1. When I was a student in the 1960s, we basically thought of economics as being divided into two segments: there was "economic theory" on the one side, and there was a whole mish-mash of economic history, "Indian economics" and "development economics" on the other. Those who were very bright and mathematically competent (the two were taken as synonymous) did the former and were on a par with the most highbrow sahibs , while those who were less so did the latter (large chunks of which were sometimes pejoratively referred to as "cowdung economics"). Nowadays, ever since a host of game theorists and other mathematically-oriented economists have invaded the domain of development economics, this absurd stigma attached to the latter has diminished somewhat, though it has by no means disappeared, but the basic dichotomy still persists, between "economics" (or "economic principles" if you like) and "development economics". In almost every university in the world, including every Indian university, there is a whole set of lectures on "economic theory" which makes absolutely no reference to the developing countries or their problems. Then there is a whole set of separate lectures on "development economics" which deal exclusively with these problems, in which some of the tools of analysis learned in the first set of lectures are occasionally applied. Now, the question arises: how can such a dichotomy be justified?

2. The most obvious answer to this question would probably be as follows: the developed economies and the underdeveloped economies have very different institutions, notwithstanding the fact of an overlap of institutions between them. The former represent by definition a higher level of development than the latter. It follows then that the institutions prevailing in the former must constitute the central focus of analysis, from which the study of the latter economies, though necessarily separate because of their specificities, can still benefit, both because of the overlap referred to, and also because the former represent the ultimate goal. Looking at it differently, since developing countries are those which happen to lag behind the developed ones along the road of development, there is an obvious rationale in studying the institutions of the former as the central focus and then looking at the specific reasons why the laggards lag behind.

This perception, interestingly, is not confined to "mainstream" thinking alone; it is to be found even within the Marxian tradition. Marx's remark that "the country that is more developed industrially only shows, to the less developed, an image of its own future" (1977, p.91), which was made in the context of Britain and Germany, is often universalized to cover the case of even the underdeveloped economies, which are presumed to follow, quite spontaneously in the normal course, the path traversed by Britain and the other developed countries. From this it follows that focussing on the analysis of these developed economies taken in isolation constitutes the first priority while the reason why some others lag behind can be discussed subsequently. At any rate the two inquiries are essentially separate according to this perception.

Of course one can have an alternative, and more sophisticated justification for this procedure within the Marxist tradition. One can argue that the law of motion of a particular mode of production, whether capitalist or feudal, seen in its internal unity and as an "ideal type", must take precedence over, or form a prelude to, the "concrete analysis of the concrete conditions", of economies caught in a process of transition, with admixtures of capitalist and pre-capitalist characteristics. On this reasoning, whether or not the underdeveloped economies ever manage to make the transition to capitalism , i.e. whether or not they are only "laggards" who would eventually "catch up", there is a purely theoretical case for a dichotomy between two areas, which we traditionally designate as "theory" (focussed primarily on, but not referring concretely to, the advanced countries) and "development economics" (referring concretely to the underdeveloped countries).

3. There is however a basic flaw in this entire perception: it sees the developed economies, or "capitalism" for that matter, in exclusive isolation. It is predicated on the view that capitalism is a self-contained system which can therefore be legitimately analyzed in isolation from its pre-capitalist environment. Indeed the reason why both "mainstream" theory and certain tendencies within the Marxist tradition converge on this question of dichotomy is that they both view capitalism as a self-contained system, no matter how basic their other differences about its nature may be. To be sure, the existence of trade relations is recognized, but "mainstream" economics sees trade merely as enlarging the opportunities available to a country, but not as essential for an economy's existence and growth (on this more later). Trade in short is seen to benefit every country, and to constitute an additional bonus for all. Trade theory, though a part of "economic theory" (indeed a special case for the application of its basic conclusions) does not therefore enter the core of it.

Even within the Marxist tradition, the scope for viewing capitalism in isolation, as being self-contained, arises, despite Marx's acute awareness of the phenomenon of colonial exploitation and his remarks about the "primitive accumulation of capital" through colonial loot, because in his model of Capital (Volume I) international trade scarcely figures. Whether this is a result of the imprint of Classical Political Economy (especially Ricardo) on Marx, or a reflection of the unfinishedness of Marx's project, is beside the point. It makes possible the emergence of the above-mentioned theoretical dichotomy within the Marxist tradition as well, even though the literature on imperialism that came up around the First World War within this tradition, makes the tradition as a whole immune to the charge of upholding this dichotomy.

The result of this dichotomy, however, which "mainstream" economics so assiduously cultivates, is to impoverish both "economic theory" as well as "development economics" . In other words it does not merely detract from an understanding of the problems of development; it makes the so-called economic theory itself largely irrelevant. I shall cite three examples from economic theory to illustrate this point.

4. My first example relates to "mainstream" Growth Theory. I shall refer explicitly only to Solow's model (1956), even though what I say applies to Growth Theory in its other incarnations as well. The basic conclusion of the theory is that the rate of growth of a capitalist economy is tethered in the long-run to the exogenously given rate of growth of its labour force in "efficiency units" (in recent "endogenous" growth theories, this rate of growth is not entirely exogenous, but the exogenous component nonetheless exercises a restraining effect on long-run accumulation^[1]). This is because, unlike in the Classical or Von Neumann models, labour here is a "rent good"; it cannot be internally produced or acquired in adequate quantities at the going wage rate.

At first sight this appears a plausible assumption. Even though the advanced countries may not have full employment, they have reserve armies of labour which can get exhausted quickly; and the very process of exhaustion would put pressure on the wage rate and lower the rate of accumulation (as Goodwin (1967) had argued). What is wrong then in asserting that accumulation is constrained by the exogenously given rate of growth of the labour force?

The error consists in believing that capital has available to it only the labour force of its country of origin. In fact the entire labour force of the world is at its command, and the reserve army at the world level is so large that the constraint on accumulation cannot conceivably be attributed either to any absolute labour shortage or to any pressures on the wage rate arising from the process of its exhaustion.

This is no idle speculation. In fact, throughout history, capital accumulation by the metropolis, whether located in the metropolis itself or outside of it, has always drawn on world labour reserves (located in particular in India and China) whenever the need has arisen. In the nineteenth century, there were two huge streams of migration: of European labour migrating to the temperate regions of white settlement, occupying land by forcibly dispossessing the local population, and having capital migration following in its wake; and of tropical labour, mainly from India and China, migrating to other tropical lands to serve the needs of European capital in mines and plantations. These two streams were kept strictly separate by migration controls, and their wage rates (or incomes per head) too were vastly different (owing to the former's access to "free land"). But each of these was a mighty stream, involving as many as 50 million persons^[2]. Similarly, in the post-Second World War years when Keynesian demand management lowered unemployment in the metropolis, substantial migration took place from less developed countries to the metropolis to serve the needs of capital, from Turkey to Germany, from Algeria to France, from the Indian subcontinent to the U.K., from Mexico to the U.S., and so on. Given this enormous scale of migration that has unfailingly occurred historically to serve the needs of capital, to believe that capital meekly adjusts to the "natural rate of growth" of labour force within its country of origin is nothing short of a travesty. Indeed, as Nicholas Kaldor came to emphasize in his later writings (Kaldor 1978), the one thing that capital accumulation has never been constrained by is shortage of labour. But the poverty of "mainstream" theory here arises because of looking at the capitalist economy in isolation, as a self-contained entity, with no access to labour from pre-capitalist economies.

5. The second example relates to the so-called Non-Accelerating Inflation Rate of Unemployment (NAIRU)^[3]. The monetarist version of it is the Natural Rate of Unemployment (NRU) which is nothing else but de facto full employment. Let us consider the monetarist version first. At NRU the supply price of labour equals the marginal product of labour. The level of employment can be increased beyond what is given by the NRU, but only if the workers do not anticipate inflation. And even in such a case, it can be maintained at this level only if expectations are adaptive and not rational, and then too only at the expense of accelerating inflation. The presumption underlying all this is that a given amount of labour is supplied only if a particular real wage is expected to be obtained; if the expected real wage is lower, then only a smaller amount of labour would be supplied. Now, in third world economies saddled with massive labour reserves this is patently untrue; even when real wages secularly decline, without there necessarily being any divergence between the expected and the actual real wage, the labour supply never starts drying up. From this fact it follows that in so far as commodities produced by third world labour enter the production process in the metropolis as inputs, talking of a Natural Rate of Unemployment is meaningless.

Even if we take the more general concept of NAIRU which is in no way linked to de facto full employment, the same question again arises: as long as there are workers, located within an ocean of labour reserves, whose ex-ante wage-claims are compressible in real terms, we cannot talk in terms of a NAIRU. If these workers are located outside the capitalist economy, so that their lower real wage rate expresses itself as lower terms of trade for the outside economy's product vis-a-vis that of the capitalist economy, then one can talk of a particular level of NAIRU within the capitalist economy, associated with each level of the terms of trade . But since the real wage rate of these workers is ex ante compressible, i.e. can be actually lowered without causing accelerating inflation, it follows that the terms of trade can always be turned against their products, to the extent required to prevent accelerating inflation in the metropolis, no matter what the level of employment. It follows in other words that the capitalist economy can settle at any level of employment without experiencing accelerating inflation, as long as labour reserves exist outside of it, in the surrounding pre-capitalist economies. The fact that the contrary has been asserted, and on the basis of it the possibility of Keynesian demand management has been denied, is only because the capitalist sector has been seen in isolation, as a self-contained entity.

6. My third example concerns Trade Theory^[4]. There is a basic presumption underlying the theorems of Trade Theory which is worth re-iterating, if only because it is not always appreciated. And that is the following.

[1]This is true of all "endogenous" growth theories in which the attainment of an equilibrium is not dependent on a specific rate of labour force growth. In these models the larger the rate of labour force growth the higher is the rate of growth of capital stock and output in steady state. There are however some theories in which equilibrium can be attained only with a stagnant population (any positive rate of population growth makes the economy explode). In the case of these, the assertion in the text that the rate of accumulation would be higher for a higher growth rate of the exogenous component (viz. labour force) would not be justified. But these models, no matter what one thinks of their insights, must be considered structurally flawed in this respect.

[2]W.Arthur lewis (1978, p.14) writes: "The development of the agricultural countries in the second half of the nineteenth century was promoted by two vast streams of international migration. About fifty million people left Europe for the temperate settlements...About the same number- fifty million people- left India and China to work mainly as indentured labourers in the tropics on plantations, in mines, or in construction projects...".

[3]The argument of this section has been developed at length in Patnaik (1978).

[4]The argument given in this paragraph is taken from U.Patnaik (2003).