9.64 10.59 15.86 16.97 16.33 13.84 14.52 17.86 10.63 13.14
11.98 13.55 14.58 12.47 12.51 12.92 14.13 11.57 11.49 13.51 11.74 11.69
14.79 12.36 11.05 11.58 11.79 13.62 12.18 11.33 12.42 12.72 11.82 11.62
15.96 15.94 15.75 15.26 15.55 14.76 15.02 13.74 15.86 14.66 13.87 13.65
5.20 6.18 6.01 6.64 7.20 7.90 7.44 9.52 7.14 7.16 7.04 7.53
13.39 12.03 11.54 15.59 9.42 7.02 15.50 13.15 12.12 9.24 8.91 10.50

 

“\IQMhLANr—n

I—‘h—I
V-‘OO

 

NNNb-‘HI—OD—dI—lb—II—IID—l
IJHOOWQOM-BWN

 

 

 

173

 

 

 

0_N_.0 N000.0 N000.0 0000.0 N000.0 0N0 ~N~ 0000.0 0.0 NN00.0 0N00.0 0N00.0 0000.0 00N0.0 000—
:0N.0 0000.0 00 0.0 0 00.0 0000.0 0. _0 00— 000 .0 0.0 0— 00.0 00N0.0 0000.0 0N00.0 N000.0 000—
000N.0 0N00.0 0000.0 00N0.0 0N—0.0 0.00 00N NON0.0 0.0 0000.0 0000.0 0NN—.0 000N.0 0000 N00_
0000.0 030.0 0000 .0 0000.0 0000.0 0. _0 00N 000N.0 N.0— 0 _ NN.0 _N00.0 0000.0 0000.0 0000.0 00—
0000.0 0000.0 0000.0 300.0 000.0 0N0 NNO 000_.0 0.0— 0—000 0000.0 N000.0 N000.0 0000.0 000~
0N00.0 0000.0 0000.0 0NON.0 0000.0 0.00 0—0— 0000.0 0.N_ 0000.0 200.0 000.0 000.0 0000.0 000—
00000 000 _ .0 N000.0 82.0 000N.0 0.00 000~ 0000.0 0.0— 000N.0 080.0 00N0.0 0N00.0 0—000 000—
00000 000 ~ .0 000N.0 0000.0 000 ~ .0 0N0 0000 0000.0 N.0_ 000 _ .0 0000.0 NON0.0 000 0 .0 800.0 000 _
00000 000N.0 000—0 0000.0 002.0 0.00 00N 00N0.0 0.0— 0000.0 0200.0 0N00.0 0000.0 00N0.0 000_
.0000 ~00N.0 00000 0N00.0 0000.0 0.00 00N 0N—0.0 _.0— 0000.0 0—0—0 0000.0 00N0.0 0000.0 000—
_000.0 N00—.0 00N—.0 0000.0 0000.0 0.00 0: 0000.0 0.0— 0000.0 N000.0 0000.0 0000.0 0000.0 000—
00000 002.0 200.0 :00 0N00.0 0.0 00 0000.0 0.0— 0000.0 0000.0 000.0 0000.0 E000 000_
0000.0 00000 00N0.0 000.0 000.0 N00 00 N000.0 0.0- 000.0 N000.0 000N.0 00000 000.0 N00_
0000.0 0000.0 0000.0 0000.0 0000.0 0.00 00— 0000.0 N.0~ 0000.0 0000.0 0000.0 ~00N.0 0000.0 00—
N000.0 00N0.0 00N0.0 0000.0 0000.0 0.00 00N 0000.0 N.0_ N00_.0 _N00.0 000_.0 ~00N.0 0000.0 000—
0000.0 0000.0 0000.0 N000.0 0000.0 0.N0 00_ 0000.0 0.0_ 020.0 NNN0.0 0000.0 0000.0 000—0 000_
000N.0 0000.0 0000.0 0000.0 0000.0 000 N0_ 000N.0 0.0N 0000.0 N000.0 0N00.0 000N.0 NN00.0 0000
000N.0 0000.0 0000.0 0000.0 0000.0 0N0 00 0000 .0 0.0— 000N.0 0000.0 00000 300.0 0000.0 000_
300.0 0000.0 0000.0 0000.0 0000.0 0.00 00 0000.0 0.0_ ~0N0.0 0000.0 0000.0 050.0 0000.0 000~
N000.0 ~0N0.0 0000.0 ~NON.0 0000.0 0.00 N0 0000.0 0.0- 0000.0 0NN0.0 0_N0.0 0000.0 0000.0 0000
0N00.0 0000.0 N000.0 000N.0 0NN0.0 0N0 00 0000.0 0.0— 00N0.0 N000.0 N0_0.0 0000.0 300.0 000~
0000.0 0000.0 0000.0 00~N.0 0000.0 0.00 0 0N00.0 _.0_ 000N.0 0000.0 0000.0 000N.0 N000.0 000—
0000.0 0000.0 0_ 00.0 0000 .0 00N0.0 0.00 0 0000.0 0.0~ 000N.0 0000.0 0000.N 0000.0 N000.0 N00_
0000.0 0000.0 0000.0 0000.0 0_N0.0 0.00 0_ 500.0 0.3 000N.0 ~00N.0 000.0 N000.0 .00—.0 ~00_
“000.3% m0<>2MOD£ m0<305 m0<325 H0355 200m 300 0055 00800.0. REE 09.0035 09.005 FEES REE g

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.0092 57.8003— 2: E 03.: «:5 .3 930.0

174

 

 

 

0000.0 000.0 0000.0 N0_0.0 00N0.0 0000.0 003.0 003.0 NONN.0 0N00.0 000N.0 0003
0000.0 0200.0 0000.0 0000.0 0000.0 N000.0 003.0 0002.0 000N.0 00N0.0 000N.0 0003
800.0 00N0.0 0000.0 0000.0 0000.0 N000.0 000_.0 002.0 3002.0 030.0 000N.0 N002
0000.0 000N.0 0200.0 080.0 0000.0 0000.0 003 .0 0 200.0 002 _ .0 000N.0 0002 .0 ~00—
_N00 .0 0N0 _ .0 0000.0 0000.0 0000.0 0000.0 0303 .0 0000.0 00— 2 .0 000N.0 0002 .0 000—
00000 000— .0 0N00.0 0000.0 0000.0 0000.0 000 2 .0 003 _ .0 000N.0 0000.0 ~00N.0 000—
_ 000.0 0000.0 2000.0 0000.0 0000.0 0000.0 0003 .0 2000.0 003 .0 00N0.0 000N.0 000—
0N _0.0 0000.0 220.0 0000.0 0000.0 0000.0 000— .0 002 .0 000N.0 0000.0 NO0N.0 000—
02000 000N.0 0000.0 0000.0 0N00.0 0000.0 0002.0 22.0 0030.0 3000.0 N020.0 0003
0000.0 2000.0 002.0 0000.0 0N00.0 N000.0 0003.0 NO0N.0 ~00N.0 0000.0 0030.0 0002
800.0 00N0.0 0NON.0 0000.0 0N00.0 N000.0 0N03.0 0002.0 80N.0 0000.0 020N.0 000—
0000.0 0000.0 0NON.0 0000.0 0200.0 0000.0 02N.0 0N0~.0 0NON.0 0000.0 030.0 000—
00000 0200.0 0NON.0 200.0 N000.0 00N0.0 002 _ .0 0 3 02 .0 000N.0 0N00.0 000N.0 N00—
00N0.0 0N00.0 0003.0 0000.0 0000.0 2030.0 0000.0 003.0 0N0—.0 000N.0 0002.0 200—
N000.0 0000.0 0N00.0 0000.0 0N00.0 0000.0 0000.0 0000.0 0002 .0 2 _ ..0 N03 2.0 0002
_N03.0 0000.0 0200.0 0000.0 0000.0 N000.0 0000.0 0000.0. 008.0. 202.0 0000.0 0002
.N0— .0 _000.0 0000.0 0000.0 0000.0 N000.0 N000.0 00N0.0 0000.0 .NON.0 N02 2.0 000—
_N0_.0 0003 .0 0000.0 0000.0 00N0.0 :000 0000.0 0000.0 N000.0 00_N.0 003 .0 000—
~N02.0 0000.0 0000.0 0000.0 00N0.0 300.0 0000.0 0000.0 N00_.0 NO0N.0 0N0—.0 000—
_N0_.0 N000.0 0000.0 0000.0 00N0.0 200.0 0000.0 0000.0 008.0. ~N00.0 2000.0 000—
0800 0000.0 230.0 0N00.0 0000.0 00N0.0 8N0.0 002.0. 0000.0. 0000.0. 3000.0. 000—
0000.0 0000.0 0020.0 0N00.0 0000.0 00N0.0 800.0 200.0. 2000.0 003 3.0 0000.0 000—
0000.0 0000.0 0030.0 020.0 0200.0 0000.0 200.0 0000.0 00N0.0 0002.0 080.0 N002
0000.0 0000.0 N200 0000.0 0N00.0 0000.0 0000.0 0003.0 0002.0 0NON.0 02N.0 ~00—
0DXmS Emmi Emms 0.5202 Mpg Emvm 00003 000030 00002 DOD—>03 DOOM g

 

 

 

 

 

 

 

 

 

 

 

 

 

8.2.80 2 2.3.

175

LIST OF REFERENCES

REFERENCES

Ahn, Jong S. 1994. The Effects of Tax Policy on Foreign Direct Investment.
Research Report 94-12. Korean Tax Institute.

Agarwal, J. P. 1980. Determinants of Foreign Direct Investment: A Survey.
Weltwirtschaftliches Archiv. Vol. 116., pp. 739-773.

Alworth, Julian S. 1988. The Finance, Investment and Taxation Decisions of
Multinationals, Oxford, Basil Blackwell.

Auerbach, Alan J. 1979. “Wealth Maximization and the Cost of Capital,”
Quarterly Journal of Economics, Vol. 93, pp. 433-446.

Auerbach, Alan J. 1983. “Taxation, Corporate Financial Policy and the Cost of
Capital,” Journal of Economic Literature, Vol. 21, pp. 905-940.

Bajo-Rubio, Oscar and Simon Sosvilla-Rivero. 1994. “An Econometric Analysis of
Foreign Direct Investment in Spain, 1964-1989,” Southern Economic Journal,
Vol. 61, pp. 104-120.

Bank of Korea. Economic Statistic Yearbook, various issues

Bernheim, B. D., and John B. Shoven. 1989. Comparison of the Cost of Capital in the
U.S. and Japan: The Roles of Risk and Taxes. Working paper.

Boadway, Robin, Neil Bruce, and Jack M. Mintz. 1984. “Taxation, Inﬂation, and
the Eﬂ‘ective Marginal Tax Rate on Capital in Canada,” Canadian Journal of

Economics, Vol. 17, pp. 62-79.

176

Boskin, Michael J ., and William G. Gale. 1987. New Results on the Effects of the Tax
Policy on the International Location of Investment. In The Effects of Taxation
on Capital Accumulation. Ed. Martin F eldstein. Chicago: University of Chicago
Press.

Bruce, Neil. 1989. “The Impact of Tax Reform on International Capital Flows and
Investment,” in The Impact of Taxation on Business Activity, eds, J. M. Mintz and
DD. Purvis. Kingston, Ontario: John Deutsch Institute.

Chang, Ching H., and Peter W. Cheng. 1992. Tax Policy and Foreign Direct
Investment in Taiwan. In The Political Economy of Tax Reform, eds., T. Ito and
A. O. Kreuger. Chicago: The University of Chicago Press.

Dunning, J. H. 1973. “The Determinants of International Productions,” Oxford Economic
Papers, Vol. 25, pp. 289-336.

Dunning, J. H. 1981. International Production and the Multinational Enterprise.
London: Allen and Urwin.

EURO Money Ofﬁce. EUROMoney, various issues.

Gerard, Mar. 1993. “Cost of Capital, Investment Location and Marginal
Effective Tax Rate: Methodology and Application,” in Empirical
Approaches to Fiscal Policy Modeling, eds, Alberto Heimler and Danielle
Meulders. London & New York: Chapman & Hall Press.

Goldberg, M. A. 1972. “The Determinants of U.S. Direct Investment in the EEC:

Comment,” American Economic Review, Vol, 62, pp. 692-699.

177

Gravelle, Jane G. 1994. The Economic Effects of Taxing Capital Income.
Cambridge: The MIT Press.

Grubert, Harry and John Mutti. 1991. “Taxes, Tariffs and Transfer Pricing in
Multinational Corporate Decision Making,” The Review of Economics and
Statistics, Vol. 73, pp. 285-294.

Hall, R. E. and D. W. Jorgenson. 1967. “Tax Policy and Investment
Behavior,” American Economic Review, Vol. 57, pp. 391-414.

Hartman, David G. 1984. “Tax Policy and Foreign Direct Investment in the United
states,” National Tax Journal, Vol. 37 (4), pp. 475-488.

Hartman, David G. 1985. “Tax Policy and Foreign Direct Investment,” Journal of Public
Economics, Vol. 26, pp. 107-121.

Hines, James R., Jr. 1997. “Tax Policy and the Activities of Multinational Corporations,”
in Fiscal Policy: Lessons ﬁom Economic Research, ed. Alan J. Auerbach.
Cambridge and London: MIT Press.

Hines, James R., Jr. and Eric M. Rice. 1994. “Fiscal Paradise: Foreign Tax Havens
and American Business,” Quarterly Journal of Economics, Vol. 109, pp. 149-182.

Hines, James R., Jr. 1994. “Credit and Deferral as International Investment Incentives,”
Journal of Public Economics, Vol. 55, pp. 323-347.

Hulten, C. R., and F. C. Wycoﬂ‘. 1981. “The Measurement of Economic Depreciation,”
in Depreciation, Inﬂation, and the Taxation of Income ﬁom Capital,
ed. C. R. Hulten. Washington, D. C .: Urban Institute.

International Labor Ofﬁce. Yearbook of Labor Statistics, various issues.

178

International Monetary Fund. International Financial Statistics, various issues.

Jorgenson, D. W. 1963. “Capital Theory and Investment Behavior,” American
Economic Review, Vol. 53, pp. 247-260.

Jorgenson, D. W., and K. Y. Yun. 1991. Tax Reform and the Cost of Capital.

Oxford: Oxford University Press.

J orgenson, D. W., and Ralph Landau. 1993. Tax Reform and the Cost of Capital.
Washington, DC: The Brookings Institution.

J orgenson, D. W., and Martin A. Sullivan. 1981. “Inﬂation and Corporate Capital
Recovery,” in Depreciation, Inﬂation, and the Taxation of Income from Capital,
ed. C. R. Hulten. Washington, D. C.: Urban Institute.

Jun, J. S. 1995. “The Impact of International Tax Rules on the Cost of Capital,”
in The Eﬂects of Taxation on Multinational Corporations, ed., Martin Feldstein.
Chicago: The University of Chicago Press.

King, M. A. 1977. Public Policy and the Corporation. London: Chapman & Hall

King, M. A. and D. Fullerton. 1984. The Taxation of Income from Capital
Chicago: The University of Chicago Press.

Korea Development Institute. 1996. Facts and Figures on the Public Sector In Korea.
Seoul, Korea: KDI.

Kim, Yun C. 1994. The Effects of Foreign Direct Investment in Korea. Monthly Research
Statistics. Bank of Korea.

Kwack, Tae W. 1985. “Cost of Capital for Korean Corporate Firms”, in Public

Policy, Corporate Finance and Investment, The Experiences of Japan,

Korea and Taiwan, ed., Kwang Choi. Toyko: Institute of Developing Economics

179

Kwack, Tae W. 1985. Depreciation Method and Capital Income taxation
Korea Development Institute Research Paper. 85-05. Seoul, Korea: KDI

Kwack, Tae W. and Kye S. Lee1992. "Tax Reform in Korea,” in The Political Economy
of Tax Reform, eds, Takateshi Ito and Anne 0. Krueger.
Chicago: The University of Chicago Press.

Ministry of Finance and Economy. Fiscal and Financial Statistics, various issues.

Ministry of Finance. Korean Taxation, various issues.

Ministry of Finance and The Korea Development Bank. 1993. Foreign Capital
Inducement in Korea for 30 years.

Mintz, Jack M., and Thomas Tsiopoulos. 1994. “The Effectiveness of Corporate Tax
Incentives for Foreign Investment in the Presence of Tax Deferral”, Journal of
Public Economics, Vol. 55, pp. 233-255.

Mintz, Jack M. 1990. “Corporate Tax Holidays and Investment,” The World
Bank Economic Review, Vol. 4, No. 1, pp. 81-102.

Mintz, Jack M 1993. “On the Taxation of Multinational Corporate Investment When
the Deferral Method is used by the Capital Exporting Country,” Journal of Public
Economics, Vol. 51, pp. 75-96.

Modigliani, Franco, and Merton H. Miller. 1958. “The Cost of Capital, Corporation
Finance, and the Theory of Investment,” American Economic review,

Vol. 48 (3), pp. 261-297.

Modigliani, Franco, and Merton H. Miller. 1963. “Corporate Income Taxes and

the Cost of Capital: A Correction,” American Economic review,

Vol. 53 (3), pp. 433-443.

180

OECD (Organization for Economic Cooperation and Development). 1991.
Taxing Proﬁts in a Global Economy: Dynamic International Issues.
Paris: OECD.

Poterba, James M. 1991. “Comparing the Cost of Capital in the United States and Japan:
A Survey of Methods,” Federal Reserve Bank of New York Quarterly Review,
Winter, pp. 20-32.

Sharpe, William F. 1970. Portfolio Theory and Capital Markets
New York: McGraw Hill Press.

Singh, Harinder and Kwang W. Jun. 1995. Some New Evidence on Determinants of
Foreign Direct Investment in Developing Countries.

Working Paper 1531. International Finance Division, The World Bank.

Slemrod, Joel. 1990. “Tax Effects on Foreign Direct Investment in the United States:
Evidence ﬁ'om a Cross-Country Comparison,” in Taxation in the Global Economy,
eds, Assaf Rasin and Joel Slemrod. Chicago: The University of Chicago Press.

Slemrod, Joel, and Kenneth A Timbers. 1991. “Japanese and U.S. Tax Treatment of Their
Resident Multinationals. Who Has the Comparative Advantage?,” in Tax Reform

in the United States and Japan. eds, David F.Bradford and Kent E. Calder.

Center of International Studies, Princeton University.

Tajika, Eiji, and Yuji Yui. 1988. “Cost of Capital and Eﬂ‘ective Tax Rate:
A Comparison of U.S. and Japanese Manufacturing Industries,”

Hitotsubachi Journal of Economics, Vol. 29, pp. 181-200.

181

Torrisi, C. R 1985. “The Determinants of Direct Foreign Investment in a Small LDC,”
Journal of World Development, Vol., 10, pp. 29-45.

Trela, Irene, and John Whalley. 1992. "The Role of Tax Policy in Korea's
Economic Growth," in The Political Economy of Tax Reform eds, Takateshi Ito
and Anne 0. Krueger. Chicago: University of Chicago Press.

Tsai, Pan L. 1991. “Determinants of Foreign Direct Investment in Taiwan: An Alternative
Approach with Time-Series Data,” World Development, Vol. 19, pp. 275-285.

U. S. Department of the Treasury Internal Revenue Service. Corporation Income Tax
Returns, various issues.

U. S. Department of Commerce, Bureau of Economic Analysis. Survey of Current
Business, various issues.

U. S. Department of Commerce, Bureau of Economic Analysis. U.S. Direct Investment
Abroad: Benchmark Survey, Final Results, various issues.

Wheeler, D ., and A. Mody. 1992. “International Investment Location Decisions,
The Case of U.S. Firms,” Journal of International Economics, Vol. 33, pp. 57-76.

World Bank. World Table, various issues. '

Wu, D. 1989. “An Empirical Analysis of Japanese Direct Investment in

Taiwan: A Neoclassical Approach,” in Advances in Financial Planing:
Taiwan ’s Foreign Investment, Exports, and Financial Analysis, eds,
C. F. Lee and S. C. Hu. pp. 65-89. Greenwich, Conn: JAI Press.
Wu, D., and Ji Chou. 1994. “The Cost of Capital and the Eﬁ‘ective Tax Rate in
Taiwan, 1961-1985,” in Taxation and Economic Development Among
Paciﬁc Asian Countries, eds, R A. Musgrave and C. H. Chang, and John Riew,

pp. 117-139. Westview Press.

182

Young, Kan H. 1988. “The Eﬂ‘ects of Taxes and Rates of Return Foreign Direct
Investment in the United states,” National Tax Journal, Vol. 41 (1), pp. 109-121.
Yun, K. Y. 1992. “Taxation of Income from Foreign Capital in Korea,”
in The Political Economy of Tax Reform, eds, T. Ito and A. O. Kreuger.

Chicago: The University of Chicago Press.

183

"Illllllll’lllll7llf