258 BULLETIN business in a given foreign market.At that point,the areas to be investigated as possible production sites have been narrowed while the size of the risk has been more explicitly defined.Accordingly,the decision whether to invest in added information is more readily made.Once having felt compelled to focus on the issue,the innovator will decide in some cases to set up a local producing unit in order to prolong some of the advantages that were created by its original monopoly. Two Critical Changes The networks'spread.For the past three decades or so,the process of innovation, export,and investment has been progressing full tilt.One result has been a transformation in the industries in which innovations tend to be especially prominent,such as chemicals,electronics,machinery,and transportation equip- ment.In industries such as these,innovating firms that are limited to their own home markets no longer are very common.Instead,enterprises with highly developed multinational networks of producing units typically account for more than half the global output in their respective product lines. In spreading their networks of subsidiaries around the world,multinational companies have followed some reasonably well-defined patterns.These patterns offer some strong clues regarding the changing perceptions of the enterprises and their likely lines of future behaviour. First,a word on the extent of the spread itself.Table I compares the scope of TABLE 1 Networks of Foreign Manyfacturing Subsidiaries of 315 Multinational Companies 1950and1970s Number of enterbrises 180 US-based 135 MNCs based in UK with neliorks including MNCs and Europe 19501975 1950 1970 Fewer than 6 countries 138 9 116 31 6 to 20 countries 43 128 16 75 More than 20 countries 0 44 29 Source:Harvard Multinational Enterprise Project. the overseas subsidiary networks of a group of the world's largest firms in 195o with the networks of those same firms in the I97os.The dramatic increase in the overseas networks of such firms is apparent. Detailed data have been developed for the 180 US firms in the group,indicating more exactly how the overseas spread took place.10 According to these data,the overseas spread of the firms in our sample was consistent and stable throughout the three decades following World War II.Firms typically set up their subsidiaries, product lines,and new products in a sequence that began with the geographical areas with which they were most familiar,such as Canada and the United Kingdom, and eventually spread to those that had originally been least familiar,such as Asia and Africa.As time went on,however,the unfamiliar became less so,and the 10 The data on which the next few paragraphs are based are presented in detail in Raymond Vernon and W.H.Davidson,'Foreign Production of Technology-Intensive Products by U.S.-Based Multinational Enterprises,cited earlier.258 BULLETIN business in a given foreign market. At that point, the areas to be investigated as possible production sites have been narrowed while the size of the risk has been more explicitly defined. Accordingly, the decision whether to invest in added information is more readily made. Once having felt compelled to focus on the issue, the innovator will decide in some cases to set up a local producing unit in order to prolong some of the advantages that were created by its original monopoly. Two Critical Changes The networks' spread. For the past three decades or so, the process of innovation, export, and investment has been progressing full tilt. One result has been a transformation in the industries in which innovations tend to be especially prominent, such as chemicals, electronics, machinery, and transportation equip￾ment. In industries such as these, innovating firms that are limited to their own home markets no longer are very common. Instead, enterprises with highly developed multinational networks of producing units typically account for more than half the global output in their respective product lines. In spreading their networks of subsidiaries around the world, multinational companies have followed some reasonably well-defined patterns. These patterns offer some strong clues regarding the changing perceptions of the enterprises and their likely lines of future behaviour. First, a word on the extent of the spread itself. Table i compares the scope of TABLE 1 Networks of Foreign Manufacturing Subsidiaries of 315 Multinational Companies 1950 and 1970s Source: Harvard Multinational Enterprise Project. the overseas subsidiary networks of a group of the world's largest firms in i 950 with the networks of those same firms in the i 970s. The dramatic increase in the overseas networks of such firms is apparent. Detailed data have been developed for the 180 US firms in the group, indicating more exactly how the overseas spread took place.'° According to these data, the overseas spread of the firms in our sample was consistent and stable throughout the three decades following World War II. Firms typically set up their subsidiaries, product lines, and new products in a sequence that began with the geographical areas with which they were most familiar, such as Canada and the United Kingdom, and eventually spread to those that had originally been least familiar, such as Asia and Africa. As time went on, however, the unfamiliar became less so, and the 10 The data on which the next few paragraphs are based are presented in detail in Raymond Vernon and W. H. Davidson, Foreign Production of Technology-Intensive Products by U.S-Based Multinational Enterprises', cited earlier. Number of enterprises with networks including 180 US-based MNCs 1950 1975 135 MNCs based in UK and Europe 1950 1970 Fewer than 6 countries 138 9 116 31 6to 20 countries 43 128 16 75 More than 20 countries 0 44 3 29