THERORETICAL BACKGROUND

Diversification Classification

Prior studies classify a firm’s diversification strategy mainly by product, market, or

industry. There are also other types of categorization. For instance, based on a firm’s supply

chain, product, market, and industry, Ansoff (1957) classifies a firm’s diversification as

vertical diversification, horizontal diversification, and lateral diversification. Furthermore,

from firm growth perspective, Ansoff (1965) regards diversification as firms providing new

products to new markets based on the existing products and markets. Thus, by using these two

constructs, product and product, Ansoff (1957, 1984) defines four types of the diversification

strategy: market penetration, market development, product development, and product

proliferation. Similarly, Wrigley (1970) also uses product category to define diversification.

while Varadarajan and Ramanujam (1987) attempt to employ the number of industries a firm

involves and the average number of products in each industry to classify a firm’s

diversification strategy.

The advantages of this traditional method to classify diversification include objective

measurement and easy operationalization whereas the limitations are that it may neglect the

importance of capabilities or resources, which is transferable among different business units.

For instance, computer retailers and food retailers are two different industries but their

retailing activities may be similar. If diversification categorization is based on products,

markets, or industries, then computer retailing and food retailing may be classified as

un-related diversification. However, since the capabilities or resources of computer retailing

and food retailing, firms may transfer its existing capabilities or resources from the original

business unit to the new business unit without losing competitive advantage. Therefore, the

classification by product, market, or industry may not fully capture the generic picture of the

diversification strategy. Thus, in addition to product, market or industry, we attempt to explore

other constructs for classifying a firm’s diversification.

Rumelt (1974) asserts that a firm’s diversification strategy not only should consider its

positioning of products and markets, but also should consider whether diversified businesses

share the same activities in technologies, capabilities, and resources. Aaker (1984) also

suggests that if two business units have similarity in operational activities, such as research

and development, production technologies, or distribution channels, then synergies can be

created due to the scale of economies as well as the exchange of technologies or resources

between two business units. Moreover, Hamel and Prahalad (1994) use two constructs,

competence and industry, to examine how firms attain competitive strategy. Hamel and

Prahalad (1994) assert that it is not market or industry but a firm’s core competence to

determine the firm’s competitive advantage. Base on the above discussion, we conclude that

capabilities or resources (Rumelt, 1974), operational activities or value activities (Aaker, 1984)

and competences (Hamel and Prahalad, 1994) are more important constructs to classify the

diversification strategy which determines a firm’s competitive advantage.

Related or Unrelated Diversification

As mentioned earlier, prior studies use product, market, or industry to measure the

degree of diversification. Gort (1962), for example, defines diversification as the concept of

heterogeneity of output depended on the number of markets served to those outputs. A firm’s

degree of diversification is also measured by the number of industries (Berry, 1975) or the

number of businesses (Wood, 1971; Pitts, 1977). Traditionally, two approaches are used to

measure the degree of diversification (Davis and Duhaime, 1992). The first approach is a

categorical measurement, which give scores based on the characteristics of diversification

while the second approach is a continuous measurement, which uses a scale value between

related and unrelated diversification. The former is mainly built from the works of Wrigley

(1970), and the latter is derived from Standard Industrial Classification (SIC) Code system,

which is the most common measurement for diversification. Wood (1971) uses SIC code to

calculate and define broad spectrum diversification and narrow spectrum diversification.

Rumelt (1974) also uses SIC code to develop indicators for measuring the relatedness of diversified businesses, including specialization ratio (Rs), related ratio (Rr), vertical ratio (Rv）

and related-core ratio (Rc). Based on these four indicators, he then defines nine types of

diversification. The advantage of using SIC code to calculate the degree of diversification is

its comprehensive definition and easy operation. However, since this approach does not

consider sale volume for each product, it is unable to understand the real focus business of a

diversified firm. Therefore, some scholars attempt to provide amended measurement.

Jackquemin and Berry (1979) develop the Herfindahl indicator and Entropy indicator, which

is the most well-known tool for measuring a firm’s diversification. Both the Herfindahl

indicator and Entropy indicator use sale volume of each product to measure diversification

but the Entropy indicator further weights different products (or business units), which can

authentically reflect the degree of a firm’s diversification. Ramanujam and Varadarajan (1989)

also use two dimensions of the Entropy indicator to define four types of diversification.

However, the standard SIC code, Herfindahl indicator, or Entropy indicator all base on

the types of industries or products. As discussed earlier, these classifications are unable to

explain what or why the competitive advantage can be transferred from an old business to a

new business. It can only evaluate the relatedness between diversified businesses, but hardly

shows how these diversified businesses achieve the synergy of resources, technologies or

capabilities. This urges us to further explore the other constructs or dimensions to measure the

relatedness of diversification.