BOOK REVIEW 73 accounting u^it and production d^mon^ ^re regulated by ptofitaNfity, A profit retention system, according to him, would accelerate techno-» logical progress and the operation of the "market" would close the demand-supply gaps. However, rp^jor exp^n$iron decisions of the enterprises ought to be controlled, both by a sysUm of bank loans and centralised investment decisions by the state. The problems relating to th® 'centralised' and 'decentralised systems of planning have be^n extensively debated in the East European Case. The general consensus has been that the centralised planning system would be necessary in the "extensive phase^ of growth of these economies, where it is possible to shift people from low productivity to high productivity occupations. However, with the transition to an "intensive phase" of growth, when this is ao W>re possible because of labour shortages, there have been stro»g advocates for a decentralised planning system, because growth can take place only by a conscious effort to raise labour productivity through technological innovations, for which the centralised planning mechanism might be unsuitable. Xue gives no indication as to whether such a transformation has occurred in the case of China, in arguing for a decentralised planning system. Within the decentralised planning system, there are two variants. One is the Yugoslavian system, where individual enterptis^s have almost complete autonomy over all decisions, including those ®f substantial expansion of capacities. This could very well lead to all kinds of disproportionalit^s, as i^ fact has happened in Yugoslavia. The other is the Hungarian variant, where enterprises are-free lo exchange commodities with each -other and also carry certai4 technological innovations independently, but major expansion decision^ of individual enterprises are still controlled by the state in tU interests of orderly planning. It is not clear as to which of thes^ variants Xue is advocating for China, as their implications fo^ socialist development are radically different, j However, there are indications to suggest that Xue's posing of the problems of planning in China as a choice between the central-Used and decentralised systems is probably unjustified. As note<^ above, in the "extensive phase" of growth, with a limited set of priorities, centralised target fixation is likely to be a more efficient mechanism of planning. In fact, it is possible to argue that tht problems in China stem not so much from too much centralise<)l planning, as Xue assumes, but the lack of a proper system of centralised planning. While to start with. planning was highly central lised, with alt important targets being set by ministries in Beijin^ responsible for the various industries, by 1957 a large measure of regional decentralisation had occurred, with a whole set of decision making coming under the control of the provincial authorities. In China after 1957, material balances were mostly fixed by the