Title: Some environmental correlates of foreign market service strategies
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00098142/00001
 Material Information
Title: Some environmental correlates of foreign market service strategies
Physical Description: xiv, 221 leaves : diagrs. ; 28cm.
Language: English
Creator: Hilterman, Robert James, 1942-
Publication Date: 1975
Copyright Date: 1975
 Subjects
Subject: Investments, American   ( lcsh )
Commerce -- United States   ( lcsh )
Management thesis Ph. D   ( lcsh )
Dissertations, Academic -- Management -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis--University of Florida.
Bibliography: Bibliography: leaves 215-220.
General Note: Typescript.
General Note: Vita.
Statement of Responsibility: by Robert James Hilterman.
 Record Information
Bibliographic ID: UF00098142
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000163957
oclc - 02770779
notis - AAT0316

Downloads

This item has the following downloads:

PDF ( 8 MBs ) ( PDF )


Full Text














SOME ENVIRONMENTAL CORRELATES OF FOREIGN
MARKET SERVICE STRATEGIES










By

ROBERT JAMES HILIERMAN


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PIIILOSPHY


UNIVERSITY CF FLORIDA
1975
































































Copyright By

ROBERT JAMES HILTERYMAN

1975




































Dedicated to

JOAN and ROBIN












ACKNOWLEDGEMENTS


This dissertation could not have been completed without the

assistance and guidance of many individuals. I wish to express my

deep appreciation to the members of my supervisory committee, Dr.

J. M. Feldman, Chairman, Dr. M. B. Connolly, and Dr. R. E. Klippel,

for their advice and direction in this research. A special debt of

gratitude is owed to Professor J. M. Feldman for his support, counsel,

and encouragement throughout the project.

I also wish to thank Dr. P. Krishnakumar, Dr. G. A. Moore, Jr.,

Dr. E. R. Wittkopf, and Dr. J. S. Fitch for the suggestions and

insight they provided in defining operational variables.

Finally, special thanks go to my wife, Joan Hilterman, and to

Mrs. Alice Cullu for their editorial and typing assistance they so

willingly gave in preparing this manuscript.











TABLE OF CONTENTS


PAGE

ACKNOWLEDGEMENTS . . . . . . . .... . . i iv

LIST OF TABLES . . . . . . . . . . . . viii

LIST OF FIGURES . . . . . . . . . . ... xi

ABSTRACT . . . . . . . . . . .. . . xii

Chapter

I. INTRODUCTION. . . . . . . ... ... .. 1

The Internationalization of Business. . . . .. 1

International Business Theories . . . .... .. 2

Market Service Strategies . . . .... .... 12

Statement of Purpose. . . . . . . ... 15

II. THE DETERMINANTS OF FOREIGN MARKET SERVICE STRATEGIES 16

Litvak and Banting Model of International Business
Arrangements. . . . .. .. . . . . 16

Conceptual Model of Foreign Market Service Strategies
and their Determinants. . . . ... . . 22

The Need for Empirical Research . . . ... 36

Research Hypotheses . . .. . . . . . 38

III. METHODOLOGY . . . . . . . . . . . 40

Research Design . . . . ... . . . 40

Subject Population. . . . ... .. . . . . 45

IV. OPERATIONAL DEFINITION OF VARIABLES, DATA COLLECTION,
AND CONSTRUCT VALIDATION. . . . . . .. . 47







Chapter


Operational Definition of Variables . . . .

Data-Collection. . . . . . . . . .

Construct Validation . . . . . . . .

Additional Methodological Considerations . . .

Criterion and Predictor Variables. . . . . .

Data Base Variables. . . . . . . . .

V. MODEL DEVELOPMENT, METHODOLOGY OF VALIDATION, AND DATA
ANALYSIS . . . . . . . . . . .

First-Order Factor Model . . . . . . .

Second-Order Factor Model . . . . . . .

Environmental Factor Model . . . . . . .

Test for Differences in Predictive Ability for All
Models . . . . . . . . . . .

VI. INTERPRETATION OF RESULTS AND CONCLUSIONS . . .

Environmental Structure and Its Relationship to the
Three Criterion Variables . . . . . .

Test of Lead or Lag Relationships . . . . .

Implications for International Management . . .

Suggestions for Future Research. . . . . .


APPENDIX

A. Definitions of Litvak and Banting Environmental
Factors. . . . . . . . . . . .

B. Countries Included in Subject Population . . . .

C. Variables and Data Sources . . . . . . .

D. Test Results of the First-Order Factor Model .. ...

E. Test Results of the Second-Order Factor Model . .

F. Test Results of the Environmental Factor Model . .

G. A Study of Lead and Lag Associations Between Predictor
and Criterion Variables . . . . . . .


Page

47

50

52

57

61

100


104

104

117

129


136

139


1.40

160

162

164


. .


.


.


. .







Chapter Page

BIBLIOGRAPHY ...... ... ....... ....... . 215

BIOGRAPHICAL SKETCH. ........... ... ...... . . 21


vil











LIST OF TABLES


TABLE PAGE

1. COMPARATIVE GROWTH OF UNTIED STATES FOREIGN DIRECT
INVESTMENTS, EXPORTS, AND GROSS NATIONAL
PRODUCT, 1950 to 1970 . . . . .. . . . 2

2. FACTOR LOADINGS FOR POLITICAL STABILITY FACTORS . . 72

3. FACTOR LOADINGS FOR MARKET OPPORTUNITY FACTORS. ... . 78

4. FACTOR LOADINGS FOR ECONOMIC DEVELOPMENT AND PERFORMANCE
FACTORS . . . . . . . . ... . . .82

5. FACTOR LOADINGS FOR CULTURAL UNITY FACTORS. . . .. 88

6. FACTOR LOADINGS FOR LEGAL BARRIERS FACTORS. . . .. 91

7. FACTOR LOADINGS FOR PHYSIOGRAPHIC BARRIERS FACTORS. . 94

8. FACTOR LOADINGS FOR GEO-CULTURAL DISTANCE FACTORS . .. 97

9. DATA BASE VARIABLES AND VARIABLE IDENTIFICATION ... 101

10. TEST SUMMARY OF THE FIRST-ORDER FACTOR MODEL. CRITERION
VARIABLE=INDEX OF FOREIGN MARKET SERVICE STRATEGIES . 114

11. TEST SUMMARY OF THE FIRST-ORDER FACTOR MODEL. CRITERION
VARIABLE=UNITED STATES EXPORTS 115

12. TEST SUMMARY OF THE FIRST-ORDER FACTOR MODEL. CRITERION
VARIABLE=UNITED STATES FOREIGN DIRECT INVESTMENT. . 116

13. FACTOR LOADINGS FOR SECOND-ORDER FACTORS. . . . ... 119

14. SECOND-ORDER FACTOR NAMES AND IDENTIFICATION. . . .. 121

15. TEST SUMMARY OF THE SECOND-ORDER FACTOR MODEL. CRITERION
VARIABLE=INDEX OF FOREIGN MARKET SERVICE STRATEGIES . 125

16. TEST SUMMARY OF THE SECOND-ORDER FACTOR MODEL. CRITERION
VARIABLE=UNITED STATES EXPORTS. . . . . . . 126

17. TEST SUMMARY OF THE SECOND-ORDER FACTOR MODEL. CRITERION
VARIABLE=UNITED STATES FOREIGN DIRECT INVESTMENT. . 128








TABLE PAGE

18. RELATIONSHIP BETWEEN ENVIRONMENTAL FACTORS AND
FIRST ORDER FACTORS. . . . . . . . . . 132

19. TEST SUMMARY OF THE ENVIRONMENTAL FACTOR MODEL. CRITERION
VARIABLE=INDEX OF FOREIGN MARKET SERVICE STRATEGIES . 133

20. TEST SUMMARY OF THE ENVIRONMENTAL FACTOR MODEL. CRITERION
VARIABLE=UNITED STATES EXPORTS. . . . . . .. 134

21. TEST SUMMARY OF THE ENVIRONMENTAL FACTOR MODEL. CRITERION
VARIABLE=UNITED STATES FOREIGN DIRECT INVESTMENT. ... . 135

22. TEST RESULTS OF THE DIFFERENCES BETWEEN MULTIPLE
CORRELATION COEFFICIENTS. . . . . . . . ... 137

C-1 VARIABLES AND DATA SOURCES. . . . . . . . ... 172

D-1 TEST RESULTS OF THE FIRST-ORDER FACTOR MODEL. CRITERION
VARIABLE=INDEX OF FOREIGN MARKET SERVICE STRATEGIES . 178

D-2 TEST RESULTS OF THE FIRST-ORDER FACTOR MODEL. CRITERION
VARIABLE=U.S. EXPORTS . . . . . . . . .. 180

D-3 TEST RESULTS OF THE FIRST-ORDER FACTOR MODEL. CRITERION
VARIABLE=U.S..FOREIGN DIRECT INVESTMENT . . . .. 183

E-1 TEST RESULTS OF THE SECOND-ORDER FACTOR MODEL. CRITERION
VARIABLE=INDEX OF FOREIGN MARKET SERVICE STRATEGIES . 186

E-2 TEST RESULTS OF THE SECOND-ORDER FACTOR MODEL. CRITERION
VARIABLE=U.S. EXPORTS . . . . . . . . .. 188

E-3 TEST RESULTS OF THE SECOND-ORDER FACTOR MODEL. CRITERION
VARIABLE=U.S. FOREIGN DIRECT INVESTMENT . . . ... 191

F-1 TEST RESULTS OF THE ENVIRONMENTAL FACTOR MODEL. CRITERION
VARIABLE=INDEX OF FOREIGN MARKET SERVICE STRATEGIES . 194

F-2 TEST RESULTS OF THE ENVIRONMENTAL FACTOR MODEL. CRITERION
VARIABLE=U.S. EXPORTS. .... . . . . . . 196

F-3 TEST RESULTS OF THE ENVIRONMENTAL FACTOR MODEL. CRITERION
VARIABLE=FOREIGN DIRECT INVESTMENT. . . . . ... 199

G-1 CORRELATIONS BETWEEN OBSERVATIONS OF PREDICTOR VARIABLES
AND RESIDUAL VALUES FOR THE FIRST-ORDER FACTOR MODEL.
CRITERION VARIABLE=INDEX OF FOREIGN MARKET SERVICE
STRATEGIES. . . . . . . . . ... . . 206








TABLE PAGE

G-2 CORRELATIONS BETWEEN OBSERVATIONS OF PREDICTOR VARIABLES
AND RESIDUAL VALUES FOR THE FIRST-ORDER FACTOR MODEL.
CRITERION VARIABLE=U.S. EXPORTS . . . . . .. 207

G-3 CORRELATIONS BETWEEN OBSERVATIONS OF PREDICTOR VARIABLES
AND RESIDUAL VALUES FOR THE FIRST-ORDER FACTOR MODEL.
CRITERION VARIABLE=U.S. FOREIGN DIRECT INVESTMENT . . 208

G-4 CORRELATIONS BETWEEN OBSERVATIONS OF PREDICTOR VARIABLES
AND RESIDUAL VALUES FOR THE SECOND-ORDER FACTOR MODEL.
CRITERION VARIABLE=INDEX OF FOREIGN MARKET SERVICE
STRATEGIES. . . . . . . . . ... . . 209

G-5 CORRELATIONS BETWEEN OBSERVATIONS OF PREDICTOR VARIABLES
AND RESIDUAL VALUES FOR THE SECOND-ORDER FACTOR MODEL.
CRITERION VARIABLE=U.S. EXPORTS . . . . . ... 210

G-6 CORRELATIONS BETWEEN OBSERVATIONS OF PREDICTOR VARIABLES
AND RESIDUAL VALUES FOR THE SECOND-ORDER FACTOR MODEL.
CRITERION VARIABLE=U.S. FOREIGN DIRECT INVESTMENT . . 211

G-7 CORRELATIONS BETWEEN OBSERVATIONS OF PREDICTOR VARIABLES
AND RESIDUAL VALUES FOR THE ENVIRONMENTAL FACTOR MODEL.
CRITERION VARIABLE=INDEX OF FOREIGN MARKET SERVICE
STRATEGIES. . . . . . . . . ... . . 212

G-8 CORRELATIONS BETWEEN OBSERVATIONS OF PREDICTOR VARIABLES
AND RESIDUAL VALUES FOR THE ENVIRONMENTAL FACTOR MODEL.
CRITERION VARIABLE=U.S. EXPORTS . . . . . ... 213

G-9 CORRELATIONS BETWEEN OBSERVATIONS OF PREDICTOR VARIABLES
AND RESIDUAL VALUES FOR THE ENVIRONMENTAL FACTOR MODEL.
CRITERION VARIABLE=U.S. FOREIGN DIRECT INVESTMENT . . 214













LIST OF FIGURES


Figure Page

1. FIRM USING SIMILAR FOREIGN MARKET
SERVICE STRATEGIES IN DIFFERENT MARKETS. . . . ... 14



2. FIRM USING DIFFERENT FOREIGN MARKET
SERVICE STRATEGIES IN DIFFERENT MARKETS. .... .. 14



3. FIRM USING DIFFERENT FOREIGN MARKET
SERVICE STRATEGIES IN A SINGLE MARKET. . . . ... 14



4. RELATIONSHIP OF FOREIGN MARKET
SERVICE STRATEGIES AND ENVIRON-
MENTAL TEMPERATURE FOR A SINGLE FIRM . . . ... 28



5. RELATIONSHIP OF FOREIGN MARKET
SERVICE STRATEGIES AND ENVIRON-
MENTAL TEMPERATURE FOR MANY FIRMS. . . . . ... 31








Abstract of Dissertation Presented to the Graduate Council of the
University of Florida in Partial Fulfillment of the Requirements cor the
Degree of Doctor of Philosophy


SOME ENVIRONMENTAL CORRELATES OF FOREIGN

MARKET SERVICE STRATEGIES


By

Robert James Hilterman

August, 1975

Chairman: Dr. Jack M. Feldman
Major Department: Management

The purpose of this study was to investigate the relationship of

several environmental characteristics of countries to (1) the marketing

channel roles of domestic firms in foreign countries, (2) the level of

United States exports, and (3) the level of United States foreign

direct investment.

Three sets of constructs were developed as alternative methods of

representing the political, social, cultural, economic,and other

characteristics of a country that may affect international business

operations. A theoretical model (the "environmental temperature model")

of international business arrangements was the source for one set of

constructs. The other two sets were developed inductively and

empirically.

Construct validation was performed on measures of the charac-

teristics postulated by the environmental temperature model using

factor analytic techniques and correlational analysis. Data used in

this study were obtained from a sample of fifty-three countries for

the years 1965 through 1967. The results of the validation indicate

xii








that the several constructs originally proposed are not unitary, as

hypothesized. Instead, they appeared to represent a more complex

environmental structure involving many different elements.

Following construct validation, three theoretical models were

developed (the environmental temperature model and two empirically

derived alternative models) using the three sets of measures as

potential correlates of an index of foreign marketing channel roles,

the level of exports, and the level of investment. The relative

validities of the models were tested by multiple regression analysis

and stepwise multiple regression analysis. The results of this test

do not entirely support the validity of the environmental temperature

model of international business arrangements. While some of the con-

structs were found to be significantly related to the three criterion

variables, others were not. In addition, interaction effects between

constructs as well as nonlinear relationships between the constructs

and criteria were found to be significant, although these relationships

were not predicted by the environmental temperature model.

The results obtained in testing the models indicate that a

country's language and culture are important elements of environmental

structure and are related to a domestic firm's foreign market service

strategies. These environmental elements along with the environmental

elements that define a country's transportation network and its

advertising and promotional media were found to be related to United

States exports. A country's population, and the environmental elements

defining a country's language and culture, its transportation network,

and its advertising and promotional media were found to be related to

United States foreign direct investment.








The continental location of countries was also found to be an

important element of environmental structure. This element both

affected and moderated the associations between other environmental

elements and the criterion variables. Exports to and investment in

countries located on the American continent were especially found to

be different from other countries of the world and more sensitive to

differences in environmental conditions than would otherwise be

expected.

Data were also examined to investigate lead or lag associations

between the constructs and the criterion variables. No such associa-

tions were found. However, there were indications that the values of

the constructs and criterion variables were systematically changing

with time or that the functional relationships between constructs

and criterion variables were changing.














CHAPTER I
INTRODUCTION


The Internationalization of Business

In the past two decades, American business has increasingly looked

abroad for the production and marketing of goods and services. This

"internationalization of business" can be measured in part during this

period by the growth in United States exports and increased foreign di-

rect investments.* As can be seen from Table 1, the book value of for-

eign direct investments has grown rapidly from only 11.8 billion dollars

in 1950 to 78.2 billion dollars in 1970, while exports have increased

from 14.4 billion dollars to 62.9 billion dollars during the same inter-

val. With an average annual growth rate of 9.9 percent for these two

decades, foreign direct investments have outpaced the annual 6.4 percent

growth rate of the United States gross national product. Similarly,

United States exports, with a slightly smaller annual growth rate of 7.6

percent, have also exceeded the annual rate of growth of the gross na-

tional product.

The high levels of United States foreign direct investments and

exports during these past two decades are but two indices of the in-

creasing internationalization of business. The internationalization of

business has, in fact, been accomplished in many different ways. For


*The term "foreign direct investments" should be considered a
stock concept representing domestic ownership of foreign equity. "For-
eign direct investment" is a flow concept representing net capital out-
flows to foreign countries plus reinvested foreign earnings.

1







TABLE 1

COMPARATIVE GROWTH OF UNITED STATES FOREIGN DIRECT INVESTMENTS,
EXPORTS, AND GROSS NATIONAL PRODUCT, 1950 TO 1970
(Money Amounts in Billions of Dollars)


Average annual
1950 1970 growth (%)


U.S. foreign direct
investments 11.8 78.2 9.9

U.S. exports of goods
and services 14.4 62.9 7.6

U.S. gross national
product 284.8 876.4 6.4


SOURCE: Survey of Current Business, United States Department of
Commerce (Various Issues); United States Statistical Abstract, United
States Department of Commerce (Various Issues).


instance, joint ventures, licensing agreements, and branch offices are

some of the strategies that firms have used to expand their current for-

eign activities and penetrate new foreign markets. Other strategies and

combinations of strategies, however, have also been used.

Since the internationalization of business can be accomplished in

many different ways, the question can be raised: what determines the

particular strategy or strategies that domestic firms will use in their

international operations? This question is the central issue of this

study.


International Business Theories

For most of human history, international business has consisted

almost entirely of trade--that is, the importation and exportation of

goods and services between countries. For this part of history,








international trade theory, which will be discussed in a following

section of this chapter, has been useful in explaining almost all

aspects of international business. Today, however, international

business consists not only of trade but also the movement of money,

assets, and personnel through a global network of interrelated organ-

izations. This more complex structure of international business cannot

be explained by international trade theory alone. Unfortunately, no

single general theory has been developed that deals adequately with the

present situation. Robock and Simmonds maintain that:

The development of a theoretical basis, or framework, for ex-
plaining and predicting international business patterns is still
in an early stage. There has been no movement to discard trade
theory. But something more is needed. (1973, p. 16)

Numerous alternative theories have, however, been advanced to ex-

plain the different structures and processes of today's international

business. It is a useful exercise to review briefly some of the theo-

ries of international business as well as the theory of international

trade. Such a review can show the direction of current thought and

illuminate its areas of weakness and strength.

A review of international business theories is useful for another

reason. International business involves interrelated activities in

many fields such as marketing, management, economics, and politics. A

study of international business, therefore, often requires a multidis-

ciplinary approach. Because the reader may not be acquainted with all

of these fields, the following review may serve to illuminate these

fields in the perspective of international business and provide a back-

ground for further discussion.








International Trade Theory

Beginning with Ricardo and Torrens in the early nineteenth cen-

tury and extending into the present, the theory of international trade

has had a long period of development and refinement (Schumpeter, 1954).

This theory attempts to explain why trade occurs between countries,

what products and services are traded, how gains from trade are distrib-

uted among the trading parties, and how international trade patterns

are affected by disturbances. According to the theory, international

trade will occur when two or more regions of the world have comparative

advantages in the production of goods and services. It will be advan-

tageous for each of these regions to specialize in producing the prod-

ucts or services for which it has a relatively greater production effi-

ciency. That is, each region should produce those products and serv-

ices for which its factor costs are comparatively the smallest.

When regions specialize in products or services for which they

have a comparative advantage, then resources are used most efficiently

to produce the greatest output. Trading this output for goods from

other regions results in each region being better off than it would

have been without specialization and trade.

This theory of international trade is rather simple and ideal-

istic, since it is based on somewhat restrictive assumptions. These

assumptions include (1) the absence of transportation costs between

regions, (2) perfect competition between regions, (3) full employment

of productive factors, (4) homogeneity of factor endowments, (5) the

absence of technological innovation, and (6) immobile productive fac-

tors (Root, 1973). Rarely, however, are these assumptions matched by








real world conditions. Thus, the assumptions need to be and can be

modified to make the theory more meaningful and realistic.

Root (1973) suggests that modifications can be included within

the theory of international trade with little difficulty, even though

they may complicate its explanation. He maintains that these modifi-

cations can improve the ability of the theory to predict present trade

patterns.

One modification to the theory of international trade in which

an assumption was relaxed was suggested by Gruber, Mehta, and Vernon

(1967). They hypothesize that technological innovation is an impor-

tant element of the dynamics of international trade and that innova-

tion is not equally shared by all countries of the world. They main-

tain that because of high levels of expenditures on research and de-

velopment, United States firms are able to develop products superior

to those of firms in other countries and to exploit this technological

advantage over foreign competition by exporting. Horst (1972) ob-

tained empirical support for this hypothesis when he found that the

research and development expenditures of United States manufacturing

firms was significantly related to the level of exports to the Cana-

dian market.


General Investment Theory

A general economic theory of international investment suggests

that capital will move from one location to another in response to

differences in marginal productivity. That is to say, capital will

move from areas where it is abundant and earns a low rate of return to

areas where it is scarce and earns a high rate of return. The general








investment theory thus suggests that firms will invest in foreign

rather than domestic ventures when the foreign ventures provide higher

rates of return.

The general investment theory, however, fails to account for

more recent investment activity. Between the United States and Europe,

for instance, investment dollars move simultaneously in both directions.

That is, as domestic firms are investing in European ventures, European

firms are investing in United States ventures (Robock and Simmonds,

1973).

To overcome this inability to account for simultaneous investment,

various modifications to the general theory have been proposed.

Lamfalussy (1961), for instance, has suggested that direct investment

will occur whenever there are large and growing markets--regardless of

their profitability. Scaperlanda and Mauer (1969) found support for

the Lamfalussy version of investment theory in their study of the de-

terminants of United States direct investment in the European Economic

Community. Their research indicated that a country's national product,

in fact, was directly related to the magnitude of United States foreign

direct investment.


Oligopolistic Theory

A recent advance in investment theory by Hymer (1960) suggests

that international investment and also other forms of international

business come about because of the oligopolistic nature of domestic

firms. That is, some United States firms possess a quasi-monopolistic,

or oligopolistic, advantage over foreign firms. A United States firm,

for instance, may have a superior production technology; may produce








highly differentiated products; or may possess managerial and organi-

zational advantages. Consequently, the domestic firm may license its

superior technology, products, or managerial know-how to foreign firms

or it may invest in foreign production and marketing facilities and

compete directly with foreign-owned producers.

A number of empirical studies have provided support for the

hypothesized link between the oligopolistic nature of firms and vari-

ous international business activities (see for example: Brash, 1966;

Dunning, 1958; Horst, 1972). Caves (1971), in an interpretation of a

study by Eastman and Stykolt (1967), also found that in the Canadian

market foreign ownership of production facilities is significantly

related to product differentiation.


Product Life-Cycle Theory

Raymond Vernon (1966) has advanced the theory that international

trade and investment are accounted for by changes in a product's

life-cycle. He applies the term "life-cycle" to the time-span during

which a product will be produced for consumption.

Vernon hypothesized that United States entrepreneurs are likely

to be the first innovators of a high-income or labor-saving product.

In the early stage of a product's life-cycle, it will be produced in

the United States primarily for domestic consumption.

After penetration of the home market, however, the producer

will begin to look eagerly at the untapped potential of foreign

markets and will begin to export the domestically manufactured product.

Exports of the product will then increase as the product becomes more

widely known in the foreign markets.








With time, the nature of both product and producer will change.

Domestic competitors will begin to enter the market with standardized

products and cost-saving production innovations. Foreign entrepreneurs

will also begin to consider the possibility of local production to meet

the expanding demand in their home market. These factors, in turn,

will prompt domestic producers to supply the foreign demand for the

product from within the foreign country. Consequently, investment dol-

lars from the domestic firm will be channeled into the foreign country

for the development of production and marketing facilities. With for-

eign production increasing, domestic exports of the product may level

off and begin to decrease.

In the late stages of a product's life-cycle, labor cost differ-

ences between the host country and the United States may be crucial to

production. The foreign labor costs may be low enough to offset trans-

portation costs to the United States, and the product can then be ex-

ported from foreign countries into this country. More investment dol-

lars would then be needed by domestic firms to expand their foreign

production facilities to supply both foreign and domestic markets.

Thus the product life-cycle model explains how both investment

and exports occur. In the early stages of a product's life-cycle ex-

ports to foreign countries predominate over investment. But as the

product matures, investment replaces exports.

For the durable goods industry, L. T. Wells, Jr. (1969), found

support for the product life-cycle theory. Other studies have also

found it useful in explaining the past history of domestic firms (see

for example: Stobaugh, 1968; Hirsch, 1967).








Recently, however, Vernon himself has concluded:

Though this may be an efficient way to look at enterprises in
the United States economy that are on the threshold of devel-
oping a foreign business, the model is losing some of its rel-
evance for those enterprises that have long since acquired a
global scanning capacity and a global habit of mind. (1971,
p. 107).


Nonrational Investment Theory

The previously cited theories of international business all con-

tain elements of a rational, objective decision process. In these

models a firm's decision to invest in or export to a foreign country

is determined by such objective, quantifiable, and rational criteria

as cost of production, tariff barriers, research and development ex-

penditures, and rate of return on investment.

Aharoni (1966) presents, however, quite a different view of in-

ternational investment. He suggests that before a firm can make a

decision to invest in a foreign country two things will occur. First,

an initiating force will motivate a finr to consider investment in a

foreign country. Secondly, the firm will then analyze the investment

potential of the foreign country.

According to Aharoni, the initiating force may be nonrational,

subjective, and nonquantifiable. It may be a force that arises within

the domestic organization or outside of it. For example, while on va-

cation a corporate officer of a domestic firm may discover a potential

market in a foreign country. His discovery of this potential market

may be the initiating force that prompts him and his firm to consider

the possibilities of foreign investment. The initiating force in this

situation has occurred by "chance," since the consideration of foreign






10

investment might not have been made if the foreign country had not been

included in the corporate officer's vacation itinerary.

Initiating forces can also arise in other ways. For instance,

articles in newspapers, trade journals, and magazines which describe a

foreign country may be read by a corporate officer. The information

supplied by these articles may prompt the officer and his firm to con-

sider the investment potential of the foreign country. The initiating

force in this case was the reading of the articles by the corporate

officer and the information about the foreign country that the articles

contained. This initiating force has again occurred by "chance." Had

the articles not been written or had they not been read by the corpo-

rate officer, the initiating force would not have occurred.

The analysis of potential foreign investment may also be influ-

enced by nonrational and subjective criteria. For instance, a firm

may follow a policy of not investing in South American countries. This

policy may be based purely on the false belief that all South American

countries are highly unstable and subject to internal disorder. Be-

cause the firm has neither rationally nor objectively questioned this

belief, the firm's investment dollars may be directed toward countries

with less desirable investment opportunities.

Although the Aharoni theory of foreign investment has been widely

referenced in international business literature (see for example: Hays,

1971; Robock and Simmonds, 1973; Schreiber, 1970; Vernon, 1972), little

empirical work has yet been done in testing this theory. A single ex-

ception is the study by Miller and Weigel (1972). They found that the

decision by domestic firms to invest in Brazil was effected by initi-

ating forces experienced by the firms. Unfortunately, no attempt was






11

made in their study to differentiate between rational and nonrational

elements that might be included in these initiating forces. The pos-

sibility that nonrational or chance elements may enter into the deci-

sion to invest abroad and also may enter into the analysis of differ-

ent investment opportunities, therefore, awaits empirical verification.

The theories discussed previously in no way exhaust the number

that have been suggested to account for international business. For

international trade alone, a great body of theory exists, bearing on

individual trade problems and situations (Tinbergen, 1954). The same

can be said about studies of international investment. However, the

foregoing theories seem to be representative of current thought. They

are often cited in international business texts (see for example:

Root, 1973; Robock and Simmonds, 1973; Hays, Korth, and Roudiani,

1972), and, as demonstrated by the studies cited previously, they have

been the basis of recent research. While not exhausting the theoreti-

cal approaches to international business, they do contain a common

theme also found in most other theories of international business--they

tend to describe international business by emphasizing exports and for-

eign direct investment.

While foreign direct investment and trade may be important to

the study of capital flows, balance of payments, and similar interna-

tional economic and political matters, it is not the only measure of

the internationalization of business. One could also consider such

strategies as sales agencies, sales branches, joint ventures, and li-

censing agreements as indicators of the internationalization of busi-

ness. Research that considers only foreign direct investment and trade






12

may be ignoring important phenomena that could help describe the oper-

ations of the modern international business organization more

accurately.

Internationalization of business occurs not simply because a

firm is investing or trading abroad; it comes about because a firm de-

cides to provide services or products in one or more foreign markets.

Through some combination of production and marketing strategies, the

firm services the demand of both its domestic markets and markets in

foreign countries.* These strategies, then, may be useful indices for

analysis of the internationalization of business.


Market Service Strategies

The strategies that a firm may use to service a foreign market

are numerous and varied. Some firms, for instance, have all of their

production facilities located domestically and export directly to for-

eign marketing agents. Other firms service foreign markets locally

through parent-owned manufacturing and marketing subsidiaries. Still

other strategies involve different combinations of foreign production

and marketing activities, different levels of parent-ownership and

foreign participation, and different amounts of domestic and foreign

investment capital.


*In many instances in international literature the term "exploit"
is often used to describe the action of a firm in a market. Many
negative connotations are associated with the word exploit, however,
and the assumption may be made that firms are insensitive to customers
and society and have a proclivity to "gouge" a market for all it is
worth. Because this policy is clearly not true of all firms, the term
exploit will not be used to describe market activities. Instead the
word "service" will be used to describe how firms supply
want-satisfying goods and services to a market.







13

The availability of diverse strategies provides business organi-

zations with alternative solutions to servicing foreign markets. For

instance, some firms, as shown in Figure 1, may service foreign markets

with the same strategy. Other firms, as shown in Figure 2, may use

different strategies in different markets. Still other firms, as shown

in Figure 3, may find that multiple strategies may be the most effi-

cient method to service a single foreign market.

Foreign direct investment and exports are important and sometimes

necessary elements in implementing and maintaining some of these market

service strategies. Foreign direct investment and exports, however, do

not completely describe the service strategies used in foreign markets.

Different strategies require different levels of investment, exports,

foreign participation, and foreign production and marketing. Because

companies use different strategies and combinations of strategies to

service foreign markets, studies that have considered only foreign di-

rect investment and exports have only partially explained the servicing

of foreign markets by international organizations.

McDonald and Parker (1962) maintain that exports and a high level

of foreign investment are representative of market service strategies

that lie at opposite ends of an evolutionary process. They suggest

that a firm may become international through the following steps:

1. Export-import activity with little change in management ori-
entation or production lines.
2. Foreign licensing and international transfer of technological
know-how, with little change in management or orientation and
domestic operations.
3. Investments in overseas operations, including assembly plants
and full manufacturing, sometimes through joint ventures but
largely independently. This stage involves substantial in-
vestment of funds and management efforts and leads to the
development of international skills. Still, the domestic op-
erations dominate company policy.









FIGURE 1

FIRM USING SIMILAR FOREIGN MARKET SERVICE
STRATEGIES IN DIFFERENT MARKETS


FIGURE 2

FIRM USING DIFFERENT FOREIGN MARKET SERVICE
STRATEGIES IN DIFFERENT MARKETS


FIGURE 3

FIRM USING DIFFERENT FOREIGN MARKET SERVICE
STRATEGIES IN A SINGLE MARKET


iT EXPORTING lI FOREIGN

DIRECT INVESTMENT IN
4 WHOLLY OWNED SUBSIDIARY MARKET









4. Substantial increase in foreign investment, with foreign
assets becoming an important part of total assets and over-
seas profits contributing a substantial part of overall" com-
pany profits. At this point, the company emerges as a world
enterprise, and there may be an integrated global approach
to production, sales, finance, control, and other matters.
(1962, pp. 17-24)

Dymsza (1972) points out that many other strategies are possible

and that it is not necessary for all enterprises to evolve the same

way. He also suggests that firms may bypass steps, evolve at different

rates, and use more than one market service strategy simultaneously.

Since various strategies can be and are used to service differ-

ent markets, the question can be raised "what determines the strategy

that will be used to service a market?" This question, unfortunately,

cannot be easily answered. International trade theory and investment

theories provide insight only into exports and investment strategies.

International business literature usually refers in general terms to

the importance of firm, industry, or foreign market parameters in

making market service strategy decisions. Very little theoretical or

empirical work has focused on specific determinants of market service

strategies. In order to better understand and explain the interna-

tionalization of business, however, the determinants of market serv-

ice strategies need to be more thoroughly defined.


Statement of Purpose

The purpose of this study is to conduct an exploratory investi-

gation of the determinants of market service strategies. More spe-

cifically, models will be developed relating market service strategies

to potential determinants. The validity of these models will then be

tested by analysis of empirical data.














CHAPTER II
THE DETERMINANTS OF FOREIGN
MARKET SERVICE STRATEGIES


A model that provides a theoretical basis for the investigation

of foreign market service strategies will be discussed in this chap-

ter. This model was developed as a corollary to a model of interna-

tional business arrangements suggested by Litvak and Banting (1968).

Before presenting the model used in this study, first consider the

Litvak and Banting model.


Litvak and Banting Model of International
Business Arrangements

The theoretical model suggested by Litvak and Banting indicates

that certain characteristics or environmental factors of a country de-

termine the institutional structure of foreign marketing channels in

terms of the roles of foreign marketing middlemen. For example, in

some countries environmental factors may cause foreign marketing chan-

nels to take the form of foreign marketing middlemen operating as

agents of domestic firms. In other countries the environmental fac-

tors may cause middlemen to become merchant wholesalers for domestic

firms. In still other countries the environmental factors may cause

middlemen and domestic firms to form joint ventures.

In the Litvak and Banting model the environmental factors are

related to the degree of control that domestic firms exercise over

their foreign marketing middlemen. This control, in turn, forces the

foreign marketing middlemen to evolve and to assume specific

16









institutional roles. Hence, the environmental factors determine the

structure of foreign marketing channels.

Litvak and Banting suggest that the institutional structure of

foreign marketing channels is determined by the following environmen-

tal factors of a country:

1. Political stability

2. Market opportunity

3. Economic development and performance

4. Cultural unity

5. Legal barriers

6. Physiographic barriers

7. Geo-cultural distance

As explained by Litvak and Banting, each of these seven factors can be

considered as a measure of a country's environment. Conceptualizing

a country's environment in terms of a temperature scale, high politi-

cal stability contributes to a "hot" environment. Low political sta-

bility contributes to a "cold" environment. In the same way each of

the other six environmental factors contributes in part to the "envi-

ronmental temperature" of a country.


Cold Environments

According to Litvak and Banting, a country will possess a cold

environment when the environmental factors of political stability,

market opportunity, economic development and performance, and cultural

unity are low and the environmental factors of legal barriers,


*Definitions of environmental factors and their contribution to
environmental temperature are presented in Appendix A.









physiographic barriers, and geo-cultural distance are high. For a

domestic firm, such environmental conditions may be viewed as poten-

tially risky for investment or involvement in foreign marketing and

production activities. Consequently, the firm may be content to use

a foreign marketing agent as its representative in that country's

market and to grant him a major role in the foreign marketing channel.

In cold environments, therefore:

The manufacturer's agent finds few threats to his continuation
as a pure agent. His principal is more inclined to let the
agent handle his accounts as he sees fit, and competitive pres-
sures toward institutional change are less dynamic. (Litvak
and Banting, 1968, p. 461)


Hot Environments

A country possesses a hot environment when the environmental

factors of political stability, market opportunity, economic develop-

ment and performance, and cultural unity tend to be high and the envi-

ronmental factors of legal barriers, physiographic barriers, and

geo-cultural distance tend to be low. A domestic firm could view such

a country as being safe for investment or involvement in foreign mar-

keting and production activities. In addition, the firm may believe

that increasing its control of the foreign marketing channel may pro-

vide greater profits. Therefore, the firm may form licensing agree-

ments, partially owned subsidiaries, or wholly owned manufacturing

subsidiaries in the foreign country.

A foreign marketing middleman, however, could not continue to

exist as an agent of a domestic firm in a country with a hot environ-

ment. With the increased presence, power, and profit expectation of

the domestic firm, the foreign marketing middlemen would have to enter






19

into licensing agreements, participate in joint ventures, or serve in

some smaller channel role than pure agent. In hot environments:

Environmental forces irresistibly push the agent toward new in-
stitutional structures. If he does not adapt, he finds he no
longer has any principals to represent, or competitors have
taken most of his business away. In other words, the environ-
ment is "too hot" for his continued existence as a pure agent.
(Litvak and Banting, 1968, p. 461)


Related Studies

The concepts of environmental factors and environmental temper-

ature were originally defined by Litvak and Banting in their theoreti-

cal model of international business arrangements. These two concepts

have, subsequently, been used in related studies. Goodnow and Hansz,

for instance, hypothesize:

A firm will tend to pursue an entry strategy involving greater
control and greater investment in marketing channel activities
as the country's environment becomes "hotter" in the Litvak and
Banting sense. (1972, p. 33)

In a study of 222 firms, Goodnow and Hansz compared the environ-

mental temperature of countries with the strategies that were used

when the firms initially entered and penetrated the foreign countries

to market their principal products. In their study one hundred coun-

tries were first assigned an environmental temperature of hot, moder-

ate, or cold. The entry strategies of the firms were then compared

with the respective environmental temperatures of the countries in

which the firms were establishing operations. They found that the ex-

tent of investment and the level of control that these firms exercised

over foreign marketing channels were related to environmental temper-

atures of the countries. In countries with hot environments, the

firms tended to invest heavily and use a large number of majority-owned








or jointly owned manufacturing facilities. In countries with cold

environments, exporting through company-owned or agent-representative

facilities predominated over all other entry strategies (Goodnow and

Hansz, 1972).

While the results of the Goodnow and Hansz study tended to sup-

port their hypothesis, there were some important exceptions. A number

of South- and Central-American countries, along with Spain, South

Africa, the Philippines, and India, possessed larger numbers of

majority-owned manufacturing facilities than would seem to be warranted

by the countries' environmental temperatures. Goodnow and Hansz sug-

gest that these discrepancies may occur because firms assign different

levels of importance to the different environmental factors. In their

study the environmental factors were considered to be of equal impor-

tance in determining the environmental temperature of a country. That

is, environmental factors were given equal weight in each factor's

contribution to environmental temperature. Firms, however, may be in-

fluenced unequally by market conditions, and the environmental factors

should be given different weights when the environmental temperature

of a country is determined. An environmental temperature obtained

with the use of unequally weighted environmental factors may be quite

different than if equally weighted factors were used. This difference

may account for the discrepancies noted by Goodnow and Hansz.

Another possible reason for the discrepancies found by Goodnow

and Hansz may be the need to include variables in their study to mod-

erate other predictor variables. These moderator variables, for ex-

ample, may define developed or undeveloped countries or the continental

locations of countries and function in a prediction equation by forming








interactions with predictor variables. These interactions may then

increase or decrease, only for those countries defined by the moder-

ator variables, the weight that is associated with the predictor var-

iables and that equates the predictor variables to the criterion

variables.

As a hypothetical example of moderation effects, assume that

the environmental factor political stability is associated with the

level of United States foreign direct investment for all countries of

the world. Further assume that this association is linear and posi-

tive and that an increase in political stability is associated with

an increase in investment. A variable defining countries located in

Europe will moderate political stability if the interaction between

the variable and political stability is also found to be positively

associated with United States foreign direct investment. This inter-

action indicates that the change in investment associated with a

change in political stability is greater for European countries than

for all other countries. The variable defining European location is,

therefore, said to moderate the association between political stabil-

ity and investment.

In another study, Masson (1973) hypothesized that the Litvak

and Banting environmental factors could be used to determine the level

of different types of United States foreign direct investment. Masson

found that environmental factors were related to foreign direct invest-

ment in the areas of public utilities, manufacturing, petroleum, min-

ing and smelting, and general investments. He also reported that dif-

ferent weights were associated with each of the environmental factors

when they were used to predict the levels of the different types of








direct investment. He suggested that these weights might not be the

same for all countries but may vary for different groups of countries.

A literature review revealed that although extensive reference

to the Litvak and Banting model has been made in some marketing man-

agement texts (see for example: Lazer, 1971; Miracle and Albaum, 1971),

the two studies by Goodnow and Hansz, and by Masson are apparently the

only empirical tests of the model or its concepts. In each of these

two studies the conceptsof environmental factors and temperatures were

used to study specific international business activities. In the

Goodnow and Hansz study, environmental temperature was found to be re-

lated to foreign market entry strategies of firms. In the Masson

study the environmental factors were found to be related to the levels

of different types of foreign direct investment. Besides being useful

in the study of entry strategies and of foreign direct investment,

however, the Litvak and Banting model may also be useful in providing

a framework and theoretical basis for examining and investigating the

determinants of foreign market service strategies.


Conceptual Model of Foreign Market Service
Strategies and their Determinants

I suggest that a conceptual model relating foreign market service

strategies to their determinants can be defined as follows:

The market service strategies used by domestic firms operating

in a foreign market are determined by the Litvak and Banting environ-

mental factors, the market's economic development, and the continental

location of the market.

An explanation of this model can be accomplished in two

parts--first, by considering how environmental factors may determine








foreign market service strategies and, second, how a market's eco-

nomic development and continental location may determine foreign mar-

ket service strategies.


Foreign Market Service Strategies and Environmental Factors

In the Litvak and Banting model each environmental factor pro-

vides a measure of a country's environmental temperature. Environ-

mental temperature is thus an artificial device that combines the meas-

ures of the seven environmental factors into a single parameter. In

the previous explanation of the Litvak and Banting model the use of

environmental temperature provided a simple method for demonstrating

the combined effects of all environmental factors upon the roles of

foreign marketing middlemen. While environmental temperature may be

related to middlemen's roles, the environmental factors determine

environmental temperature; and these factors are thus the basic and

fundamental determinants in the Litvak and Banting model.

In this discussion of the relationship between environmental

factors and foreign market service strategies, the concept of environ-

mental temperature will again be used. As in the Litvak and Banting

model, environmental temperature will be used to combine the measures

of the seven environmental factors and to provide a simple and suc-

cinct method for describing the effects of environmental factors upon

foreign market service strategies. The reader is cautioned to remem-

ber, however, that environmental factors determine environmental tem-

perature and these factors are hypothesized to be the fundamental and

basic determinants of foreign market service strategies.






24

In the Litvak and Banting model, environmental factors determine

the roles of foreign marketing middlemen. If environmental factors

change so that environmental temperature changes, foreign middlemen

may be forced to evolve and assume new institutional roles. But the

level of control, production participation, and marketing participation

of foreign middlemen are only a consequence of the particular foreign

market service strategies that domestic firms elect to use in the for-

eign market. For each strategy that a domestic firm uses, a different

channel role may be required for the foreign middlemen. His role must

thus follow as a consequence of the strategy chosen by the domestic

firm. The environmental factors of a foreign market, therefore, may

be only indirectly related to the roles of foreign middlemen. A di-

rect relationship may occur, however, between the environmental fac-

tors and the adoption of specific foreign market service strategies by

domestic firms.

As an example of the possible relationship between environmental

factors and foreign market service strategies, consider the situation

that occurs when environmental factors produce a hot environmental

temperature. According to the Litvak and Banting model, a hot environ-

mental temperature will be associated with minor channel roles for for-

eign marketing middlemen. Minor channel roles for foreign marketing

middlemen, however, indicate that domestic firms have major roles in

their foreign marketing channels. This conclusion follows since mar-

keting functions that are not included in the roles of foreign mar-

keting middlemen must be accomplished by domestic firms. Therefore, a

hot environmental temperature will also be associated with major chan-

nel roles for domestic firms.






25

Since a hot environmental temperature is associated with major

channel roles for domestic firms and minor channel roles for foreign

marketing middlemen, a hot environmental temperature can be considered

as a determinant of the foreign market service strategies that produce

those roles. In addition, since environmental factors alone determine

a hot environmental temperature, the environmental factors can be con-

sidered as the basic and fundamental determinants of the same foreign

market service strategies.

The relationship between environmental factors and foreign mar-

ket service strategies can also be considered from the point of view

of a cold environmental temperature. If the environmental factors

produce a cold environmental temperature, then this cold environmental

temperature, according to the Litvak and Banting model, will be asso-

ciated with major channel roles for foreign marketing middlemen. Do-

mestic firms, consequently, will have major roles in the foreign mar-

keting channel.

Since a cold environmental temperature is associated with minor

channel roles for domestic firms and major channel roles for foreign

marketing middlemen, a cold environmental temperature can be consid-

ered as a determinant of the foreign market service strategies that

produce those roles. The environmental factors, as determinants of

the environmental temperature, can also be considered as the basic and

fundamental determinants of the foreign market service strategies.


Perception of risks and opportunities

The environmental factors of a foreign market may determine the

foreign market service strategy used by a domestic firm because of






26

the firm's perception of business risks and opportunities in the mar-

ket. A low environmental temperature, from the point of view of the

seven Litvak and Banting environmental factors, indicates a market

with low political stability, market opportunity, economic development

and performance, and cultural unity along with high legal barriers,

physiographic barriers, and geo-cultural distance. These market con-

ditions may be perceived by a domestic firm as detrimental to inter-

national business operations. The firm may consider the market as a

risky business opportunity and may attempt to limit its investment,

control, and foreign production activities in the foreign marketing

channel.

A high environmental temperature, however, indicates a market

with high political stability, market opportunity, economic develop-

ment and performance, and cultural unity, along with low legal bar-

riers, physiographic barriers, and geo-cultural distance. These mar-

ket conditions may be perceived by a domestic firm as being conducive

to international business operations. Because of perceived opportu-

nities in such a market, the firm may attempt to maximize its invest-

ment, control, and foreign production activities in the foreign mar-

keting channel and utilize a service strategy that provides it with a

major role in the channel.


Environmental temperature and the single firm

Figure 4 provides an example of how the environmental factors of

a foreign market may be related to the market service strategy used by

a single domestic firm. If the environmental factors of a foreign mar-

ket combine to form an extremely low environmental temperature, a firm






27

may participate in a very minor way in the foreign marketing channel.

The firm may service the market by exporting, for instance. If the

environmental temperature of the market is slightly higher, however,

the firm may play a more major role in the foreign marketing channel

by using a strategy such as sales agency. Progressively higher envi-

ronmental temperatures may be associated with strategies that provide

such progressively greater foreign channel roles for the domestic

firm as sales branches, licensing agreements, jointly owned subsidi-

aries, and full subsidiary ownership.

Figure 4 may also be used as an example of the possible relation-

ship between the foreign market service strategies used by a single

domestic firm and the environmental temperature of many different for-

eign markets in which the firm is operating. For each market the firm

may use a strategy associated with that market's environmental temper-

ature. For example, in a market in which the environmental factors

combine to form an extremely cold environmental temperature, the firm

may use a foreign market service strategy such as exporting. In a

market with an extremely hot environmental temperature, however, the

firm may use a majority-owned subsidiary to service the market.

Since market conditions may change with time, the environmental

factors may change and, consequently, the environmental temperature

of the market may also change. If the environmental temperature of a

market changes, a domestic firm operating in the foreign market may

alter its foreign market service strategy. For example, again con-

sider Figure 4 and a domestic firm using the strategy of sales agency

in a foreign market with a moderately cold environmental temperature.

If market conditions change and the environmental temperature of the










FIGURE 4

RELATIONSHIP OF FOREIGN MARKET SERVICE STRATEGIES AND
ENVIRONMENTAL TEMPERATURE FOR A SINGLE FIRM


Wholly owned
subsidiary


Jointly owned
subsidiary




Licensing agreement


COLD (Environmental temperature) HOT


market warms slightly, the firm may wish to assume a more major role

in the foreign marketing channel. The firm might adopt a foreign

market service strategy, such as licensing agreement or joint venture.

Larger increases in environmental temperature may result in the firm's

use of a majority-owned subsidiary to service the market. The environ-

mental temperature can, of course, decrease. In that case, the firm

may change its service strategy and assume a more minor role in the

foreign marketing channel. It might, for example, close a sales

branch and open a sales agency or even adopt an exporting strategy


HIGH





I4


0
to




-4A 0
U 4
C 0

0.
0-
o






4j







LOW








if the environmental temperature decreases to an extremely cold

level.

The foreign market service strategies used in Figure 4 are not

the only strategies available to domestic firms. This figure is

presented only as a possible example for a single firm. Firms have

many options available to them by which they can service foreign

markets. Depending on the specific firm and foreign market, there

may be entirely different sets of strategies other than those in

Figure 4. Additionally, the relationship between environmental

temperature and strategies may not be linear. However, for single

firms I suggest that the tendency should be for high environmental

temperatures to be associated with foreign market service strate-

gies that provide firms with major channel roles and low environ-

mental temperatures to be associated with foreign market service

strategies that provide firms with more minor channel roles.


Environmental temperature and many domestic firms

The relationship between environmental temperatures and for-

eign market service strategies may also be considered in terms of

many domestic firms. Figure 5 indicates a possibility for this re-

lationship. This example indicates that (1) many different foreign

market service strategies are being used at any specific environ-

mental temperature, (2) the number of firms using a specific market

service strategy increases with increasing environmental temperature,

and (3) the number of firms using strategies that provide major chan-

nel roles increases at a greater rate with increasing environmental

temperature than does the number of firms using strategies that provide

more minor channel roles.









There are two reasons why multiple foreign market service

strategies may be used simultaneously at any specific environmental

temperature. First, as was suggested in the previous example of

Figure 4, a single domestic firm should tend to assume a foreign mar-

ket service strategy that provides an increasing channel role as the

environmental temperature increases. It would be difficult to imag-

ine, however, that every firm operating in a foreign market would use

the same strategy at the same environmental temperature. Different

firms may, in fact, use the same strategy at different environmental

temperature levels. Each firm may also have an entirely different

set of strategies associated with the environmental temperature

range. Thus, throughout the entire environmental temperature range,

any foreign market service strategy could conceivably be used by one

or more firms.

The second reason that multiple foreign market service strate-

gies have been represented throughout the entire temperature range

was to account for single domestic firms using multiple strategies in

the same foreign market. A firm may service a foreign market, for

example, both by export of domestic output and by local production in

a wholly owned foreign subsidiary. Other combinations of multiple

foreign market service strategies are, of course, possible and rea-

sonable. Thus, many different types of strategies can conceivably be

used simultaneously at any specific environmental temperature.

Figure 5 indicates that the number of firms using each foreign

market service strategy increases with increasing environmental tem-

perature. This increase occurs in part because, as was suggested

before, firms should tend to adopt foreign market service strategies









FIGURE 5

RELATIONSHIP OF FOREIGN MARKET SERVICE STRATEGIES
AND ENVIRONMENTAL TEMPERATURE FOR MANY FIRMS


Wholly owned
subsidiary



Jointly owned
subsidiary


Licensing
agreement



Sales branch



Sales agency


Exporting


(Environmental temperature)


that provide increasing channel roles as environmental temperature

increases. Therefore, the number of firms using the strategies that

provide major channel roles should increase with increasing environ-

mental temperature. If only a constant number of domestic firms op-

erates in a foreign market, it might then be expected that the move-

ment of firms into the strategies that provide major channel roles

would cause a decline in the number of firms using strategies that

provide minor channel roles. However, increasing environmental


MANY












0


-4
0




F





FEW


COLD


-----'
~- ~--









temperature is determined by market conditions associated with less

business risk and with greater market opportunity. New domestic firms

may be motivated by these market conditions to penetrate and service

the markets. The market service strategies of these new firms may be

of any type, including the strategies associated with minor channel

roles. The penetration of the market by new domestic firms at an in-

creasing environmental temperature may then produce a net increase in

the number of firms using each foreign market service strategy.

Figure 5 has been drawn to indicate that the number of domestic

firms using strategies that provide major channel roles is increasing

at a faster rate with increasing environmental temperature than is

the number of firms using strategies that provide minor channel roles.

The penetration of the foreign market by new firms may help provide an

increase in the use of all market service strategies. But new firms

can be expected to follow the suggested relationship of tending toward

the use of strategies that provide major channel roles as the environ-

mental temperature increases. In addition, firms previously operating

in the market should tend to switch from minor to major channel role

strategies with increasing temperature. The effect of these actions

by both new and old firms may be not only a net increase in the use of

all strategies but also a higher rate of increase in the use of the

strategies providing major channel roles as the environmental temper-

ature increases.

Figure 5 shows a linear relationship between environmental tem-

perature and each of the foreign market service strategies. These

relationships were used to simplify the discussion of the figure.

Conceivably, a nonlinear relationship may exist between environmental






33

temperature and foreign market service strategies. Regardless of the

linearity of the relationships, however, I suggest that the tendency

should be for: (1) high environmental temperatures to be associated

with foreign market service strategies that provide firms with major

channel roles and (2) low environmental temperatures to be associated

with foreign market service strategies that provide firms with more

minor roles.


The effect of foreign market service strategies on environmental
factors and temperature

In the examples of the single and multiple firms discussed previ-

ously, I have suggested how environmental temperature may affect and

determine the marketing strategies used to service markets. Possibly,

however, the use of foreign market service strategies by domestic

firms can also affect environmental factors and the environmental tem-

perature of foreign markets.

Investment in foreign subsidiaries and the production and mar-

keting of new products, for instance, may create new consumer demand,

increase purchasing power, stimulate the development of transportation

and communication networks, and provide component parts and materials

for other local industries. The licensing of new products and proc-

esses may increase the productivity of labor and provide an impetus to

local firms to expand their research and development. Other strategies

may also affect similar or additional elements of a market. Since

environmental factors reflect different aspects of a market, activities

of international firms that alter the market may also alter the envi-

ronmental factors and temperature.






34

Foreign Market Service Strategies and a Market's Economic Development
and Continental Location

The model of foreign market service strategies defined previous-

ly included the provision that a market's economic development and

continental location are determinants of foreign market service strat-

egies. Both economic development and continental location were in-

cluded in this model to account for possible moderation effects. Both

economic development and continental location may provide moderation

effects because of firms' perceptions of different markets. If a mar-

ket is economically developed, for instance, a domestic firm may have

a generally favorable attitude toward that market because of its past

performance, its future potential, and because it is economically sim-

ilar to the United States (the home country of the firm and the market

with which it is probably most familiar). Firms may have an unfavor-

able attitude toward undeveloped markets because the markets are eco-

nomically dissimilar to the United States or because the firms may be-

lieve that the markets provide little present or future potential.

Possibly, though, firms may believe that just the opposite conditions

prevail for the two groups of markets. Economically developed markets,

for example, may be perceived by firms as having reached the peak of

their economic growth or of containing too many present or potential

competitors. Undeveloped markets may be considered by firms as having

the potential for future growth.

While no a priori justification exists for believing that firms

perceive developed and undeveloped markets either one way or another,

the distinction is reasonable. Economic and international business

literature has tended to dichotomize countries into those that are






35

developed or undeveloped (see for example: Kuznets, 1972; Fink, 1972;

Moyer, 1968). Firms may make the same dichotomy with markets and may

form separate and different attitudes about the two groups.

Firms may also perceive markets differently, depending on each

market's continental location. Some firms, for instance, may consider

South, Central, and North American countries as economic satellites of

the United States. These firms may believe that they can exert a

strong influence on United States government policy toward these areas

to benefit their own foreign operations (Bennettand Green, 1972). Con-

sequently, these firms may be more favorably disposed toward business

activities in the South- Central- and North-American markets than if

United States government influence were not present.

European markets may also be perceived differently than are other

markets. European countries share a common heritage with the United

States; they have traditionally been trading and business partners with

the United States and European countries have many formal economic,

political, and military agreements. The attitudes and beliefs that

firms have about past relationships with European countries may influ-

ence the firms' present and future activities in European markets.

If firms do perceive markets differently in terms of economic

development or of continental location, then these perceptions may af-

fect the manner in which other market conditions are perceived. The

differential viewing of market conditions for separate groups of mar-

kets may have an effect upon the use of foreign market service strate-

gies in these markets. Consequently, the ability to predict foreign

market service strategies by the use of environmental factors may be

enhanced by the use of economic development and continental location to

classify markets.









The Need for Empirical Research

In previous sections of this chapter a conceptual model relating

foreign market service strategies to possible determinants of these

strategies was defined. A discussion then followed in which an expla-

nation and interpretation of the model was presented, along with exam-

ples showing the possible relationship between foreign market service

strategies and their determinants; Three important questions are sug-

gested by this discussion: First, are the environmental factors mean-

ingful conceptualizations of environmental structure? Second, can the

model of foreign market service strategies be interpreted in quantita-

tive terms so that foreign market service strategies and their hypo-

thetical determinants can be systematically related? Third, does the

model have any validity?

The question of whether the environmental factors are meaningful

conceptualizations of environmental structure has been posed because

these constructs were apparently developed by Litvak and Banting on a

purely a priori basis. They provided no supporting evidence that

would indicate the factors originated from empirical findings. Nor did

they indicate that the environmental factors are suggested by or relat-

ed to other meaningful and empirically verified concepts. Conceivably

there are other constructs representing environmental structure that,

compared to the environmental factors, provide equal or superior pre-

dictive ability and that are also justified on an empirical or theore-

tical basis.

The question concerning the quantitative interpretation of the

model of foreign market service strategy was raised because the model

has so far been interpreted only in qualitative terms. In previous






37

discussion it has been hypothesized that foreign market service strat-

egies are determined by a market's environmental factors, economi- de-

velopment, and continental location. However, no suggestions have

been made pertaining to the quantitative structure of the model--such

as the weight associated with each determinant, the interrelationship

between determinants, and the manner of combining and relating the de-

terminants to foreign market service strategies. Litvak and Banting

imply that the environmental factors are each monotonically related to

the roles of foreign marketing middlemen. Because the model of for-

eign market service strategies is a direct corollary of the Litvak and

Banting model, the environmental factors also may each be monotonically

related to foreign market service strategy. However, the exact rela-

tionship between the hypothetical determinants and foreign market serv-

ice strategies can only be determined from empirical evidence.

It might be suggested that the relationship between the determi-

nants and environmental temperature should also be quantified. This

relationship need not be quantified, however. In the discussion in

previous sections of this chapter, environmental temperature was used

in examples demonstrating the possible combined effects of the environ-

mental factors on foreign market service strategies. If foreign market

service strategies can be shown to be determined by some such combina-

tion of variables as the environmental factors, then environmental tem-

perature can be arbitrarily defined as any monotonically increasing

function of the same combination of variables.

The need then is not to quantify the relationship between the

suggested determinants, environmental temperature, and foreign market

service strategies, but to quantify the model in terms of the determi-

nants and foreign market service strategies.









Empirical research can assist in providing an answer to the va-

lidity of the environmental factors, the quantification of the me el,

and the validity of the model. Therefore, to investigate the concep-

tual model of foreign market service strategies, empirical data will

be examined and analyzed in this study. In addition, the ability of

a market's environmental factors, economic development and continen-

tal location to predict United States exports and foreign direct in-

vestment will also be considered in the empirical research.

United States exports are included in this study, since they

are easily observable and are a major element of international busi-

ness. But more importantly, exports are the result of the market serv-

ice strategy of exporting. Since a market's environmental factors,

economic development, and continental location may determine marketing

service strategies, in general, they may also determine such a single

marketing service strategy as exporting.

United States direct investment is included in this study because

of the suggestion by Masson that it is determined by environmental fac-

tors and because, like exports, it is easily observable and is an im-

portant and major element of international business.


Research Hypotheses

Three hypotheses will be treated in this investigation. These

hypotheses are stated as follows:


Hypothesis I

Foreign market service strategies of domestic firms are related

to a market's environmental factors, economic development, and conti-

nental location.






39

Hypothesis II

The level of United States exports is related to a market's en-

vironmental factors, economic development, and continental location.


Hypothesis III

The level of United States direct investment is related to a

market's environmental factors, economic development, and continental

location.














CHAPTER III
METHODOLOGY


In Chapter II three hypotheses were developed from a conceptual

model of foreign market service strategies. An exploratory investiga-

tion of this model and a test of these hypotheses was accomplished

through an examination of empirical data. This empirical study was

carried out in two phases. Each of these phases requires extensive

discussion to describe the research methodology involved. Therefore,

each phase will be considered separately, in detail, in following

chapters. However to provide continuity for the following chapters, a

general outline of the methodology used in the research design and the

selection of the subject population will be briefly described here.


Research Design

The first phase of the research was designed to accomplish three

objectives: (1) the operational definition of variables, (2) the col-

lection of data, and (3) the validation of constructs. In the first

step of this phase of this study, the three hypotheses developed in

Chapter II were examined to determine independent (predictor) and de-

pendent (criterion) variables. The predictor variables determined by

this examination consisted of the seven environmental factors, a mar-

ket's economic development, and a market's continental location. The

criterion variables consisted of foreign market service strategies,

United States exports, and United States foreign direct investment.

For each of these predictor and criterion variables, explicit






41

definitions were then developed. Finally, from these definitions, op-

erational variables, which could be used as surrogates or proxies for

the predictor and criterion variables, were defined.

Following the definition of operational variables, data were col-

lected from secondary archival sources. These sources were used to ob-

serve the operational variables for a subject population for each of

the years 1965, 1966, and 1967. These observations were designated for

future references as the "observed data values." For each operational

variable and subject, the mean of the observed data values for the

three years was then calculated. These means were designated for fu-

ture reference as the "mean data values."

The validation of constructs was the final step of the first

phase of this study. This validation of constructs was necessary be-

cause concepts, such as political stability, are conceptualizations

that are inherently unobservable. However, these concepts are theoreti-

cally reflected in the operational variables developed for each concept.

If the operational variables representing a specific concept tend to

covary--that is, agree with each other--and if these variables are re-

lated to other variables in theoretically meaningful ways, then we have

evidence that the operational variables are useful measures of the con-

cept and that the concept is a useful conceptualization of phenomena.

The Litvak and Banting environmental factors and a country's eco-

nomic development are constructs and theoretical conceptualizations of

environmental structure. Two methods were used in validating these con-

structs. The first method was used on all of the environmental factors

and consisted of the factor analysis of data for the operational vari-

ables representing each environmental factor. From this analysis the






42

underlying concepts, which these operational variables measure, were

determined. Variables based on these underlying concepts were ti'en

defined, and values were calculated for these variables for the sub-

ject population.

The underlying dimensions defined by factor analysis represent

groupings of those operational variables that covary. If these dimen-

sions are similar to the environmental factors, then the dimensions are

valid measures of the environmental factors and the environmental fac-

tors are meaningful conceptualizations of environmental structure. If

the dimensions are not similar to the environmental factors, the fac-

tors are either not valid measures of environmental factors or the en-

vironmental factors do not represent meaningful conceptualizations of

environmental structure. Because multiple operational variables were

defined for each environmental factor, it is unlikely that among all

the multiple variables, none provides a valid measure for the environ-

mental factor. Therefore, if the underlying dimensions are dissimilar

to the environmental factors, then quite possibly the environmental

factors do not meaningfully represent environmental structure.

A second method to test the validity of constructs was used on

the environmental factor political stability and on the environmental

factor economic development and performance. The method was also used

to test the construct representing a country's economic development.

This test of construct validity consisted of the use of correlation

analysis to determine how the constructs of this study were related to

other independently derived constructs. Those constructs that are re-

lated in a meaningful way to other theoretical constructs represent

elements of a much larger body of knowledge--a body of knowledge which






43

has independent validity. Consequently, support is also provided for

the validities of this study's constructs (Heeler and Ray, 1972;

Campbell and Fiske, 1959).

Following the validation of constructs, variables developed from

the factor analysis, along with the operational variables for a mar-

ket's economic development and continental location, were designated

as the operational variables to be used in the second phase of the

study.

The second phase of this study consisted of the development and

test of analytical models that relate environmental structure to for-

eign market service strategies, United States exports, and United

States foreign direct investment. Because a literature review had

revealed that there were no existing alternatives to the model of

foreign market service strategies developed in Chapter II, two empir-

ically derived models were developed, tested, and compared against the

model of foreign market service strategies. This comparison of models

provided the opportunity to assess the ability of the environmental

factors to predict international business activities and to determine

whether they are superior to other alternative constructs. Testing

the validities of the three models also provides evidence indicating

the quantitative nature of the relationship between environmental

structure and the three criterion variables. This evidence could be

used to (1) check the Litvak and Banting assumptions that the environ-

mental factors are all individually and monotonically related to the

roles of foreign marketing middlemen and (2) check the presence of

moderation effects.






44

Each of the three analytical models tested in the second phase

of the study contained potential predictor variables composed of vari-

ous combinations of main-effect, cross-product, and polynomial terms.

The same combination of terms in each model was tested as potential

predictors of the three criterion variables. Thus, from each model

three equations were developed: one to predict foreign market service

strategies, a second to predict United States exports, and a third to

predict United States foreign direct investment.

In the first analytical model, the operational variables that

were designated at the end of the first phase of the study were used

directly as various main-effect, cross-product, and polynomial terms.

For the second model, the set of values for the variables that

were developed from the factor analysis in the first phase of the

study were themselves factor analyzed. This second factor analysis

provided the means to determine and define underlying concepts, which

may be more fundamental than those concepts developed in the first

factor analysis. For each of those concepts obtained in the second

factor analysis, variables were defined and values for those variables

were calculated for the subject population. These variables, along

with the operational variables for a market's economic development and

continental location, were then used in various main effect,

cross-product, and polynomial terms in the second model.

For the third model, the variables that were developed from the

factor analysis in the first phase of the study were combined to obtain

single variables representing each of the Litvak and Banting environ-

mental factors. These variables, along with the operational varibles

for a market's economic development and continental location, were then






45

used to construct various independent main-effect, cross-product, and

polynomial terms in the third model.

When the validity of the three models was tested, both multiple

regression and stepwise multiple regression were used. The data base

for these tests consisted of mean data values obtained in the first

phase of the study and data values calculated for the variables devel-

oped by factor analysis.


Subject Population

The conceptual model and the three hypotheses developed in Chap-

ter II relate foreign market service strategies, United States exports,

and United States foreign direct investment to their determinants

within the concept of a market. The use of the term "market" to des-

ignate an area of foreign business activity may provide greater valid-

ity for the model and hypotheses than if the term "country" were used.

While a market and a country can be considered as synonymous terms,

firms can conceivably perceive markets as existing outside the limita-

tions created by the arbitrary and artificial boundaries of nations.

For example, a market may be considered to exist across national

boundaries. In addition, multiple markets may also be perceived to

exist within a single country. Thus, the use of market, and not coun-

try, to specify an area of foreign business activity increases the po-

tential of the model and hypotheses to describe valid relationships.

In this study, however, observations of operational variables

could only be obtained for areas consisting of individual countries.

While the model and hypotheses provide the capability for investigating

a wider scope of international business activities, this study was






46

constrained to the investigation of those situations where a country

and a market were considered as identical.

Observations of each of the operational variables were obtained

for a total of fifty-three countries throughout the world. These

countries are listed in Appendix B. The methodology used in selecting

these countries will be discussed in detail in the following chapter.














CHAPTER IV
OPERATIONAL DEFINITION OF VARIABLES, DATA
COLLECTION, AND CONSTRUCT VALIDATION


In this chapter, a detailed review of the methodology involved

in three operations of this study will be presented. These operations

are the operational definition of variables, the collection of data,

and the validation of constructs. Each of these operations will be

discussed separately. In addition, the specific results of these

operations performed on predictor and criterion variables also will be

presented.


Operational Definition of Variables

The operational definition of variables involves the specifica-

tion of observables that can be used as underlying theoretical con-

structs involved in a hypothesis. This operationalization process is

important because it allows inherently unobservable phenomena (such

as "political stability") to be represented quantitatively instead of

verbally. With quantitative observations of variables, the validity

of a hypothesis can then be statistically tested.

The operationalization process is additionally important because

in specifying how phenomena can be measured, the researcher is forced

to consider the exact meaning of the phenomena. Dunnette suggests:

The simple act of specifying measurement operations forces us to
define explicitly what we wish to talk about, thereby ruling out
ambiguities of speculation, loose definition, and distortions of
personal perception that contribute so greatly to imprecision in
everyday discourse. (1966, p. 14)

47







48

The operational definition of variables for this study was accom-

plished in two steps. First, those variables that permit an empirical

test of the theory of foreign market service strategies were determined.

These variables were designated as predictor and criterion variables.

The predictor variables consisted of the seven environmental factors, a

market's economic development, and a market's continental location.

The criterion variables consisted of foreign market service strategies,

United States exports, and United States foreign direct investment.

For each of these predictor and criterion variables, explicit

conceptual definitions were developed. For the seven environmental

factors, the definitions suggested by Litvak and Banting were generally

used. However, in a few cases, it was necessary to make some changes

to their definitions both to conform to current knowledge and to make

the environmental factors separate and mutually exclusive of each

other. The remaining predictor and criterion variables were defined

by reference to current literature.

In the second step of the operational definition of variables,

operational (observable) variables that could serve as proxies for

the conceptual predictor and criterion variables were selected. The

operational variables that were selected fulfilled two requirements.

First, the operational variables were consistent with the definitions

developed previously for the conceptual predictor and criterion vari-

ables; and, second, the operational variables were observable in many

different countries.

Multiple operational variables were selected for each of the

seven environmental factors and for a market's continental location.

For the environmental factors, multiple operational variables were








selected for three reasons. First, no single operational variable

was deemed sufficient to measure adequately all elements of each

environmental factor. For instance, the factor geo-cultural distance

contains both geographic and cultural elements. The second reason

for selecting multiple operational variables was to increase the

likelihood that at least one operational variable provided an adequate

measure of a specific element of an environmental factor. Finally,

multiple operational variables were selected because any single meas-

ure is subject to both systematic and random error. Adding over many

indicants of the underlying construct tends to cancel out these sources

of error, especially random error.

Multiple operational variables were selected for a market's

continental location to define separate continents of the world. These

operational variables could then be tested separately to determine

moderation and interaction effects that might be associated with the

specific continental location of a market.

A single operational variable was selected as a proxy for a mar-

ket's economic development, and single operational variables also were

selected as proxies for each of the criterion variables that consisted

of foreign market service strategies, United States exports, and United

States foreign direct investment. Single operational variables were

deemed adequate for the criterion variables because there existed

single observables that were essentially synonymous with each of the

criterion variables. A single operational variable was chosen for the

economic development of a market because a single variable allowed the

dichotomization of markets into those that were economically developed

and those that were economically undeveloped.







Data-Collection

Initially an attempt was made to obtain measurements of all the

operational variables for each of the years 1965, 1966, and 1967 for

a subject population of 113 countries. Countries that were members

of the United Nations or countries that were included in recent inter-

national business literature were chosen for this sample. The years

1965, 1966, and 1967 were chosen since the most data were available

for this period of time. For the years prior to 1965, United States

foreign direct investment data were not available for a large sample of

countries. For the years following 1965, the available data sources

were inadequate to provide more current values of the operational

variables.

Data values for the operational variables were obtained from

secondary-archival sources and from the United States Department of

Commerce. For future reference these data values will be called ob-

served data values. The sources of these observed data values for all

the operational variables are listed in Appendix C. In many cases ob-

served data values could not be obtained for a specific country in the

subject population. The factor analysis that would be used in later

stages of this study, however, does not allow for missing data. There-

fore, those countries for which complete data could not be obtained

were dropped from the subject population and were excluded from further

analysis.

The final sample for which all operational variables could be

measured consisted of fifty-three countries.* These countries are

listed in Appendix B.


*Belgium and Luxembourg have highly interrelated, if not common,
economic and political systems. These two countries were therefore






51

The arithmetic mean of the observed data values for the years

1965, 1966, and 1967 was calculated for each operational variable for

each country in the subject population. These mean observed data

values were calculated because extreme fluctuations were noted in the

values of the operational variables from year to year. These fluctua-

tions may be due to transient or random events--events that may not be

associated with any of the parameters in the model of foreign market

service strategies. Averaging the observed data values over a three-

year period reduces these fluctuations and results in values that may

be more representative of long-term conditions.

After computing the means of the observed data values, the distri-

bution of these values for each operational variable was standardized.

These standard distributions were obtained by transforming each mean

observed data value into a standard score with the use of the following

equation:


S= Xij Xi
SDi


Whereas Zij is a standard score for operational variable i and country

j, Xij is the mean observed data value for operational variable i and

country j, Xi is the mean of the mean observed data values for opera-

ticnal variable i, and SDi is the standard deviation of the mean ob-

served data values for operational variable i.

The standard score data associated with the operational variables

for the three criterion variables--foreign market service strategies,

domestic exports, and foreign direct investment--were included in the


counted in the subject population as one country and all operational
variables were also measured as if they were one country.






52
data base that was subsequently used to test the model of foreign mar-

ket service strategies. Similarly, the standard score data for conti-

nental location and economic development were also included in this

data base. However, the standard score data associated with the seven

environmental factors were used to determine the environmental structure

of a country. The method by which the environmental structure of a

country was determined will be discussed in the following section of

this chapter.


Construct Validation

The validation of constructs as discussed in Chapter III involves

the determination of whether theoretical concepts are validly repre-

sented by operational measures and whether the theoretical concepts

are meaningful and useful conceptualizations of phenomena. Two methods,

which will be discussed in following sections, were used in this study

to validate constructs. The first method was used on all the environ-

mental factors and consisted of the factor analysis of data for the

operational variables representing each environmental factor. The

second method was used on only the environmental factor political sta-

bility, the environmental factor economic development and performance,

and the variable economic development. This second method consisted

of the use of correlation analysis to determine how constructs were

related to other independently derived constructs.


Construct Validation by Factor Analysis

Litvak and Banting hypothesized that the environmental factors

are unitary concepts and that each factor represents a single element

of environmental structure. To determine the validity of these con-

structs, the standard scores associated with each environmental factor






53

were factor analyzed. This process of factor analysis, which will be

discussed in the following section of this chapter, produces factors

that define separate dimensions of a country's environmental structure.

These dimensions can be compared to the definitions of the environ-

mental factors to determine whether the environmental factors are

valid conceptualizations of the environmental structure that is inherent

in the operational measures.

Measures for each of the factors obtained by factor analysis

were also calculated for each of the countries in the subject popula-

tion. These measures, representing observations of the environmental

dimensions, were then included in the data base that was subsequently

analyzed to test the relationship between environmental structure and

the three criterion variables.


Factor analysis

In factor analysis, a correlation matrix is analyzed to determine

which variables cluster together to provide information on the same

concept. This clustering of variables is called a factor. Each factor

can be considered as the central concept or underlying dimension that

the clustered variables measure.

The amount of information that clustered variables measure is

determined in factor analysis by the factor loadings of each variable.

Those variables with high factor loadings for a specific factor provide

more information about an underlying dimension than do variables with

smaller factor loadings.

In factor analysis, absolute values for factor loadings from .40

to .50 are considered significant and absolute values equal to or greater

than .50 are considered as quite significant (Frank, Kuehn, and Massey,






54

1962). Variables with significant or quite significant factor loadings

for a single factor can be considered to measure individually the same

underlying dimension. By examining the variables that load highly on

a factor, the researcher can often determine the underlying dimension

that the variables measure. The factor can then be named for this

dimension.

The high loading of more than one variable on a factor occurs

because variables provide separate measures of the same factor. The

separate measures of the factor can be combined to form a single

measure of the factor as follows: first, the algebraic signs of the

factor loadings for the variables that load highly on a factor are

determined. These signs are interpreted as unit vectors. That is,

a plus sign is interpreted as a +1 and a minus sign is interpreted as

a -1. The standardized scores for the variables for a subject of the

subject population are then multiplied by each variable's unit vector

to form vector scores. Finally, the vector scores are summed to pro-

duce a single measure of the factor for the subject. This measure is

called a factor score.

Like variables, factors also can covary. In factor analysis

the relationship between the factors can be changed by "rotation."

This rotation is accomplished by changing the initial factor structure

to obtain a second set of factors which are defined by new clusters of

variables. The second set of factors then represents new underlying

dimensions.

Many different techniques are available for factor rotation. The

specific technique that is used is determined by the interpretability

of the rotated factors--that is, how well the underlying dimensions







55

can be defined in meaningful terms. The most desirable rotation tech-

nique is also the one producing factors that have been used and

validated in other research. Rummel (1963) suggests that if underlying

structure is completely unknown,then orthogonal rotation should first

be used. Each of the factors produced by this rotation contains vari-

ables that are maximally related to each other, and the total factor

structure is most parsimonious mathematically (Guertin and Bailey,

1970).


Factor analysis of environmental factor data

Recall that for each environmental factor, operational variables

were selected and measured for each country of the subject population

and the distributions of these measurements were then standardized to

produce standardized scores. These standardized scores associated with

each environmental factor and each subject country were factor analyzed.

Principal axis factor analysis with varimax orthogonal rotation was used

to extract factors for each environmental factor.

Factor loadings with an absolute value equal to or greater than

.50 were used to select operational variables that measured the under-

lying dimension of a rotated factor. The .50 value was used because it

minimized the selection of the same variable for more than one factor.

It was felt that this conservative approach to developing factors would

allow a country's environment to be defined by the simplest and most

basic factor structure.

For some of the factors that were extracted by factor analysis,

more than one operational variable was found to load significantly on

a single factor. Single measures of factors were obtained by calculating

factor scores for each factor for each of the subject countries. The









factor scores for all factors for all subjects were then included in

the data base that was subsequently analyzed to test the model of for-

eign market service strategies.


Construct Validation by Correlational Analysis

Construct validation was accomplished in this study not only by

factor analytic technique but also through correlational methods.

Correlational methods were used because the validity of constructs

is supported if the constructs are related to other concepts and also

operate in meaningful and expected ways (Heeler and Ray, 1972; Runkel

and McGrath, 1972; Cronbach et al, 1963).

Three constructs were included in this test of construct valida-

tion. These constructs were the environmental factor political stabil-

ity, the environmental factor economic development and performance,

and the variable economic development. No independent measures that

might be meaningfully related to the other constructs of this study

could be defined. Consequently, these other constructs were excluded

from this correlational method of construct validation.

As a first step in validation of each construct, an independent

measure of a concept that should be meaningfully related to the con-

struct was defined and measured. Next, the correlations between the

observations of the independent measures and the observations of the

construct's operational variables were calculated. Finally, the

resultant correlation coefficients were examined to determine whether

the independent measure was related to the construct as expected.









Additional Methodological Considerations


Random Error

Errors can be introduced into research data and they may seri-

ously affect the results and conclusions of a study. These errors

may be either systematic or random. An error occurs when data have

been either overstated or understated. This error will be a random

error when all elements of data have an equally probable chance of

being overstated or understated and the direction and magnitude of

the error is uncorrelated with any other data. Random errors may

result for such reasons as human mistakes, mechanical malfunctions, or

electronic failures; and random errors may occur during the collection

of data or any of the subsequent steps involved in analyzing the data.

The use of secondary-archival data in a study also increases the possi-

bility of random errors because of the additional chance of typographi-

cal and printing mistakes.

A number of positive actions were taken in this study to reduce

the occurrence of random errors. Whenever possible, data obtained

from one secondary source were cross-checked with data from one or

more alternative sources. Data inconsistencies between sources were

then examined and any differences were resolved before the data were

incorporated into the study. In this way, typographical errors,

printing errors, or other errors originating in a single secondary data

source could be detected and the propagation of these errors into this

study's data base could be prevented.

The activity of handling and processing data by physical, mechani-

cal, and electronic methods can also result in the generation of random







errors. "Dummy data" were included with the research data in many

steps of the analysis to guard against these errors. These dummy data

were selected to provide specific predetermined results from the anal-

ysis. Deviations from the expected results were then used to indicate

that data errors were being generated.

In addition to dummy data, checksums and other parity types of

information were included with the research data in some steps of the

analysis to detect erroneous changes in data values. In other steps

of the study, the accuracy of data that had been transcribed or re-

corded from other sources were verified by physically cross-checking

the data with their original source.

Although a great deal of effort was expended in reducing the

generation of random errors, the probability of their occurrence in

this study must be presumed to be greater than zero. Given that ran-

dom errors have been introduced into the data base, the question can

be posed: what effect will such error have on the results of this

research?

To answer this question, one must consider the effects of random

data errors on both correlational analysis and factor analysis, the

two primary analysis tools used in this study. In correlational anal-

ysis, data containing random errors will reduce the value of correla-

tions between measures from the value that would be obtainable with

data that are free of random errors. Measures that would be signifi-

cantly correlated when data do not contain random errors may then be

insignificantly correlated when data contain random errors. Conse-

quently, random errors may have the effect of reducing the number of

correlations that are found to be significant.






59

The presence of random data errors, however, does not damage any

significant results that are obtained by correlational analysis. On

the contrary, the existence of random errors serves to strengthen the

confidence that may be placed in significant correlations, since the

level of significance would be higher if the data were free of random

errors.

The effects of random error on factor analysis have been analyzed

by Cattell (1952) and Rummel (1963) and experimentally investigated by

Mosier (1939). These investigators indicate that an orthogonal rota-

tion with data contaminated by random errors will produce factor

loadings somewhat lower than would be produced with uncontaminated data.

But the factor structure as a whole will not be distorted. In factor

analysis, then, the same factors will be extracted with or without the

presence of random errors.


Systematic Error

Systematic error occurs when a measure is consistently understated or

overstated so that the induced error of the measure is correlated with

the same measure or with another measure. Some of the same causes of

random errors may also cause systematic errors. For example, human

mistakes, mechanical malfunctions, or electronic failures that may occur

during the collection and subsequent analysis of data may result in the

generation of systematic errors.

The same safeguards and actions for detecting and preventing

random errors were applicable and were used to detect and prevent

systematic errors. For instance, dummy data, checksums, and parity

information were used in the processing and handling of research data





60
to detect the occurrence of systematic error. The cross-checking of

data between alternative secondary sources also provided a means for

detecting the systematic overstatement of data by one source.

While every effort was made to prevent and detect systematic

errors, it would be a mistake to assume that no systematic error was

inherent in the political, economic, social, and cultural data used in

this study. For example, some of the observations of these parameters

originated with sources located within developing countries. These

are countries that may have a need to present to other countries of

the world the appearance of a politically stable and economically sound

nation. Consequently, the data sources may have been influenced to

understate politically destabilizing events and overstate economic

measures.

Systematic errors may be multiplicatively related to a measure

and cause a systematic understatement or overstatement of data. The

systematic understatement or overstatement of data may result in corre-

lations between measures that are overly low or overly high, respec-

tively. If systematic errors in data cause overly low correlations,

then the effects on the results of this study will be the same as in

the case of data with random errors. That is, a smaller number of

significant correlations may be obtained with the contaminated data

than would have been obtained with data that were free of systematic

errors. However, the lowering of the correlational levels by system-

atic errors, as with random errors, will strengthen the confidence that

may be placed in the significant correlations that are found.

If systematic data errors create overly high correlations between

measures, then some correlations that would not be significant with data







61

that do not contain systematic errors may be accepted as significant.

The systematic overstatement of data, therefore, requires that less

confidence be placed in the results of correlational analysis, since

significant correlations may have been erroneously produced.

Systematic data errors can also affect the results of factor

analysis. Rummel (1963) maintains that the factor structure obtained

with data containing systematic errors may be seriously distorted.

This distortion may allow incorrect factors to be extracted from the

data.

Because of the damaging effects that systematic errors can have

on the results of a study, an attempt was made to measure the system-

atic error that might be inherent in some of the data. Even though

very little can be done to remove the effects of systematic error, the

knowledge of its presence or absence can provide a measure of confidence

in the results and conclusions of the study. The techniques used and

theresults obtained in measuring systematic error will be considered

in other sections of this chapter.


Criterion and Predictor Variables

In the remaining sections of this chapter, the specific results

that were obtained from the definition of operational variables and

collection of data will be presented. The results of these operations

will be presented for each of the criterion and predictor variables

used in this study. In addition, development of the underlying dimen-

sions for the environmental factors will also be presented in each

discussion of an environmental factor.








Foreign Market Service Strategies

The criterion variable foreign market service strategies was

defined in this study as:

The different methods that domestic firms use to produce and
market goods and services in foreign markets. These methods
require different foreign roles for the domestic firm in terms
of foreign channel control, foreign production, and foreign
marketing.

It might be expected that an operational variable could be devel-

oped and used to enumerate the foreign market service strategies of

domestic firms. Enumeration of foreign market service strategies,

however, is a difficult task. Except in a few isolated instances,

firms have not specified in published reports or annual statements the

market service strategies that they have used in foreign countries.

Foreign countries, likewise, have not published or made available infor-

mation concerning the foreign market service strategies that domestic

firms have used within their borders.

Instead of enumerating the different types of foreign market

service strategies, I suggest that the levels of United States exports

and United States foreign direct investment may provide a means of

measuring the use of different foreign market service strategies. To

understand how these two parameters may be used, first consider the

definition cited above for the criterion variable foreign market serv-

ice strategies. According to this definition, the service strategies

that a domestic firm uses in a foreign market require different foreign

roles for domestic firms. Of all the strategies available to a firm,

the strategy of exporting results in the smallest foreign role because

the tasks of foreign channel control, foreign production, and foreign

marketing for the firm are either minor or nonexistent. And, in





63
exporting, little or no foreign direct investment is required for the

firm.

Strategies other than exporting bring about a greater allocation

of the channel tasks to the domestic firm. These strategies may reduce

the amount exported by the firm, but they will, undoubtedly, require

foreign direct investment. In fact, strategies that result in large

foreign roles with high levels of foreign channel control, foreign pro-

duction, and foreign marketing for the domestic firm may be associated

with high levels of foreign direct investment.

As an example of the different foreign market service strategies

available to a firm, consider the following:

Strategy 1. Wholly owned subsidiary

Strategy 2. Joint venture

Strategy 3. Sales branch

Strategy 4. Sales office

Strategy 5. Exporting

These strategies are rank-ordered according to the foreign role

of the domestic firm. Strategy 1 requires the greatest foreign role,

in terms of channel control, foreign production, and foreign marketing;

and Strategy 5,the smallest foreign role. Besides being ordered

according to foreign role, however, the strategies are rank-ordered

according to the amount of foreign direct investment required to imple-

ment that strategy. Strategy 1 requires the greatest foreign direct

investment and Strategy 5 the smallest.

The parallel ranking of strategies by the level of foreign direct

investment and of a domestic firm's foreign role occurs because a firm

usually acquires an increased foreign role--increased channel control,






64

foreign production, and foreign marketing--by foreign direct invest-

ment. Strategies other than the preceding five strategies could have

been used and ranked according to the level of foreign direct invest-

ment required to implement them. Similar results, however, would have

been obtained: strategies ordered by foreign direct investment would

also be ranked by the level of a domestic firm's foreign role.

If a market is serviced with strategies, such as Strategy 1 or 2,

that produce large foreign roles for domestic firms, then the level of

foreign direct investment to the market, compared to the level of

exports to the market, should be high. If another market is serviced

with strategies, such as Strategies 4 or 5, that produce smaller for-

eign roles for the domestic firm, then the level of foreign direct

investment to the market, compared to the level of exports to the

market, should be low.

Comparing the ratio of foreign direct investment to exports for

two markets may yield the following inequality.


[Foreign direct investment > Foreign direct investment
S Exports J a Exports J


here: a = a market where Strategy 1 or 2 is predominantly used.

8 = a market where Strategy 4 or 5 is predominantly used.

In fact, this inequality should hold for any two markets in which the a

market is being serviced by strategies that tend to provide larger for-

eign roles ftr the domestic firm than do those strategies in the B

market. Therefore, the ratio of the value of foreign direct investment

to the value of exports apparently provides an indication of the use of

different foreign market service strategies in different markets. This






65

indication of different foreign market service strategies is not pro-

duced directly in terms of specific strategies but indirectly in terms

of the different levels of roles that domestic firms tend to use in

different markets.

The ratio of the value of foreign direct investment to the value

of exports may also be useful for studying the changing use of strate-

gies in a single market. For instance, if the ratio is observed to

increase from year to year for a single market, then this increase may

indicate that firms are adopting strategies that provide them with

greater foreign roles. Conversely, a declining ratio may indicate that

firms are adopting strategies with smaller foreign roles.

The values of foreign direct investment to a market and exports

to a market are the aggregate values resulting from many domestic firms

using many different strategies. Using aggregate values to obtain a

measure of the use of different foreign market service strategies may

be superior to enumerating the specific strategies that a sample of

firms use. This superiority results because aggregate values allow

for a study of the average behavior of many firms instead of specific

behavior of a few firms. Grunfeld and Griliches (1960) point out that

the use of aggregated economic variables may produce a net gain in

information by reducing the specification error inherent in microeco-

nomic relationships. In situations where the micro behavior cannot be

specified perfectly in terms of micro equations, the use of aggregated

values of a variable, while producing an aggregation error, often pro-

duces a greater aggregation gain.

R. Eisner (1967) also suggests that aggregation of investment

data is sometimes useful because specific investment decisions of firms








are often the results of both permanent and transitory determinants.

To the extent that strategies require foreign direct investment, then

these strategies may also be the result of permanent and transitory

determinants. Aggregate values tend to eliminate transitory phenomena

in much the same way that multiple measures of a construct tend to

eliminate random error. The use of aggregate values then, allows one

to focus on the permanent or average elements of a decision process.

The operational variable for foreign market service strategies

was defined in this study as the ratio of United States foreign direct

investment to United States exports.* For future reference this

operational variable will be called the index of foreign market service

strategies. Measurements of this operational variable were obtained

for each subject country for each of the years 1965, 1966, and 1967.

The arithmetic mean of the observed data values for the three years was

calculated for each subject country, the distribution of the resultant

mean observed data values was standardized, and the standard scores

were then included in the data base.


United States Exports

The criterion variable United States exports was defined as

the value of domestically produced merchandise and of foreign produced

merchandise that is shipped from domestic sources to foreign locations

(U.S. Department of Commerce, Business Statistics, 1967).

United States exports was considered as an operational variable,

and observations for this variable were obtained for each of the sub-

ject countries for each of the years 1965, 1966, and 1967. The


*Unless otherwise indicated, all operational variables used in
this study were quantitative variables.








arithmetic mean of the observed data values for the three years was

calculated for each subject country, the distribution of the result-

ant mean observed data values was standardized, and the standard scores

were then included in the data base.


United States Foreign Direct Investment

The criterion variable United States foreign direct investment was

defined as the net capital outflows to foreign countries plus reinvested

foreign earnings that are used by United States persons, organizations,

or affiliated groups to acquire equity interest in foreign located

business organizations (U.S. Department of Commerce, U.S. Direct

Investments Abroad: 1966, 1970).

United States foreign direct investment was considered as an

operational variable, and observations for this variable were obtained

for each of the subject countries for each of the years 1965, 1966, and

1967. The arithmetic mean of the observed data values for the three

years was calculated for each subject country, the distribution of the

resultant mean observed data values was standardized, and the standard

scores were then included in the data base.


Political Stability

Litvak and Banting defined political stability as:

A system of government which permits representative of the major
segments of its society, enjoys the confidence of its people,
generates conditions for continuity of business operations, and
is sympathetic to private enterprise. (1968, p. 461)

This definition would seem to be more descriptive of a country

with representative democratic government and a laissez-faire economy

than a definition of political stability. Because Litvak and Banting's

definition did not appear to define political stability adequately,






68

a second definition was developed and used in this study. This second

definition defines political stability as a system of government that

maintains continuity and consistency within laws, policies, and legal

enforcement aad provides enduring and steady character and purpose.

Operational variables were not selected to measure political

stability directly. Instead, political instability was defined as

opposite to political stability and opposite in all respects to the

second definition of political stability given. Operational variables

that measure political instability--that is, reverse-scored measures

of political stability--were selected and were then used throughout the

analysis of data. Results from the analysis that contained these oper-

ational variables, however, were translated into terms of political

stability.

Two different methods were used to select operational variables

that measure political instability. The first method was suggested by

Feierabend and Feierabend (1966). They maintained that political

instability is indicated by the amount of aggressive conflict that

occurs within a country. According to them, political instability is

The amount of aggression directed by individuals or groups
within the political system against other groups or against
the complex of office holders and individuals and groups
associated with them. Or, conversely, it is the amount of
aggression directed by these office holders against other
individuals, groups, or office holders within the polity.
(p. 250)

Five operational variables were selected to measure some of the

events that Feierabend and Feierabend suggested represent conflict

behavior and political destabilization. These five variables were

1. Number of protest demonstrations during previous eight years

2. Number of riots during previous eight years








3. Number of armed attacks during previous eight years

4. Number of deaths from domestic violence during previous

eight years

5. Number of government sanctions during previous eight years

Each of these five variables was considered as a separate measure of

political instability. Observations of these variables were obtained

for each of the subject countries for each of the years 1965, 1966,

and 1967. The arithmetic mean of the observations for the three years

was calculated for each variable for each subject country, and the dis-

tribution of the resultant mean observed data values for each variable

was then standardized.

Another method used to measure political instability was suggested

by Taylor (1969). He maintained that political instability may be meas-

ured by considering the number of illegitimate changes in the central

government to the total number of illegitimate and legitimate govern-

mental changes. Four variables were selected that could be measured

and could be used to represent Taylor's concept of governmental change.

These four variables were

1. Number of regular executive transfers during previous eight

years

2. Number of renewals of executive tenure during previous eight

years

3. Number of executive adjustments during previous eight years

4. Number of irregular executive transfers during previous eight

years.

Observations for these four variables were obtained for each subject

country for each of the years 1965, 1966, and 1967. An approximation








of Taylor's measure of political instability was then computed for

each subject country and year of the sample population by dividing the

sum of the values of the last two variables by the sum of the values of

all four variables. For further reference this measure will be called

the index of political instability.

The arithmetic mean of the computed values for the index of polit-

ical instability for the years 1965, 1966, and 1967 was calculated for

each subject country,and the distribution of the resultant mean observed

data values was standardized. In addition, mean observed data values

were also calculated for each of the five variables that were used to

compute the index of political instability for each of the subject

countries, and the distribution of these mean observed data values for

each variable was then standardized.

The standard score data for the five operational variables of

conflict behavior, the four operational variables used to compute the

index of political instability, and the operational variable index of

political instability were analyzed by principal axis factor analysis

followed by varimax orthogonal rotation. Data for the five variables

of conflict behavior and the index of political instability were in-

cluded in this factor analysis since these variables and the index

each measure political instability. Data for the five variables that

were used to compute the index of political instability were also in-

cluded in the analysis since it is possible that these variables indi-

vidually measure political instability.

Four factors were extracted from the data by factor analysis.

The factor loadings for each variable on each factor are shown in






71

Table 2. In addition, those factor loadings that are quite significant

(an absolute value > .50) have been indicated in this table by an

asterisk.

The first factor contained four variables with quite significant

factor loadings. The common dimension measured by these variables was

defined as "executive resistance and change due to internal turmoil."

The first factor was, therefore, named after this dimension.

The common dimensions of the variables that loaded quite signifi-

cantly on each of the other three factors were also defined. The fac-

tors were named after these dimensions as follows: Factor 2 was named

"executive change by coups and by constitutional means," Factor 3 was

named "political instability," and Factor 4 was named "violent

repression."

Factor scores were calculated for each of the four factors for

each subject country, and the resultant scores were then included in

the data base.

For each of the four factors, a single score was calculated for

each country of the subject population. As an example of these


*The operational variables shown in Table 2 are reverse- scored
measures of political stability--that is, they measure political insta-
bility. The positive factor loadings of these variables, therefore,
indicate factors of political instability. However, to maintain con-
sistency with the names of the Litvak and Banting environmental factors,
the factors indicated in Table 2 will be referred to collectively as
political stability factors. To prevent confusion about the direction
of measurement, future references to political stability factors will
include the note "reverse-scored" to indicate that the factors represent
political instability.
The specific names developed for each factor in Table 2 will repre-
sent the true direction of measurement for the factor. Consequently,
future references to specific factors will not require an additional
note indicating their direction of measurement.







TABLE 2

FACTOR LOADINGS FOR POLITICAL STABILITY FACTORS


Factor loading



Variable Factor 1 Factor 2 Factor 3 Factor 4


Number of protest demonstra-
tions during previous
eight years

Number of riots during pre-
vious eight years

Number of armed attacks
during previous eight
years

Number of deaths from
domestic violence during
previous eight years

Number of governmental sanc-
tions during previous eight
years

Number of regular executive
transfers during previous
years

Number of renewals of exec-
utive tenure during pre-
vious eight years

Number of executive adjust-
ments during previous eight
years

Number of irregular executive
transfers during previous
eight years

Index of political insta-
bility


.29 -.20


.61* -.01


.02 -.19


.78* -.08


.44 -.16


.61*


.41 -.21


.78*



.64*



-.17


-.79* -.07



.20 -.03



.45 -.01


.81*


*Factor loading > an absolute value of .50.






73

calculations, measures of Factor 3, the political instability factor,

were determined with the use of the following equation:


F3,i = (-l)Zet,i + (+l)Zpi,i


Where F3,i = single measure of Factor 3 for country i.

Zet,i = standard score of the operational variable, number of

renewals of executive tenure during previous eight years,

for country i.

and Zpi,i = standard score of the operational variable, index of

political instability, for country i.

The term Zet,i is multiplied by a minus one, since the operational vari-

able number of renewals of executive tenure during previous eight years

has a negative and quite significant factor loading on Factor 3. The

term Zpi is multiplied by a positive one, since the operational variable

index of political instability had a positive and quite significant

factor loading on Factor 3. Single measures of the factors for all

the environmental factors for all subject countries were calculated in

a similar fashion.


Test for systematic error and construct validation

Observations for the five variables of governmental change that

were used to compute the index of political instability and for the

four variables for politically destabilizing conflict behavior may be

systematically understated. This possibility exists because censor-

ship of news-reporting organizations may prevent news of all conflict

behavior and governmental change from being disseminated (Rummel, 1963).

The data used in this study may have been obtained from these organi-

zations and, consequently, may be understated because of this censorhip.






74

However, the possibility that the data are overstated because of censor-

ship is highly unlikely, since it would be difficult to believe that

censorship would result in the reporting of more conflict behavior and

governmental change than actually occurred.

To test the proposition that the measures for the ten variables

of conflict behavior and governmental change were systematically under-

stated, product moment correlations were obtained between measures for

each of the ten variables and measures for freedom of the press. For

forty-seven countries of this study's subject population, these meas-

ures for freedom of the press were obtained from Taylor and Hudson

(1972) and were interpreted as measures of the absence of censorship.

All correlations were found to be negative and significant (P < .05).

These negative correlations indicate that there is less freedom of the

press--that is, more censorship of the press--as countries exhibit

greater amounts of conflict behavior and governmental change. If

censorship results in the suppression of reporting of conflict behavior

and governmental change, then the negative correlations indicate that

higher levels of conflict behavior and governmental change are more

highly understated. It must be concluded then that the measures of

conflict behavior and governmental change used in this study are quite

probably understated systematically.

As indicated previously, data containing systematic error can

distort the factor structure obtained with factor analysis. Since the

measures of conflict behavior and governmental change are probably under-

stated systematically, the underlying dimensions of political stability

that were obtained by the factor analysis of these measures may be








distorted. Any results obtained in this study that include the

political stability factors, therefore, must be interpreted with

caution.

The test for systematic error in the measures of political sta-

bility also may be used for a test of the construct validity of polit-

ical stability. Freedom of the press is a concept that can be expected

to be related to political stability. That is, it seems reasonable to

presume that less freedom of the press is associated with less political

stability. The significantly negative correlations that were obtained

between measures for freedom of the press and measures for conflict

behavior and governmental change support this presumption. If measures

of political stability are related to measures of another concept in a

meaningful and expected way,then the concept of political stability may

be part of a network of related concepts. Consequently, this evidence

supports the construct validity of political stability (Runkel and

McGrath, 1972).


Market Opportunity

The environmental factor market opportunity was defined by Litvak

and Banting as:

A sufficient number of customers with incompletely satisfied
needs and the necessary resources with which to satisfy those
needs for the product or service in question. A "hot" factor
when opportunity is high. (1968, p. 461)

Both marketing concepts and economic concepts are contained in

this definition of market opportunity. For instance, the phrase

"satisfying customer needs" is traditionally associated with marketing

while the term "resources" is usually interpreted as an economic concept.

Although no problem is presented by the marketing terminology, the








economic elements of the definition are closely aligned and overlap

with the Litvak and Banting definition for the environmental factor

economic development and performance. As will be discussed in the

following section, the economic environment of a country will be defined

and described totally by the environmental factor economic development

and performance. Therefore, the economic related concepts in the

Litvak and Banting definition of market opportunity were deleted.

In addition to removing the economic concepts from the definition

of market opportunity, the definition was also changed to indicate that

market opportunity occurs when potential customers can be informed of

the availability of goods and services. This change was made because

it was felt that market opportunity cannot exist in a country unless

there are means available for informing potential customers of the

existence of need-satisfying goods and services.

The net effect of all changes to the definition resulted in mar-

ket opportunity being defined as a sufficient number of potential cus-

tomers with incompletely satisfied needs who can be informed of the

availability of goods and services to satisfy those needs.

Using this definition for market opportunity, the following six

operational variables were selected for market opportunity:

1. Population

2. Population density

3. Number of newspapers per capital

4. Number of telephones per capital

5. Number of radios per capital

6. Number of television receivers per capital









These six quantitative variables measure market opportunity in two

ways. The first two variables, population and population density (in

number of people per square mile provide measures of the gross number

of potential customers and the concentration of potential customers in

a market. The last four variables provide measures of some of the pro-

motional media that firms use to disseminate information and to inform

potential customers of need-satisfying goods and services.

The six operational variables for market opportunity were measured

for each subject country for each of the years 1965, 1966, and 1967.

The arithmetic mean of the observed data values for the three years was

calculated for each variable for each subject country and the distri-

bution of the resultant mean observed data values for each variable was

then standardized.

The standard scores for the six operational variables that meas-

ure market opportunity were factor analyzed. Three orthogonal factors

were extracted from these data. The factor loadings for the six variables

on each factor are shown in Table 3.

The three factors obtained by factor analysis were named as

follows: Factor 1 was named "marketing media," Factor 2 was named

"population density," and Factor 3 was named "population." Factor

scores were calculated for each of these three factors for each subject

country, and these scores were then included in the data base.


Economic Development and Performance

Litvak and Banting defined the environmental factor economic

development and performance as:








TABLE 3

FACTOR LOADINGS FOR MARKET OPPORTUNITY FACTORS


Factor loading


Variable Factor 1 Factor 2 Factor 3



Population -.07 -.01 1.00*

Population density .03 1.00* -.00

Number of newspapers
per capital .91* .25 -.03

Number of telephones
per capital .94* .00 -.04

Number of radios per
capital .89* -.04 -.14

Number of televisions
per capital .97* -.06 .00



*Factor loading > an absolute value of .50.


The level of a country's economic growth, efficiency, equity, and
stability, which shape the environment for private enterprise.
Applying Rostow's classification within the context of this defi-
nition, levels would be grouped as follows:

Low. The traditional society, the preconditions for takeoff. A
"cold" factor.

Medium. The drive to maturity. A "moderate" factor.

High. The age of high mass-consumption. A "hot" factor. (1968,
p. 461).

This definition of economic development and performance is ex-

pressed totally in economic terms. As was mentioned in the previous

section, Litvak and Banting also defined market opportunity, in part,

with economic terms. In addition, several other environmental factors






79

can be operationalized with variables that measure a country's economic

condition. Unfortunately, defining or operationalizing different envi-

ronmental factors by the same economic terms or variables reduces the

ability to differentiate between those factors. For this reason, the

environmental factor of economic development and performance was con-

sidered as the only factor that describes the economic condition of a

country.

Since the definition of economic development and performance

suggested by Litvak and Banting contains only economic terms, it was

retained in its original form. All other environmental factors, however,

were redefined or operationalized, when necessary, in noneconomic terms.

It was felt that the economic development and performance of a

country is indicated by its productive output, its use of resources and

the growth rate of both its productive output and its use of resources.

Fourteen operational variables were selected to measure these four indi-

cants. These fourteen variables were

1. Gross domestic product

2. Gross domestic product per capital

3. Energy consumption

4. Energy consumption per capital

5. Steel consumption

6. Steel consumption per capital

7. Distribution of male labor force in agriculture

8. Average annual rate of growth of gross domestic product,

1960 1965

9. Average annual rate of growth of gross domestic product per

capital, 1960 1965









10. Average annual rate of growth of gross domestic product,

1965 1969

11. Average annual rate of growth of gross domestic product per

capital, 1965 1969

12. Average annual rate of growth of gross domestic product,

short period ending 1965

13. Average annual rate of growth of gross domestic product,

long period ending 1965

14. Average annual rate of growth of energy consumption

Of these fourteen variables, the two variables gross domestic

product and gross domestic product per capital were selected to measure

the productive output of a country. The five variables energy con-

sumption, energy consumption per capital, steel consumption, steel con-

sumption per capital and distribution of male labor force in agriculture

were selected to measure a country's use of resources.

Since the measurement of growth rate can be highly dependent on

the time period used, six different variables representing different

time periods were selected to measure a country's economic growth.

These variables measure both the average annual rate of growth of gross

domestic product and the average annual rate of growth of gross domestic

product per capital.

The variable average annual rate of growth of energy consumption

was selected to measure a country's growth rate in use of resources.

All fourteen of the operational variables were measured for each

subject country for each of the years 1965, 1966, and 1967. The arith-

metic mean of the observed data values for the three years was calculated

for each variable for each subject country, and the distribution of








the resultant mean observed data values for each variable was then

standardized.

The standard scores for the fourteen operational variables that

measure economic development and performance were factor analyzed.

Four orthogonal factors were extracted from the data. The factor

loadings for the fourteen variables on each factor are shown in Table 4.

The four factors obtained by factor analysis were named as

follows: Factor 1 was named "economic growth rate," Factor 2 was named

"economic development per capital Factor 3 was named "aggregate eco-

nomic performance," and Factor 4 was named "growth rate of energy con-

sumption." Factor scores were calculated for each of these four fac-

tors for each subject country, and these scores were then included in

the data base.


Test for systematic error and construct validation

The observations of the fourteen measures of economic development

and performance may contain systematic error. This possibility exists

because the less developed countries of the world may exaggerate their

economic development and performance to enhance and strengthen their

image in world affairs. Because developed countries have less of a

reason to exaggerate economic development and performance, any induced

error in economic parameters may systematically decrease as countries

become more economically developed.

The observations of the economic development and performance meas-

ures used in this study were obtained primarily from agencies of for-

eign governments. The ability of these agencies to report discrepant

data may be limited by the presence of a free press that can question








TABLE 4

FACTOR LOADINGS FOR ECONOMIC DEVELOPMENT
AND PERFORMANCE FACTORS


Factor loading


Variable Factor 1 Factor 2 Factor 3 Factor 4


Gross domestic product

Gross domestic product per
capital

Energy consumption

Energy consumption per
capital

Steel consumption

Steel consumption per
capital

Distribution of male
labor force in
agriculture

Average annual rate of
growth of gross
domestic product,
1960-1965

Average annual rate of
growth of gross
domestic product per
capital, 1960-1965

Average annual rate of
growth of gross
domestic product,
1965-1969

Average annual rate of
growth of gross
domestic product
per capital, 1965-1969


-.01


.01

-.04


-,12

.06


.04


.31 .93* -.03


.96* .17 .05

.37 .90* -.06


.26 .02

.94* .03


.08 -.86* -,27


.06 -,09


.75* -.33


.83* -.03


-.04




.01




.05


.19 .27








TABLE 4 continued.



Factor loading


Variable Factor 1 Factor 2 Factor 3 Factor 4


Average annual rate of
growth of gross
domestic product,
short period ending
1965 .96* .06 -.10 -.03

Average annual rate of
growth of gross
domestic product,
long period ending
1965 .93* -.03 .02 .02

Average annual rate of
growth of energy
consumption .10 -.03 -.05 .98*


*Factor loading > an absolute value of .50.


government accounts and independently investigate economic matters. The

more that news agencies are censored, however, the more strictly a govern-

ment may control the dissemination and reporting of erroneous information.

To test for systematic error, correlation coefficients were calcu-

lated between each of the fourteen measures of economic development and

performance and independent observations of freedom of the press. These

observations for freedom of the press were obtained from Taylor and

Hudson (1972) for forty-seven countries of this study's subject popula-

tion. The correlation between freedom of the press and the operational

variable, average annual rate of growth of gross domestic product,

1965 1969, was found to be negative and significant (P < .05). If

governmental control of the press does affect the reporting of economic






84

development and performance data and if contamination of data decreases

with decreasing governmental control, then these significant correla-

tions may be an indication of systematic error. Because systematic

error may decrease with increasing values of the average annual rate of

growth of gross domestic product, 1965 1969, positive correlations

that are obtained in this study and that involve this operational vari-

able will not be harmed by the systematic error. Instead,greater con-

fidence may be placed in any significant correlations that are obtained.

The systematic error that is potentially inherent in the obser-

vations for the annual rate of growth of gross domestic product,

1965 1969, may distort the factor structure that is obtained with

factor analysis. Any results obtained in this study that include this

variable, therefore, must be interpreted with caution.

The correlations between freedom of the press and the operational

variables energy consumption, gross domestic product per capital, energy

consumption per capital, and steel consumption per capital were all found

to be positive and significant. The correlation between freedom of the

press and distribution of male labor force in agriculture was found to

be negative and significant (P < .05). These correlations do not appear

to be indicative of systematic error in the operational variables. That

is, no reasonable link could be made between systematic error and the

operational variables that would result in the observed correlation co-

efficients. These correlations, therefore, may be attributable to

phenomena other than systematic error.

The correlation coefficients involving measures of economic devel-

opment and performance that have not previously been discussed were all






85

found to be insignificant. These coefficients, therefore, do not pro-

vide evidence that is indicative of systematic error.

The test of the economic development and performance measures for

systematic error also provides a test of construct validity. The possi-

bility that systematic error may be contained in the measures for the

annual rate of growth of gross domestic product, 1965 1969,is con-

sistent with what would be expected for this variable. The construct

validity of the variable is therefore supported by empirical evidence.

The results for all the other variables that were tested for systematic

error, however, do not provide support for the construct validity of

the variables. Conceivably, these variables are valid measures of eco-

nomic development and performance but the hypothesized relationship

between the variables and freedom of the press is not valid.


Cultural Unity

The literature on international business has emphasized the

importance of a few cultural characteristics in the development of

business strategies. Religion, ethnic background, and language, for

instance, are often cited as important cultural parameters affecting

business operations in different foreign countries (Salera, 1969;

McCarthy, 1968; Robock and Simmonds, 1973). As an example of how these

cultural parameters may affect business operations, consider the mar-

keting of the same product in both India and Japan.

India has a pluralistic society composed of many different reli-

gious, racial, and linguistic subpopulations. Multiple distribution,

pricing, and promotional strategies may be required to service each of

these separate subpopulations. Japan, however, has a more homogeneous

society composed of people with similar religious, racial, and








linguistic backgrounds. A single distribution, pricing, and promo-

tional strategy, therefore, may be sufficient to market a product in

this highly unified society.

While cultural unity is not explicitly defined in terms of reli-

gion, race, and language by Litvak and Banting, their definition is not

inconsistent with the premises that these cultural parameters do affect

business operations. They defined cultural unity as:

The values, goals, attitudes, social relationships and inter-
actions between distinct segments within a country's people in
terms of shared heritage, unassailed by competing groups. A
"hot" factor when unity is high. (1968, p. 461)

Because the Litvak and Banting definition of cultural unity is

consistent with the concept of culture that is often contained in inter-

national business literature, it was retained in its original form.

Five operational variables were selected to measure cultural unity.

These variables were

1. Religious heterogeneity

2. Linguistic heterogeneity

3. Racial heterogeneity

4. Ethnic and linguistic fractionalization

5. Literacy.

The first four variables were selected to measure the extent to

which a group of people share a single common heritage in religion,

race, and language. The last variable, literacy, was selected because

the ability to read may be important for the dissemination of beliefs,

attitudes, and other cultural elements among a group of people (Robock

and Simmonds, 1973).

The first three variables are coded or "dummy" variables that

possess a value of either 1 or 0. A value of 1 indicates the condition




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs