264 BULLETIN come from the subsidiaries themselves.9 Drawing from a shopping list of products generated by the headquarters unit,subsidiaries choose those that seem appropriate for intensive exploitation in their local markets.As long as the proposed production in the subsidiary seems to have no considerable impact on the facilities of the firm located in other countries,the managers at headquarters are disposed to give the local managers their head. Firms that pursue a policy of this sort can justify their approach readily enough:One possibility is that the firm perceives the cost of interpreting the information needed for pursuing a more centralized policy in production and marketing as exceeding the likely benefits.Another possibility is that the firm has found it impossible to fashion an organization that has the capability for absorbing and being influenced by signals that originate in the subsidiaries.20 Where this pattern of operation exists,the hypothesized behaviour of the pro- duct cycle may still be visible.But the phase of the product cycle in which the parent is responsible for serving foreign markets will be foreshortened and the oligopolistic strength of the innovating firm will be relatively weak,given the existence of firms in other markets that face similar demands and factor cost conditions. Cases in this category will of course deviate from the pattern that a global scanner would generate.First,as long as the subsidiary is the initiator,the geographical spread of products will be affected by the risk-taking propensities and drives of individual subsidiary managers and by the resource slack of individual subsidiaries rather than by a consistent set of decision rules and allocations from the centre.21 Second,in cases in which the initiative for transfer comes from the subsidiary rather than the parent,the possibility of producing in some third country where neither the parent nor the subsidiary is located is unlikely to be considered. All this leads to a simple conclusion.As we search for a hypothesis that would replace the product cycle concept as an explicator of the trading and investing behaviour of the innovating multinational company,a simple variant such as that of the global scanner will not take us very far.Global scanning is not costless,even when a network of foreign subsidiaries is already in place;costs of collecting and interpreting the information,as the firm perceives those costs,may not be commensurate with its expected benefits.In assessing the benefits,flexibility may be a problem:either the flexibility that firms have lost from decisions in the past, or the flexibility they are fearful of losing in an uncertain future 1 For illustrations,see 'IBM World Trade Corporation'and 'YKK(Yoshida Kogyo KK)',both in Stanley M.Davis,Managing and Organizing Multinational Corporations (New York:Pergamon Press. 1979).Also,from Intercollegiate Case Clearing House,see Corning Glass Works (A).(B).and (C) (numbers 9-477-024,9-477-073,and 9-477-074):International Calculators (Australia)Piy.Limited (9-572-641):Veedol France (ICH 10 M 31);The International Harvester Company (B)(9-512-009); Princess Housewares Gmb H(A)(ICH 13 M 117):General Foods Corporation-International Division(D2) (ICH 13G 214);AB Thorsten (A)(9-414-035);and Sanpix Industries (9-278-673). 20 For indications of the formidable difficulties associated with developing such an organizational capability.see Allen,Managing the Flow of Technology,op.cit. 21 This,of course,is a familiar phenomenon,long observed by business historians and organizational behaviourists.More recently the concept has been elevated to the status of theory in Harvey Leibenstein's formulation of his X-inefficiency concept;see his Beyond Economic Man:A New Foundation for Microeconomics (Cambridge:Harvard University Press,1976).264 BULLETIN come from the subsidiaries themselves.'9 Drawing from a shopping list of products generated by the headquarters unit, subsidiaries choose those that seem appropriate for intensive exploitation in their local markets. As long as the proposed production in the subsidiary seems to have no considerable impact on the facilities of the firm located in other countries, the managers at headquarters are disposed to give the local managers their head. Firms that pursue a policy of this sort can justify their approach readily enough: One possibility is that the firm perceives the cost of interpreting the information needed for pursuing a more centralized policy in production and marketing as exceeding the likely benefits. Another possibility is that the firm has found it impossible to fashion an organization that has the capability for absorbing and being influenced by signals that originate in the subsidiaries.20 Where this pattern of operation exists, the hypothesized behaviour of the pro￾duct cycle may still be visible. But the phase of the product cycle in which the parent is responsible for serving foreign markets will be foreshortened and the oligopolistic strength of the innovating firm will be relatively weak, given the existence of firms in other markets that face similar demands and factor cost conditions. Cases in this category will of course deviate from the pattern that a global scanner would generate. First, as long as the subsidiary is the initiator, the geographical spread of products will be affected by the risk-taking propensities and drives of individual subsidiary managers and by the resource slack of individual subsidiaries rather than by a consistent set of decision rules and allocations from the centre.2' Second, in cases in which the initiative for transfer comes from the subsidiary rather than the parent, the possibility of producing in some third country where neither the parent nor the subsidiary is located is unlikely to be considered. All this leads to a simple conclusion. As we search for a hypothesis that would replace the product cycle concept as an explicator of the trading and investing behaviour of the innovating multinational company, a simple variant such as that of the global scanner will not take us very far. Global scanning is not costless, even when a network of foreign subsidiaries is already in place; costs of collecting and interpreting the information, as the firm perceives those costs, may not be commensurate with its expected benefits. In assessing the benefits, flexibility may be a problem: either the flexibility that firms have lost from decisions in the past, or the flexibility they are fearful of losing in an uncertain future. For illustrations, see 'IBM World Trade Corporation' and 'YKK (Yoshida Kogyo KKY, both in Stanley M. Davis, Managing and Organizing Multinational Cooporations (New York: Pergamon Press, 1979). Also, from Intercollegiate Case Clearing House, see Corning Glass Works (A), (B), and (C) (numbers 9-477-024, 9-477-073, and 9-477-074); International Calculators (Australia) Ply. LImited (9-572-641); Veedol France (ICH 10 M 31); The International Harvester Company (B) (9-512-009); Princess Housewares Gmb H (A) (ICH 13 M 117); General Foods CorporationInternational Division (D2) (ICH 13G 214); AB Thorsten (A) (9-414-035); and Sanpix Industries (9-278-673). 20 For indications of the formidable difficulties associated with developing such an organizational capability, see Allen, Managing the Flow of Technology, op. cit. 21 This, of course, is a familiar phenomenon, long observed by business historians and organizational behaviourists. More recently the concept has been elevated to the status of theory in Harvey Leibenstein's formulation of his X-inefficiency concept; see his Beyond Economic Man: A New Fonndation for Microeconomics (Cambridge: Harvard University Press, 1976)