262 BULLETIN In a world composed of such firms,the product cycle hypothesis would play only a very little role.Although innovating firms might prefer locations in one of the advanced industrialized countries due to the supply of engineers and scientists, the preference for a location in the home market would be weaker.The exports generated by the innovations might come from the country in which the product had initially been introduced;but then again they might not.Whatever the original source of the exports might be,the hold of the exporting country would be tenuous,as the global scanner continuously recalculated the parameters that determined the optimal production location. The hypothetical global scanner,of course,is not to be found in the real world. The acquisition of information is seldom altogether costless;and the digestion and interpretation of information always entails cost.The typical patterns of behaviour that one observes in the real world reflect that fact. One typical pattern,which provides the basis for a second model,consists of firms that develop and produce a line of standardized products which they think responds to a homogeneous world demand rather than to the distinctive needs of individual markets.Some firms have been able to take this approach from the very first,because of the nature of their products;the oil,chemical,and crude metals industries,for instance,were always in a position to develop and purvey a standardized line of products to world markets.But the trend has been moving beyond such products to well-elaborated manufactures:to aircraft,computers, pharmaceuticals,and automobiles,for instance.The trends of the automobile industry in that direction are particularly striking.14 By standardizing their product on a world basis,firms can hope for two kinds of benefit:they can reduce or avoid the costs of processing and interpreting the information that bears on the distinctive needs of individual markets;and they can capture the scale economies of production and marketing on a global scale. Whether those advantages outweigh the disadvantages of being unresponsive to the needs of individual markets is an empirical question the answer to which may well vary by product lines and other factors;those firms that decide in the affirmative for some or all of their product lines cannot be said to be engaged in an irrational response. Firms in this category,innovating for a global market,are obliged to play their innovational gambles for relatively heavy stakes.Accordingly,they can be expected to maintain the central core of their innovational activities close to headquarters,where complex face-to-face consultation among key personnel will be possible;in this respect,such firms are likely to perform consistently with the product cycle pattern.To be sure,with increased ease of communication and transportation,various routine aspects of the development work,not involving the most critical choices in the development process,can be spun off to more distant 14 See A.J.Harman,'Innovations,Technology,and the Pure Theory of International Trade' unpublished Ph.D.thesis,MIT,September 1968,pp.131-134:J.M.Callahan,'GM Adopting Worldwide Purchasing Coordination',Chilton's Automotive Industries,July 1978,pp.47-49:Ford's Fiesta Makes a Big Splash,Business Week,August 22,1977,pp.38-39;and SKF Reintegrates Internationally', Multinational Business,The Economist Intelligence Unit,No.4,1976,pp.1-7.262 BULLETIN In a world composed of such firms, the product cycle hypothesis would play only a very little role. Although innovating firms might prefer locations in one of the advanced industrialized countries due to the supply of engineers and scientists, the preference for a location in the home market would be weaker. The exports generated by the innovations might come from the country in which the product had initially been introduced; but then again they might not. Whatever the original source of the exports might be, the hold of the exporting country would be tenuous, as the global scanner continuously recalculated the parameters that determined the optimal production location. The hypothetical global scanner, of course, is not to be found in the real world. The acquisition of information is seldom altogether costless; and the digestion and interpretation of information always entails cost. The typical patterns of behaviour that one observes in the real world reflect that fact. One typical pattern, which provides the basis for a second model, consists of firms that develop and produce a line of standardized products which they think responds to a homogeneous world demand rather than to the distinctive needs of individual markets. Some firms have been able to take this approach from the very first, because of the nature of their products; the oil, chemical, and crude metals industries, for instance, were always in a position to develop and purvey a standardized line of products to world markets. But the trend has been moving beyond such products to well-elaborated manufactures: to aircraft, computers, pharmaceuticals, and automobiles, for instance. The trends of the automobile industry in that direction are particularly striking.'4 By standardizing their product on a world basis, firms can hope for two kinds of benefit: they can reduce or avoid the costs of processing and interpreting the information that bears on the distinctive needs of individual markets; and they can capture the scale economies of production and marketing on a global scale. Whether those advantages outweigh the disadvantages of being unresponsive to the needs of individual markets is an empirical question the answer to which may well vary by product lines and other factors; those firms that decide in the affirmative for some or all of their product lines cannot be said to be engaged in an irrational response. Firms in this category, innovating for a global market, are obliged to play their innovational gambles for relatively heavy stakes. Accordingly, they can be expected to maintain the central core of their innovational activities close to headquarters, where complex face-to-face consultation among key personnel will be possible; in this respect, such firms are likely to perform consistently with the product cycle pattern. To be sure, with increased ease of communication and transportation, various routine aspects of the development work, not involving the most critical choices in the development process, can be spun off to more distant See A. J. Harman, 'Innovations, Technology, and the Pure Theory of International Trade', unpublished Ph.D. thesis, MIT, September 1968, pp. 131-134;J. M. Callahan, 'GM Adopting Worldwide Purchasing Coordination', Chutons Automotive Industries,July 1978, pp. 47-49; 'Ford's Fiesta Makes a Big Splash', Business Week, August 22, 1977, pp. 38-39; and 'SKF Reintegrates Internationally', Multinational Business, The Economist Intelligence Unit, No. 4, 1976, pp. 1-7