Dynamics of Industrial Location
Size Distribution of Firms
As Dynamics of Industrial Location notes (pp. 191-7), a positively skewed size distribution of firms is a long established feature of industrial economies. A recent, excellent study by Audretsch (1995), not cited in the book, both confirms the resilience of this 'empirical regularity' and offers a 'market-based' theoretical explanation that is more sophisticated and compelling than the traditional approaches which rely on stochastic processes. In particular, Audretsch interprets the size distribution of firms in terms of the differential abilities of firms to use knowledge and to innovate. In his view, the basis for startups of firms is knowledge and innovation so that the rate of startups (or new firm formation as the text calls it) varies by industry in accordance with variations in the underlying knowledge bases of industries. Many new firms die and Audretsch's analysis confirms that death rates are highest among new firms. Apparently, many new firms are unable to grow fast enough to reach minimum optimal size and cannot 'compensate' for not operating at this scale. A few firms do grow quickly, however, and some of these grow large enough to challenge the giants (see chapter 10 and what your text calls 'medium sized firms').
In established industries, 'routinization' implies production practices are well known and accepted, and a lack of new knowledge constrains further innovation and firm entry. Audretsch argues that in recent years tendencies towards routinization, and related trends towards corporate concentration (see file on corporate concentration) have been offset by the creation of new knowledge and a host of opportunities for innovation by new firms. Indeed, his analysis helps explain why throughout the 20th century, what have often seemed like inexorable trends towards ever higher levels of corporate concentration have sooner or later been arrested, at least across national economies. His analysis also has implications for ideas about business segmentation. Thus, the latter stresses the barriers facing the mobility of firms between segments. Audretsch's study, however, intimates the possibility that firms may shift between segments. Consequently, business segmentation itself must be seen in dynamic terms.
See D B Audretsch 1995 Innovation and Industry Evolution. London: MIT Press