The ''Peoples' Plan campaign'' in Kerala, which led to the devolution of nearly a third of the Plan outlay of the State to Local Self Government Institutions (LSGIs) to spend on projects of their choice, has of late generated much controversy. Many have even argued that ''decentralization'' is a part of the imperialist agenda and that the Left's adoption of it represents a capitulation (which can be explained only through the ''erroneous'' activities and understanding of some leading Comrades). This entire line of criticism however misses a basic point, namely that the ''decentralization'' proposed by the Left, propagated through the Peoples' Plan campaign, and implemented during the years of LDF rule is fundamentally different from the ''decentralization'' promoted by the World Bank and other imperialist agencies. While the ''decentralization'' agenda of the Left is a means of carrying class-struggle forward, of buttressing the class-strength of the rural poor by developing institutions where they can, in principle, assert themselves directly and hence more effectively, the ''decentralization'' promoted by the imperialist agencies has precisely the opposite objective, of blunting class-struggle, of encouraging a scenario of ''obedient-and-supplicant-villagers-being-patronized-by-NGOs'', and of substituting the concept of the ''Rights'' of the people by the concept of ''Self-Help''. In pushing their agenda, it is in the interests of the imperialist agencies to pretend as if there is no difference between their programme and the Left programme. But for that very reason it is essential for the Left to emphasize in every possible manner the difference between the two agendas, to highlight the specificity of the imperialist agenda, if it is to defeat the latter.

I

It is a deliberate ploy of imperialism to borrow concepts from the Left and incorporate them into its own lexicon after giving them a different meaning. This habit of borrowing concepts serves many purposes: first, the analytical depth that is usually associated with any Left concept is mistakenly attributed to the imperialist concept as well and imparts a prestige to imperialist theory; secondly, since the same concept is used by the enemies of the Left as by the Left itself, it creates confusion in Left ranks, and blunts a basic weapon of the latter, namely theoretical clarity; thirdly, by creating fuzziness around basic concepts imperialism seeks to destroy clarity of thinking in society as a whole, which helps to establish and perpetuate its hegemony.

Consider an example. Terms like "structure", "structural change", "structuralism", and "structural reforms" were originally developed in the ranks of the Left as a critique of imperialist theory. As against the standard imperialist view that the free market was the panacea for the ills of the third world, the Left advanced the theory that the third world needed a rupture with the structures inherited from its colonial and semi-colonial past, via land reforms, via de-linking from the imperialist-dominated world market, via building up the public sector as a bulwark against metropolitan capital, and via a rapid process of planned industrialization. This was so clear an intellectual divide that the term "structure" was as much a short-hand for Left thinking as "market" was for the Right. And yet today the term "structural adjustment" has become a central part of Fund-Bank theory; it is given a meaning that is precisely the opposite of what it originally meant within the Left lexicon, namely "opening" the economy to the free play of the so-called market forces. Imperialist agencies simply took over a Left concept and used it for their own purposes after giving it a different meaning.

The same is happening today to the concept of "decentralization", which has been, for a very long time, and particularly in India for the last quarter of a century, i.e. since 1977 when the Left Front government came to power in West Bengal, an integral part of the Left lexicon. This was long before the so-called "economic reforms" began and the Fund and the Bank started strutting around on the Indian scene. But soon they, and other imperialist agencies like the ADB and DFID, started talking about "decentralization" as if they had invented the term, and giving it a fundamentally different meaning from what it had as part of the Left lexicon. So successful has been their ploy that today, as mentioned earlier, many who claim to owe allegiance to the Left have actually started demanding that the Left should have nothing to do with decentralization which is an imperialist concept and hence a part of an imperialist conspiracy. This is "liquidationism" at the level of theory. Instead of exposing the imperialist game of taking over Left concepts and exploiting their legitimacy for its own nefarious ends, this perception actually accepts imperialist claims of having authored the concepts themselves, and abandons legitimate democratic issues on the absurd ground that if imperialism is talking "about the same thing" then they cannot be legitimate democratic issues. The whole point however is that imperialism is not talking about the same thing.

My purpose in this paper is to distinguish between imperialism's concept of decentralization and the Left concept which, as already noted, is far older and rooted in its own philosophical tradition. I shall not however, for reasons of space, be going into the whole gamut of Left thinking on this question. I shall confine myself only to the concept of decentralized planning as conceived of and practised by the LDF government in Kerala, examine its specificities, and show the difference between this practice and the one advocated by the imperialist agencies.

II

Two basic questions can be asked about Kerala's "decentralization" experiment. First, what is its relationship with planning? And secondly, what is its link with class-struggle in the countryside? In this section I shall address the first of these questions, keeping the second to the next section. Section IV discusses the contrast between this "decentralization" and the imperialist concept. The last section contains some concluding observations.

Even traditionally, i.e. prior to the intrusion of the imperialist concept, the term "decentralization" had been used in different contexts to mean several different things. To examine the specificity of the Kerala experiment therefore let me begin by distinguishing between some of these traditional meanings.

There are at least four different senses in which the term has been used in the economic literature. The first is decentralization as in a capitalist market economy, where there are a number of separate decision-makers, such as firms and consumers, and the overall functioning of the economy, its performance in any period and its dynamics over time, is a resultant of these dispersed individual decisions. By contrast a planned economy in its classical meaning is supposed to function on the basis of centralized decision-making. A comparison between the two systems of decision-making used to be a common theme in economics at one time, and the case for the superiority of the centralized system was argued on the grounds that it overcame the "anarchy of the market" to which Marx, and Keynes, in their different ways had drawn attention^[1] . (Maurice Dobb and Michael Kalecki were perhaps the most persuasive presenters of this argument).

The other three senses in which the term has been used refer essentially to the context of a planned economy itself, but one which despite being planned is not centralized in the sense of being a "command economy". The second meaning of the term has to do with decentralization as an arrangement for information flow in a planned economy, such as was proposed famously by Oskar Lange and later by Kornai and Liptak. Of course the Lange vision of decentralization, and that of Kornai-Liptak, referred not to actual decentralization but merely to a procedure of central planning, to the operation, as Joan Robinson was to put it, of a "pseudo-market" in a planned economy, so that the plan gets finalized on the basis of a two-way flow of information between the central planner and the individual enterprises.

The idea is the following. The Central Planner announces a set of prices (these prices are only announced, they do not actually rule in any market), and on the basis of these prices the enterprises inform the Planner how much they would like to produce in order to maximize the enterprise profits. Since the sum of these output decisions would differ from what the Planner might wish to have as the Plan target, it would then announce a set of revised prices; and so the ''iteration'', i.e. the to-and-fro movement of information, would continue until a final Plan is prepared. The sole advantage of this ''iteration'' proposal is that the Planner does not need to have detailed knowledge of all the resources and raw materials available with each and every enterprise^[2].

Even though this ''iteration'' is supposed to occur only through a "pseudo-market", i.e. only through an exchange of information to which there is no actual market counterpart, one could in principle even make this ''iteration'' occur in actual practice. The Central Planner could announce a set of prices, leave it to enterprises to produce amounts they like at these prices, and hold unsold amounts as stocks. In all cases where the stocks were large, the prices could be lowered and where they were small, prices could be increased, and so on. There would not then be one overall final plan, but a series of adjustments on an initial blue-print, approximating towards but never actually reaching a final consistent document. Room would have to be made in such a case through appropriate institutional provisions (e.g. larger inventories etc.) for the realization of such an approximate plan, but a planned economy of this sort could still function in a manner that curbs on the whole the anarchy of capitalism.

The third sense in which the term decentralization has been used is in the context of the view that the unit over which planning is done is not the country's economy as a whole but the economies of particular parts of it. Here we have not a change in the model of central planning, but a splitting up of the units over which the central plans are formulated. (And if we are talking not of a fully-planned economy but of a mixed or partially-planned economy, such as India was in its dirigiste phase, then "decentralization" would refer to a reduction in the size of the unit over which such partial planning is done.) An idea was current among several progressive thinkers in the country at one time that, barring a few areas such as defence, communications, and foreign policy etc., the rest of the functions of the government, including in particular development, should be undertaken by the State governments. Whether the Centre makes actual resources available to the States for this purpose, or resource-raising powers are themselves transferred to the States, this perception of "decentralization" amounts to a change not in the relation between the plan and enterprises but in the unit over which planning, of whatever variety, occurs.

The fourth concept of "decentralization" refers to a situation where, if an individual or collective unit is to be the beneficiary of a set of projects in a particular sphere, then they should be allowed to finalize the precise mix of such projects. This is nothing else but an extension of the principle that when it comes to certain kinds of goods, it is better that individuals (or families) are given the cash to purchase these goods^[3] , rather than these goods being made directly available to them, a principle that often goes under the grandiose (and, in capitalist conditions, misleading) title of "consumers' sovereignty". In certain spheres in other words, "users' sovereignty" should be respected.

The last two meanings of the term "decentralization" might appear to be too close for drawing any worthwhile distinction between them, but there is a significant difference in "quantity" between the two cases which almost amounts to a difference in "quality". I perceive these four cases in the following terms respectively: "decentralization as anarchy", "decentralization as iteration", "decentralization as multiplication" and "decentralization as users' sovereignty". Of course in practice these different cases come not in pure form but often jumbled up, and neat pigeon-holing of any actual case of "decentralization" into any one of these cases may be difficult; nonetheless this distinction is useful for assessing the implications of particular cases of decentralization. For instance it would be grossly erroneous to see the Kerala case of "decentralization" through "peoples' planning" as engendering either "anarchy" or "multiplication" or mere "iteration". It is quite clearly of the fourth kind. It does not in principle negate planning (even of the sort we have in India); rather it is based on invoking, implicitly, the concept of "users' sovereignty" as justification for the devolution of plan funds to LSGIs for undertaking certain kinds of projects within an overall plan.

Two caveats have to be entered here. First, if a plan is to have consistency, i.e. various commodity balances have to be satisfied, then the fact that the mix of projects is left to the individual LSGIs would have problematical implications. Flow excess demands and flow excess supplies of particular commodities would arise on account of the fact that the decisions of the individual LSGIs are not co-ordinated ex ante. But this problem (as in the second case discussed above of real-life ''iterative'' planning) can be resolved through the holding of appropriate inventories. Secondly, if the LSGIs are to be not merely plan-fund-using institutions but are to take some initiative in mobilizing resources of their own and using these for their own purposes, or even taking on productive roles and using the proceeds obtained from doing so for their own purposes, then these activities and the expenditures generated on account of them would constitute "add-ons" to the basic plan of the economy. These "add-ons", since they are not based on any ex ante co-ordination, could upset the basic balance of the plan by creating ex ante flow excess demands and supplies. Once again however the problem can be handled by making prior provision for such contingencies through the appropriate provision of inventories or by maintaining a higher level of planned excess capacity in the productive sectors within the framework of the plan itself.

It follows then that the "peoples' planning" experiment of Kerala, while it entails "decentralization", in the sense that the budgeted amount earmarked as expenditure on certain items in the plan, such as rural infrastructure, is handed over to the LSGIs in order to reflect better the preferences of the users, does not mean a negation of planning. On the contrary, it can be defended on the grounds of "users' sovereignty". True, the amount handed over to the LSGIs has been fixed as a percentage of the total plan outlay rather than being independently determined on the basis of the estimated expenditure under the relevant heads, but the figure for this percentage itself has been arrived at by considering the approximate proportion of expenditure actually incurred in the past on these items. Hence, the "users' sovereignty" argument is not undermined by the fact of the devolution being fixed as a percentage of plan funds.

The additional merit of the Kerala experiment consists of the fact that "users' sovereignty" is sought to be exercised through collective bodies directly or through their democratically elected representatives. This brings me however to the second question that can be asked about Kerala's decentralization experiment.


[1] Keynes who was a liberal bourgeois economist aware of the fact that capitalism was subject to anarchy, wanted to preserve capitalism against the socialist threat by reforming it so that the basic property relations remained unchanged but there was "socialization of investment" through State intervention.

[2] For a discussion of these issues see Maurice Dobb (1969).

[3] For a discussion of this principle as well as where it should not apply see Dobb (1969), Chapter 10.