time to excessive centralization still remains to be 

 seen. Writing in the Academy's main economic jour- 

 nal a year after passage of the statute, two Soviet 

 science experts admit, "While some services are cen- 

 tralized, a system has still not been found of organ- 

 izing the mutual relations of structural units and 

 the machinery of management for the complex as a 

 whole. "63 indeed until 1976 NPOs were not even reg- 

 istered as an independent institutional category at 

 the USSR Central Statistical Administration. All ac- 

 counting was done strictly in terms of their individ- 

 ual structural components. 6^ 



Underlying these problems of the continuing frag- 

 mentation of planning, financing, and management of 

 NPOs are serious and unresolved methodological issues. 

 New integrated performance criteria have not yet been 

 devised. This explains partly, in fact, why minis- 

 tries and higher planning and financial agencies per- 

 sist in issuing plans and funds to separate NPO sub- 

 divisions. Many performance indicators still relate 

 to the activities of R&D and production units in 

 their previously independent status. Existing indi- 

 cators do not differentiate between R&D subdivisions 

 that belong to NPOs and those that do not. According 

 to current methods of accounting and reporting, it is 

 not possible to aggregate the activity of organiza- 

 tions that relate to material production and to the 

 world of nonproduction.6-> 



To be sure, some efforts are being made in this di- 

 rection. Some norms have been devised for determining 

 the average length of the research-to-production cycle 

 and are used in measuring the performance of some NPOs , 

 According to Tabachnikas and Skliar, however, these 

 norms are established rather arbitrarily, largely "by 

 eye." No fixed and uniform methodology exists yet 

 for this purpose. In other associations indicators 

 are used to determine the degree to which the research- 

 to-production process has been reduced over time. Tak- 

 sir points out, however, this kind of norm is of dubi- 

 ous value because reduction of the innovation cycle 

 obviously has a limit. 66 What methodological progress 

 has been made in developing integrated evaluative in- 



217 



