ternal transactions among organizations are managed 

 through ministerial offices and departmental chan- 

 nels. A research institute or design bureau, for ex- 

 ample, reports its results to a technical administra- 

 tion, branch glavki, or industrial association to 

 which it is subordinate. The latter, in turn, de- 

 cides on what should be the next phase of work, by 

 whom, and where. 



This structure impedes innovation in at least two 

 important ways. First, the long approval route de- 

 lays decisions and prolongs the research-to-produc- 

 tion cycle. To create a new machine, for example, 

 requires typically 25 approvals at different levels. 

 To build a new technological system of 10 to 15 ma- 

 chines and mechanisms may require as many as 400 to 

 500 clearances. 2 In general these agreements are ob- 

 tained sequentially and not in parallel. 3 Forward 

 movement is constantly stalled by rounds of negotia- 

 tion; by waiting for approval of reports by depart- 

 mental and interdepartmental expert commissions or 

 for the return of tests on prototypes; by the absence 

 of supplies, equipment, and financing; by the need to 

 find the right customer with the appropriate experi- 

 mental base. Considerable time is spent on corre- 

 spondence and on trips to ministries in pursuit of 

 support for innovation. The path from conception to 

 commercialization can be especially long and precari- 

 ous if the technology entails new processes or prod- 

 ucts unrelated to established interests and activi- 

 ties or if it involves much interbranch negotiation. 

 The effort devoted to gathering signatures of approv- 

 al is due in part to statutory regulations. However, 

 it also serves, Berliner explains, "as a device for 

 limiting each organization's responsibility for its 

 own stage of the work and for reducing one's vulnera- 

 bility in the event of difficulties encountered in 

 later stages."^ 



According to studies by the State Committee for 

 Science and Technology it frequently takes as much 

 time to secure agreements and to transfer documents 

 from one organization to another as it does to con^ 

 duct the necessary scientific development.^ That is, 

 the bureaucratic process of moving research results 



178 



