Must development economists watch Hamlet without the prince?

Naqvi, Syed Nawab Haider

..... To be, or not to be: that is the question"

Hamlet.

Development economics, according to the non-believers, has changed hands in the last 30 years of its existence from the visible hand of the State to the invisible hand of the market--and, a fortiori, from being to nothingness. Correspondingly, the vision of an economic universe beset by all-pervading market failures seems to have been replaced by one of total government failure. In this revised vision, government intervention only spoils the utility/profit-maximizing show by giving rise to various activities of a rent-seeking type [Krueger (1974)]. The development experience of the last four decades--especially of the "Gang of Four", i.e., South Korea, Singapore, Taiwan, and Hong Kong--has been cited to establish the superiority of the market-based solutions to the dirigiste solutions of the development problem. Notwithstanding the Hume's Law--which prohibits deducing 'prescriptive' statements from factual statements alone--the agnostics have used these real-life happenings to prove the central point that (neo-classical) economists have made all along: optimal allocative efficiency is (best) guaranteed by following faithfully the first-best rules of competitive efficiency, which free markets alone are believed to satisfy. Citing the (glaring) contrast between free markets and centrally-planned economies, and the recent disavowal of the socialist economic philosophy by Eastern Europe, Haberler concludes: "I still maintain my early belief in the validity of classical or neo-classical theory and in the superiority of relying largely on competitive markets and private enterprise." [Haberler (1988)]. In the same vein, the ultra-conservative Economist (London) declares: "After three decades the experience of these countries [shows that] history chooses the invisible hand." [Crook (1989)].

Thus, it is claimed that development economics, born from the rib of an interventionist government, has met its nemesis. What should then be done? The standard prescription is: "Get the prices right" and leave the rest to the safe hands of the atomistic utility/profit-maximizing invisible agents in the wonderland of pure and perfect competition. And in this enthusiasm for the "right" things, externalities of various kinds as well as considerations of income distribution and unemployment are relegated to the background.

In my previous Addresses to the Society, especially in "A Tale of Two Hands" (1986) and "For Morality's Sake!" (1989), I have questioned the validity of these arguments. I have shown that development economics as a discipline exists, though not necessarily in splendid isolation; that its worldview of a "mixed economy", run with the help of both the "visible" and the "invisible" hands is essentially correct; and that its concern with essentially macro problems of growth as well as equity is well-placed. While considerations of (static) efficiency invite markets, where they exist, the dictates of equity require government intervention, especially where markets do not exist due to external economies of various kinds. The latter is the case when institutional constraints--those set by the locus of asset ownership and by the mode of property relations--must be removed as a necessary condition to achieve efficiency in resource allocation and to ensure equity.

In the present Address, I restate some of these arguments somewhat differently in the light of the recent literature on the subject. I maintain that the basic flaw in the anti-dirigiste argument flows from the error of contraposing the government and the market as rivals, and from considering the choice between the two as mutually exclusive. Such contraposition may be justifiable in communist countries; but it does not make much sense in developing countries. This is because a mixed economy has been accepted from the very first 'day' as the basic model. I show that the cult of unqualified anti-dirigisme, and the expiatory mad rush to privatize, is ill-founded both in theory and practice; and that its frequent appeals to the first-best rules of market efficiency are out of place--and nothing more than what Findlay (1988) calls "Pareto-optimality by inadvertence". But efficiency-oriented Pareto-optimality, which has been shown to be linked with competitive equilibrium, is not suitable at all to model 'reality' in the developing countries. This is because considerations of growth and equity--indeed, of equality--are inextricably bound with each other. To recognize this 'fact' we pay attention not only to individual preferences but also to the significant differences between these preferences; the intensity of these preferences should be 'weighed', and interpersonal comparisons between them should be made, too. In other words, development economics should not fight shy of making normative judgements.

But, first, let me look closely at the claim that the alleged victory of the market over the government denotes the demise of development economics, which is assumed to be incurably dirigiste in disposition.

THE STUFF THAT DREAMS ARE MADE ON

The "State of Bliss"

The essence of the anti-dirigiste cure-all prescribed for development economists is: "get the prices right"; and to do that, unshackle the market forces by 'privatizing' across-the-board. This is required to satisfy the first-best rules of competitive efficiency, which entails ensuring the postulated equality between the domestic marginal substitution (in consumption) and the domestic marginal rate of transformation (in production), and that between the domestic rate of transformation and the foreign rate of transformation (through foreign trade), which in turn is equated to international prices. The implication is that competitive equilibrium will lead to an efficient allocation of domestic resources if the solution is secured through the market rather than through state intervention. Another remarkable result cited in support of such a market-oriented efficient solution of the resource allocation problem is that--by the Fundamental Theorem of Welfare Economics--"every competitive equilibrium is Pareto-optimal, and every Pareto-optimal state is characterized by a competitive equilibrium". This two-way relationship between Pareto-optimality and competitive equilibrium rests on the "duality" property of an efficient resource allocation problem. Every maximum welfare problem has embedded in it a set of (shadow) prices, which correspond to optimal input prices--wages, rents, interest rates. But for that to happen, a perfect, self-policing competition must obtain in all markets. This, in turn, requires that (contingent) markets exist for all situations, and that these markets 'clear' at all times. The textbook world (of dreams) where all this happens is referred to as the "state of bliss", resplendent with all-round convexities and multi-market Walrasian equilibria.

The Privatization Alchemy

Does it follow, then, that a policy of "getting the (relative) prices right" will actually lead a real-world economy--made supplicant to price signals by large-scale privatization--to the state of bliss promised in the standard economics textbooks? As unbelievable as such a claim may seem, this is exactly what some economists seriously believe should, in fact, be the case. For instance, Lal (1983) thinks: "The Utopian theoretical construct of perfect competition then becomes relevant as a reference point by which to judge the health of an economy, as well as the remedies suggested for its amelioration." The proponents of the neo-classical political economy school consider the Pareto-optimal state as a counterfactual, distortion-free state, by reference to which the policy-induced (potential) waste of real resources can be measured. Furthermore, the many generalizations drawn from the development experience of the "Gang of Four" in support of the superiority of the first-best rules also belong to this category of arguments.

I think that this line of thinking is in error--theoretically and empirically.

(i) It is generally forgotten that the existence of Pareto-efficient configuration of product and input prices tells only about rules which "sustain" such a bliss configuration, which is already in place. As Bator (1958) points out, in the bliss situation "we shall be concerned only with the prior problem whether a price-market system which finds itself at the maximum welfare point will or will not remain there". It does not tell us how to get there. Thus, whether or not a policy of "getting the prices right" will in fact land an economy with "wrong" prices in a state of bliss with the "right" prices is not a sure thing--at least, not theoretically.

(ii) Another point is whether obedience to the rules of allocative efficiency must of necessity mean a shift from the government to the free markets run by the invisible hand. The answer is: not necessarily. This is because, again to quote Bator (1958), "the necessary conditions of decentralized price-profit calculations [hold] both in 'laissez faire' and in a socialist setting of Lange-Lerner civil servants". Also, the situations of failure of the basic duality property--i.e., "failure by existence", "failure by signal", "failure of incentive", and "failure by enforcement"--apply as much to "laissez faire markets with genuine profit and satisfaction seekers" as to the "decentralized efficiency of a Lange-Lerner type of [socialist] organization scheme". The only difference is that the conditions relating to a "self-policing" competitive economy, characterized by very many producers in every market, will not be relevant in a socialist (market) economy. Whether the Pareto-optimal solution, which is supposed to maximize social welfare, is attained by the 'invisible hand' or calculated electronically by a computer does not matter at this level of discussion. All that is asserted is that "an appropriate price system is associated with an efficient state", Malinvaud (1969); and that these prices are, strictly speaking, Lagrangean multipliers, which are formally equated with (shadow) profits, wages and rents. However, note that these shadow prices may or may not equal market prices. Indeed, in all cases where external economies obtain, this equality will not hold--not even theoretically.

Thus the case for a unidirectional (quick) march from the government to the market cannot rest only on the basis of the rules of allocative efficiency, which are essentially neutral with respect to the institutional or political nature of the economic set-up postulated.

On First-best Dreams in NICs

South Korea has been cited--e.g., by Little (1982)--as a showcase of a state run according to the (Pareto-optimal) efficiency rules, and as one which because of that has made it to the top. But many competent observers--e.g., Jones and Sakong 0980) and Pack and Westphal (1986)--of the South Korean miracle plead innocence of their Paretian proclivities. Findlay (1988) points out that "the experience of all NICs has been marked not only by a strong reliance on world market forces, but also by very far-reaching and pervasive intervention and control in all segments of the economy". Even the export-bias displayed by South Korea (and also by Brazil) can hardly be cited as a model of liberalism; instead, the practice of dumping through market segmentation--by making the domestic consumer pay significantly more than what the foreign consumer pays for the same product--is more an example of mercantilism than of trade liberalism. As for the argument that this latter-day strong export bias in some sense may have cancelled the early equally strong import-substitution bias, thus creating conditions approaching the first-best state of bliss promised by Pareto-optimality, is really stretching Vilfredo Pareto's meaning a little too far. According to Bardhan (1988), "it is by now well-known that the favourite neo-classical showcase of South Korea is not predominantly one of market liberalism but of aggressive and judiciously selective state intervention. The Korean state has heavily used the illiberal compliance mechanisms of selective command and administrative discretion, restricting imports for industrial promotion, disciplining the private sector through control over domestic credit and foreign exchange and underwriting of foreign borrowing, and public enterprises leading the way in many areas." In other words, South Korea, like other developing countries (e.g., Pakistan, India), has rebelled against the first-best rules of market efficiency (and Pareto-optimality) in every possible way, with state intervention being used in its qualitative as well as quantitative manifestations. Of course, one could still persist with the anti-dirigiste thesis and maintain that "it is not outlandish to believe that South Korea, especially, would have done even better if its government had intervened less ..." [Crook (1989)]. But this is more an assertion of faith, based on a metaphysical belief in the superiority of free markets, rather than a scientific statement.

The falsity of making generalizations about the benefits of privatization from the South Korean case and those of the others in the NICs group--even assuming that this is what the NICs have actually done--becomes obvious by considering the opposite case, which pushed privatization to its limits but has not done very well economically, namely the case of Chile. Yotopoulos (1989) has observed that privatization, especially of capital markets, has not led to higher levels of savings and investment. Indeed, gross fixed capital investment during 1974-82 was only 15 percent of GDP as compared with 21 percent of GDP during the 1960s, when the privatization fever had not yet taken hold of Chile. The so-called democratization of the ownership of national assets--by making ownership more broad-based--also did not come about. Indeed, the concentration of economic power in major banking groups increased. To add insult to injury, industrial efficiency also did not improve. Indeed, industrial profitability declined during the period of the privatization ferment.

THE WORLD THAT NEVER WAS

Do Markets Always Deliver?

As noted above, a regime of competitive efficiency implies a set of shadow prices--which have all the analytical characteristics of profits, wages, and rents. If these prices are set equal to each factor's marginal (revenue) product, then by the Euler's Theorem total output will be exactly apportioned among the factors of production--in cases where the production function is linear and homogeneous. As no (Marxian) surplus output is left over, no exploitation of labour (or capital) is ever possible--mathematically, that is. Also, as labour markets eventually clear, no possibility of (involuntary) unemployment emerges when bliss configurations of input and output (shadow) prices prevail.

Another related train of thought is as follows: the fundamental theorem of welfare economics asserts not only that every perfectly competitive equilibrium is Pareto-optimal, but also that, for some initial distribution of endowments, every Pareto-optimal state is also a perfectly competitive equilibrium. The second part of the theorem is important: does the market ensure that the initial distribution of endowments among the economic agents is equitable? And how does it do it, if at all it does it? The neo-classical answer is that if lump-sum, or some other nondistortionary, transfers could be made by the gainers to the losers from a change, and if still there is some leftover, then the change in question is unequivocally efficient; and, even though it is not always explicitly asserted, such a state is by the same token regarded as equitable.

Another assurance about the alleged equity of the state of bliss comes from economists like Coase (1960), Buchanan and Stubblebine (1962), and others, who show that the market fails to (re)distribute property rights not because of some inherent defect of the market mechanism, but because such rights are not adequately defined. Hence, it is shown that, in a game-theoretic framework, the outcome of a bargaining process about property rights will be Pareto-optimal--and efficient--if property rights are properly defined. The Pareto-optimal situation, according to this line of thinking, will be attained by moving on to a new "contract curve" through bilateral trading between parties.

Let us examine these arguments. First, as to the Hicksian costless lump-sum 'payments' by potential gainers to potential losers from a change, it may be noted that such lump-sum transfers that 'could be made' are never in fact made--in the best tradition of welfare economics! But this is neither here nor there. If the welfare of the losers from the change is to be raised actually, then such transfers should actually be made.

Second, it has been shown--e.g., Bowles (1985)--that market equilibrium is characterized by (involuntary) unemployment because the wage rate actually paid by the employer exceeds the market-clearing wage. This may happen in the rural areas to avoid labour shortages in the peak season [Bardhan (1988)], and in the urban areas to economise on-the-job training costs [Shapiro and Stiglitz 0984)]. Malinvaud (1984) shows that most observed unemployment is of the involuntary (disequilibrium) variety, mainly because of the non-existence of a market-clearing wage rate. He, therefore, concludes that, in general, "'permanent market clearing is an untenable hypothesis". Also, in the Harris-Todaro model, the equality of (expected) wages fails to clear the (urban) labour market, so that equilibrium in a segmented labour market coexists with urban unemployment, which is fed by continuing rural-urban migrations [Khan (1987)].

Third, for market efficiency to lead to an equitable redistribution of the existing legally sanctioned private property rights, the transaction costs are assumed to be zero. But, this assumption will seldom be met; and, with positive transaction costs, the familiar externality scenario--of a divergence between private and social benefits--will arise, which the market will fail to cure. [Furubotn and Pejovich (1972)]. But a more fundamental point is that in the Cease-Buchanan type of solution (1960 and 1962, respectively), it is essential that each individual makes available information about his initial endowments to provide a basis for choice among various Pareto-optimum states--to choose the "optimum optimorum". But such information is hardly ever truthfully revealed, without violating the ground-rules of a decentralized regime of free markets. One of these ground rules is that: the utility (or the production) functions of the individuals are not known to each other; and that there is no mechanism available to the faceless market to make any individual reveal information about his utility (production) functions to another individual. That being the case, Arrow (1979) shows that "a procedure which would achieve a Pareto-efficient allocation if each agent knew the other's utility function will have a positive probability of falling short of efficiency if this knowledge is absent": And since such knowledge is indeed absent, it follows that a Pareto-efficient solution secured through the bargaining process will not necessarily be efficient!

This last result is also relevant for examining the question of private property rights--whether market processes will lead to an equitable distribution of (initial) property rights. Arrow (1979) shows that such an outcome depends crucially on the (unstated) assumption that players in a cooperative game "know every other player's pay-off (utility, profit, whatever) as a function of the strategies played". But, as shown above, the existence of such knowledge runs contrary to the rules of the competitive markets.

Therefore, it follows that the state of bliss of the economics textbook may not even be approximated, let alone actually achieved, in the real world--for more than one reason: that the lump-sum transfers from the gainers to the losers from a change are not actually made that (involuntary) unemployment coexists with market equilibrium; and that an equitable redistribution of property and asset holdings cannot be brought about by the market.

Is Government Born to Fail?

In direct contrast to the assertions of market failure that Pigou and Marshall first made, of late there have been statements about (innate) government failure. The new-fangled "neo-classical political economy" asserts that all-pervading government intervention in economic processes has led to the creation of rent-seeking activity Krueger (1974), and to unproductive profit-seeking (DUP)activities Bhagwati (1984), thereby leading to a wastage of real resources. Brock and Magee (1984) have conjured up an (inefficient) "Invisible Foot" which tramples all over the (efficient) 'Invisible Hand'. This odd creature is another name for government, which is seen as preventing the forces of free competition from maximizing social welfare through the (unproductive) activities of the profit-seekers--the 'profits', i.e., rents, which can be eliminated by restoring the world to the joys of perfect competition.

According to this line of thought, since market processes are assumed to be costless in terms of providing the necessary information for making production decisions, a shift from the market to the government imposes an avoidable deadweight loss on the economy. But this point is not really worth much; because, as North (1984) points out, "there is no meaningful standard of Pareto-efficiency possible, since one cannot specify a least-cost structure of government for any given level of output". Furthermore, such arguments also implicitly equate competitive efficiency (and Pareto-optimality) with efficiency of (private) markets in the real world; but, as I noted above, this position is hard to defend. This is especially so because the information provided by the market is also costly.

Becker (1983) believes that the government probably can help achieve the social optimum but it would not do, because governments are representatives of vested interests, whose interests they serve. But this may be too narrow and cynical a view of governments. For notwithstanding their penchant for underhand politicking, governments do have a "conception of national interest" [Miliband (1983)]. Furthermore, such cynical neo-classical views of state do not explain how different interventionist states over time become development states in some cases though not in others. The explanation seems to lie in the ability of the development states to insulate economic management from wasteful rent-seeking activities. Even in cases where such insulation is not achieved, it will be naive to suggest that leaving it to the market will solve all problems; because, as Bardhan (1988) notes, "the very reasons why insulation is infeasible are often also the ones which will make first-best policies inoperative, and in the absence of lump-sum redistribution, a policy of relative inaction may be distributionally unacceptable".

The fact of the matter is that inherent (generalized) government failure cannot be asserted in the same sense as the Marshal-Pigou type of(selective)market failure, partly because government policies, for better or worse, set the parameters within which the market functions--efficiently, perhaps. There is always room for improvement here, and all such avenues should be exhausted. But if government will always fail to be efficient, no matter what, then there is no guarantee that the markets will also not fail, no matter what. The utility-maximizing calculus, based only on the consumption of the (private) goods owned by the individual, cannot take him too far without the availability of public goods, and without the constitutional guarantees of the free consumption of the private goods, both of which are provided by the state. At any rate, the process of growth, and the attendant (painful) structural adjustment required for this process to continue unhindered, cannot be propelled by the utility/profit-maximizing economic agents alone. A sovereign state must initiate "policy action and institutions are required [to minimize] the costs of, and resistance to, the structural shifts implicit in, and required for, a high rate of growth". [Kuznets (1971)].

THE PASSION CALLED PARETO-OPTIMALITY

As noted above, in the anti-dirigiste attacks made by the agnostics on development economics, a reference is always made to Pareto-optimality, which says that as between two alternative social states x and y, if at least one person strictly prefers x to y, and for every individual x is at least as good as y, then x is Pareto-superior to y, and the society in such a case should also prefer x to y. (In its weaker form, the Pareto-optimal rule says that if every individual strictly prefers x to y, then the society should also prefer x toy.) In simpler words, a social state qualifies as Pareto-optimal if in case of a departure from such a state not everyone can be made better off, so that the utility of an individual (say, A) cannot be raised without lowering the utility of some other individual (say, B).

In recommending free markets for developing economies--and development economics--to achieve efficiency (and growth), the neo-classical advice is for the rational individuals and for the society to go where the Pareto-optimality beckons them to go. Observance of this rule is 'desirable' because it is allegedly 'fair': since it reflects unanimity--i.e., the rule requires that preferences held universally in a society should be reflected in any scheme of social judgement. It is also liberal, because it preserves liberty, which is the basic value to be cherished by all civilized societies. Thus the passion for Pareto-optimality is not only rational, but it is also the only civilized thing to do. Let us examine these claims a little more carefully.

The Chinks in the Paretian Armour

Does Pareto-optimality preserve individual liberty? Believe it or not, the answer is firmly in the negative! Sen (1970)shows the incompatibility of the Pareto principle, both in its strong and weak forms, with even a rudimentary kind of individual liberty. Assuming an "unrestricted domain" (U), Pareto-optimality (P), and "Liberalism" (L), Sen demonstrates that "there is no social decision function that can simultaneously satisfy U, P, and L". Here liberalism--or more accurately, libertarianism--is defined in a very elementary sense of recognizing each individual's privilege to have a minimum of what Hayek (1960) calls every individual's "protected sphere". In other words, each individual should have the freedom to make "at least one social choice, for example, having his own walls painted pink rather than white, other things remaining the same for him and the rest of the society". The implication of this Impossibility Theorem, which modifies the celebrated Arrow's Impossibility Theorem, is quite disturbing: it shows that Pareto-optimality, given the assumption U--i.e., every logically possible set of individual ordering is included in the domain of the collective choice rule--cannot be combined with even a minimum dose of liberalism. Hence, if Pareto-optimality is followed to its logical (bitter) end, then "society cannot let more than one individual be free to read what they like, sleep the way they prefer, dress as they care to, etc., irrespective of the preferences of others in the community". Surely, these consequences are "most illiberal". Sen warns: "if someone takes the Pareto principle seriously, as economists seem to do, then he has to face the problem of consistency in cherishing liberal values, even very mild ones." It may seem, therefore, that even to be able to breathe freely one may have no option but to free himself from the smothering embrace of Pareto-optimality!

Is the Pareto-optimality rule 'fair'? Alas, the answer is again in the negative; and for the following reasons:

(a) The Pareto-optimality rule is distributionally neutral. Before going any further, let us understand clearly the meaning of "neutrality". Suppose there are two states x and y. The neutrality property demands that if (x, y) is replaced by (a, b) in everyone's preference ordering, then we must do the same in social ordering as well. In other words, for making a social choice it does not matter what the nature of x, y, a, b is; all that matters is the existence of individual preferences over these states. The problem is posed in the following manner: Suppose x = equal division of cake, and y = nothing for person 1 and equal division for persons 2 and 3. According to the neutrality property, it does not matter if x, y are replaced by a, b. Let a = nothing for 2 and 3 and all for 1; and b =equal division. Neutrality demands that x is socially preferred toy, if and only if a is socially preferred to b. In other words, from the society's point of view, the equal and the extremely unequal outcomes are equally preferable! As Sen (1985) observes, "once we have got to neutrality in this format, there is no real chance any more of making judgements concerning income distribution". Thus a society rolling in the luxury of a Pareto-optimal state may still be perfectly abominable.

(b) The Pareto-optimal rule is blind to whether a person is rich or poor. This follows from the utilitarian (indeed, welfarist) character of Pareto-optimality. As is well-known, Pareto-optimality--indeed, (new) Welfare Economics--regards social welfare as an increasing function of personal utility levels alone, denying admission rights once and for all to any type of non-utility information; that individual utilities are non-comparable; and that individual utilities are ranked ordinally, not cardinally. But these very characteristics incapacitate Pareto-optimality to differentiate the rich from the poor, even in broad daylight. To see this clearly, I can do no better than quote from Sen (1979).

"Can we identify the rich through the observation that they have more utility than the poor? Not in the Arrow framework, since interpersonal comparisons are not admitted. Perhaps as those with a lower marginal utility of income? No, of course not, since that will go against both non-comparability and ordinalism. Can we then distinguish the rich as those who happen to have more income, or more consumer goods (nothing about utility need be said), and bring this recognition to bear in social judgements? No, not that either, since this will go against welfarism (and against strict ranking-welfarism), since this discrimination has to be based on nonutility information."

It follows from (a), Co), and (c)above that the Pareto-optimality rule excludes economists, and especially development economists, from paying any attention whatsoever to problems of liberty, income equality, and poverty--perhaps to let them specialize completely on efficiency! But if, according to Arthur Lewis (1984), development economics is a study of the economic structure and behaviour of poor countries, then to follow Pareto-optimality alone in making collective choice is indeed a prescription for social disaster.

ENTER

THE PRINCE!

Another characteristic of Pareto-optimality is its alleged "value-free" character, in line with the stand taken by (neo-classical) economists against relying on normative judgements for making social decisions. [Robbins (1932)]. But is this so? Insofar as Pareto-optimality relies on unanimity about certain value judgements--e.g., not disturbing the status quo with respect to income distribution--this state of affairs is by no means without a value judgement: it is certainly not value-free. Be that as it may; the fact is that such a stance disqualifies Pareto-optimality as the bedrock of development economics, which aims explicitly at socio-economic change--indeed, a structural transformation of poor societies into well-to-do societies.

How Equal is Equality?

In the poor countries a central value judgement, indeed an ethical judgement, must be made: it is to push developing economies towards greater equality of distribution of income and wealth between different classes of the society--especially between the rich and the poor. An explicit commitment to some such ideal is essential to ensure a (voluntary) universal participation by all classes of society in the process of economic development, so that both the costs and the benefits of social change are equitably shared. If Pareto-optimality cannot see the difference between the rich and the poor, so much the worse for it. We then need some other social choice rule that does 'see' this vital difference.

Let it be noted that equality between the rich and the poor does not mean complete equality. Indeed, no one, including Karl Marx, has ever meant this. All that equality is meant to imply is that economic processes be directed to minimize, not necessarily to eliminate, the inter-class distributional differences as far as it is economically and socially permissible. A relevant consideration in this context is the Rawlsian maxim: "inequalities are arbitrary unless it is reasonable to expect that they will work out for everyone's advantage, and provided only the position and offices to which they attach, or from which they may be gained, are open to all." [Rawls, (1971)]. According to this rule, the creation and the existence of income inequalities may be ethically justifiable provided only that the basic institutions of the society are also restructured in such a manner that the new economic and social possibilities opened up by economic progress are equally accessible to all classes of the society.

The goal of equality, in fact, has been pursued by economists explicitly since the days of Jeremy Bentham. But equality itself has been defined differently by various schools of thought. Sen (1983) distinguishes four main types of equality: there is the "utilitarian equality", the "total equality", the "Rawlsian equality", and the "basic capability equality". Of these, the first two types of equalities are strictly for the birds. Utilitarian equality requires the equality of marginal utility of everyone. But to restrict the equalization process to the utility space alone is really to prevent it from doing anything worthwhile. As noted above, if looking at different persons' marginal utility alone does not let the observer see the difference between the rich and the poor--mainly because of the exclusion of interpersonal comparisons--then not much social change can come about by following this principle. In particular, once we come to the distribution of utilities, utilitarianism offers no comfort to the poor; for even the smallest gain in total utility sum would overbalance the worst type of distributional inequality on this scale. This problem can be avoided if it is assumed that everyone has the same utility function. But that is really to trivialize the problem of inequality--which is marked by the fact that different persons' utility functions are not the same. Equalization of total utility is a more helpful guide--particularly, its leximin version, according to which the goodness of a state is judged by the utility level of the worst-off individual. But it, too, is unsatisfactory because it, by definition, ignores the intensity of the person's needs; and it is also insensitive to the magnitude of the potential utility gains and losses.

The main problem with both these types of equality is their insistence on using utility information only as an index of individual and social welfare. This exclusive insistence on utility information, marginal or total, can lead to a situation in which more income is given to the less needy, simply because that is the hard-to-please type; and the poor person, who is easily satisfied even with small mercies, will have less income! But the most interesting information about a person's welfare is of the non-utility type--e.g., the possession of certain types of goods, or the possession of certain capabilities to do some basic things essential for man's survival. Accordingly, I shall now discuss the remaining two types of equality which explicitly use such non-utility information: (i) The Rawlsian equality; and (ii) Sen's basic capability equality.

(i) Rawlsian Equality: The central thrust of the Rawlsian conception of equality is its focus on bringing about institutional changes such as to "make the worst-off best-off'--that is, such action as would raise the welfare level of the worst-off individual in the society as far as it is possible to do so. [Rawls (1971)]. This is the so-called Rawlsian Difference Principle, Rejecting utility as the basis of individual welfare, Rawls, instead, defines welfare in terms of a bundle of "primary goods", which are defined as "things that every rational man is presumed to want". These things include "rights, liberties and opportunities, income and wealth, and the social bases of self-respect". Institutional arrangements which guarantee the access of the worst-off individuals to these primary goods are both efficient and equal. The Difference Principle is held to be "just" because it is chosen "fairly" in the "original position", which denotes (a mental) experiment of passing through a "veil of ignorance" to make impartial decisions about the structure of the society. Unlike the utilitarian principle, Rawls allows interpersonal comparisons to judge the fairness of the distribution of primary goods among individuals.

There are many problems with this principle of equality also, especially because the needs (for primary goods) of a disadvantaged person--say a cripple or a sick person--do not get registered at all in the Rawlsian calculus; and there are other technical points that need not be recounted here. But the emphasis of the Rawlsian equality on institutional change, on the welfare of the least-privileged, on "justice as fairness", and on the availability of "primary goods" to all without discrimination of any kind, are elements which should find an explicit expression in any sensible model of development economics.

(ii) Basic Capability Equality. Sen (1984) goes a step--beyond the Rawlsian emphasis on equality with respect to primary goods--towards "what goods do to human beings". Equality is insisted on with respect to such capabilities. Shifting attention from goods to capabilities has the advantage of taking into account explicitly the differences in people's "needs" and requirements--something that the Rawlsian and the utilitarian concepts of equality fail to do. Such differences, arising from the conversion of goods into capabilities, are allowed in Sen's concept of equality, as also the differences arising from the nature of different societies. "The notion of the equality of basic capability is a very general one, but any applications of it must be culture-dependent, especially in the weighting of different capabilities."

FROM "NOT TO BE" TO "TO BE"

The observations made so far may be summarized as follows. The claims made about the demise of development economics are exaggerated, to say the least. The basis for these claims--namely, an unambiguous superiority of the market over the government due to the latter's pervasive failure--is ill-founded in logic because, as Pack and Westphal (1986) point out, "the factors responsible for a government's inability to intervene effectively may also preclude its following the neoclassical prescription". Such claims are also incorrect factually, because the NICs are no laissez-faire paradises. They are not what they are because of following the neo-classical first-best road to economic salvation. Indeed, they are what they are because they used, at the right time and with speed, both the government and the market to maximize social welfare--as the government saw it. The a priori reasoning employed to condemn all government intervention as inefficient--that it entails wastage of real resources when measured with reference to an imaginary economy run according to the first-best rules of Pareto-optimality--is, by and large, empty of content. This reasoning is also at fault because Pareto-optimality, which is seen as an alter-ego of competitive efficiency, is not always efficient. And it is certainly not equitable. Thus a scenario of generalized market success is sheer neoclassical romanticism.

A guide-book up the road to economic progress, which contains directions only about efficiency and nothing of substance about equity, cannot be recommended for the development economist. Once the development process is looked at, a la Rawls, as one that requires a rearrangement of the basic structure of the society which is "fair" and "just"--that which focuses on the needs of the least-privileged in the society--the standard prescription of "getting the prices right" and then letting nature (market) take its course will not take us anywhere. Indeed, it may lead to social anarchy. An exclusive reliance on the market mechanism cannot bring about a structural transformation--especially in matters of redistributing private property rights, for the simple reason that it can neither initiate the growth process nor can it by itself adjust to a clash of vested interests to create the space, so to speak, for the post-transformation world. For this a development-oriented state is required. To ensure growth with equity, development economics should continue with its perception of a "mixed economy", whereby both the government and the market are needed to help initiate and sustain the development process. It follows that the recommendation to privatize for the sake of privatization is ill-conceived.

The brief survey of the various concepts of equality presented in the last section underscores four basic points, which are of fundamental importance for development economics. First, normative judgements are routinely made when it comes to the problems of personal or social welfare. Second, these normative judgements are based on an ethical perception of human beings, who are not only free but are also seen as free--in being equally entitled to the most extensive basic amenities and liberties that a society has to offer. Third, there is the emphasis upon changing the status quo, on doing positive things to change the basic structure of the society, and on bringing joy to the withered lives of the countless millions by producing greater equality among different human beings. Fourth, the emphasis is on justice and fairness as well as efficiency. It is not one thing to the exclusion of the other. All these points should be taken care of in our perception of the process of economic development, which should aim both at growth and equal distribution of the fruits of economic progress--things that a blind adherence to the Pareto-optimal rule would not enable us to see, much less do. Indeed, to be logically consistent and relevant, development economics must leave the amoral world of Pareto-optimality--and that of utilitarianism--for one which is ethically motivated.

The challenge facing development economics is not one of devising docking strategies to join the mothership--i.e., the neo-classical economics; it is also not a matter of declaring development economics to be 'independent' of neo-classical economics. What is required is to evolve a synthetic view of economic processes requiring both macro and micro insights, one which is also based on a social choice theory that is not restricted to Pareto-optimality. Social action should be made sensitive to the intensity of individual preferences, and not just to individual preferences; interpersonal comparisons of the welfare should be allowed to comprehend the differences in the social stations of different individuals; and due account should be taken of the nature of the societal structure with reference to the social choices made.

For once, the development economist cannot agree with his great-grandfather, Alfred Marshall, who condescendingly allowed the study of the ordinary business of life in the East to be made with the help of a limited number of concepts. We now know that development economics is a many-splendoured thing, involving above all a grand synthesis of economics and ethics. It requires us to do something as sensible and all-embracing as watching Hamlet, but not without the Prince.

REFERENCES

Arrow, Kenneth, J. (1979) The Property Rights Doctrine and Demand Revelation under Incomplete Information. Ch. 2 in Michael J. Boskin (ed)Economics and Human Welfare. New York: Academic Press Inc. 23-39.

Bardhan, Pranab (1988) Alternative Approaches to Development Economics. Ch. 3 in Hollis Chenery and T. N. Srinivasan (eds)Handbook of Development Economics 1. Oxford: North-Holland 39-71.

Bator, Francis M. (1958) The Anatomy of Market Failure. The Quarterly Journal of Economics 72 : 3 351-379.

Becker, Gary S. (1983) A Theory of Competition among Pressure Groups for Political Influence. The Quarterly Journal of Economics 97 : 3 371-400.

Bhagwati, J. N., R. A. Brecher, and T. N. Srinivasan (1984) DUP Activities and Economic Theory. In David C. Collander (ed) Neoclassical Political Economy. Cambridge: Ballinger Publishing Co. 17-32.

Bowles, S. (1985) The Production Process in a Competitive Economy: Walrasian, Neo-Hobbesian, and Marxian Models. American Economic' Review. 71 : 1 16-36.

Brock, William, and Stephen P. Magee (1984) The Invisible Foot and the Waste of Nations, Redistribution and Economic Growth. Ch. 12 in David C. Collander (ed) Neoclassical Political Economy. Cambridge: Ballinger Publishing Co. 177-186.

Buchanan, James, and W. C. Stubblebine (1962) Externality. Economica 29 371-384.

Coase, R. H. (1960) The Problem of Social Cost. Journal of Law and Economics 3 1-44.

Crook, Clive (1989) The Third World: Trial and Error; Poor Man's Burden. The Economist 23-29 September.

Findlay, Ronald (1988) Trade, Development and the State. In G. Ranis and Paul A. Schultz (eds) The State of Development Economics. New York: Basil Blackwell. 78-95.

Furubotn, Eirik, and Svetozar Pejovich (1972) Property Rights and Economic Theory: A Survey of Recent Literature. The Journal of Economic Literature 10 : 4 1137-1162.

Haberler, Gottfried (1988) International Trade and Economic Development (with a new Introduction). San Francisco: International Center for Economic Growth.

Hayek, F. A. (1960) The Constitution of Liberty. London: Routledge and Kegan Paul.

Jones, L. P., and I. Sakong (1980) Government, Business and Entrepreneurship in Economic Development: The Korea Case. Cambridge, M A: Harvard University Press.

Khan, Ali (1987) Harris-Todaro Model. The New Palgrave: A Dictionary of Economics (2 : E to J). London: Macmillan Press Ltd. 592-594.

Krueger, A. (1974) The Political Economy of the Rent-seeking Society. American Economic Review 64 : 3 291-303.

Kuznets, S. (1971) Economic Growth of Nations: Total Output and Production Structure. Cambridge, M A: Harvard University Press.

Lal, Deepak (1983) The Poverty of Development Economics. Great Britain: The Institute of Economic Affairs.

Lewis, W. Arthur (1984) The State of Development Theory. American Economic Review 74 : 1 1-10.

Little, I. M. E. (1982) Economic Development: Theory, Policy and International Relations. New York: Basic Books.

Malinvaud, Edmond (1969) Capital Accumulation and Efficient Allocation of Resources. (Econometrica, 1953). Reprinted in Kenneth J. Arrow and T. Scitovsky (eds) Readings in Welfare Economics. London: George Allen and Unwin 645-681.

Malinvaud, Edmond (1984)Mass Unemployment. New York: Basil Blackwell.

Miliband, R. (1983) State Power and Class Interests. New Left Review 138.

Naqvi, Syed Nawab Haider (1986) A Tale of Two Hands. The Pakistan Development Review (Papers and Proceedings) 25 : 3 225-245.

Naqvi, Syed Nawab Haider (1989) For Morality's Sake! The Pakistan Development Review (Papers and Proceedings) 27 : 4 357-376.

North, D. (1984) Three Approaches to the Study of Institutions. In David C. Collander (ed) Neoclassical Political Economy. Cambridge: Ballinger Publishing Co. 33-40.

Pack, H., and L. E. Westphal (1986) Industrial Strategy and Technological Change: Theory versus Reality. Journal of Development Economics 22 : 1 87-128.

Rawls, J. (1971) A Theory of Justice. Cambridge, M A: Harvard University Press. Oxford: Clarendon Press.

Robbins, L. (1932) An Essay on the Nature and Significance of Economic Science. London: Macmillan.

Sen, A. K. (1970) The Impossibility of the Paretian Liberal. Journal of Political Economy 78: 152-157. Reprinted in A. K. Sen (1983) Choice, Welfare and Measurement. Oxford: Basil Blackwell 285-290.

Sen, A. K. (1979) Personal Utility and Public Judgement. Economic Journal 89 : 355 537-558.

Sen, A. K. (1983) Equality of What? Ch. 16 in A. K. Sen Choice, Welfare and Measurement. Oxford: Basil Blackwell 353-369.

Sen, A. K. (1984) Resources, Values and Development. Oxford : Basil Blackwell.

Sen, A. K. (1985) Review of [Kenneth J.] Arrow's Social Choice and Justice. Journal of Economic Literature 23 : 4 1764-1776.

Shapiro, C., and J. E. Stiglitz (1984) Equilibrium Unemployment as a Worker Discipline Device. American Economic Review 74 : 3 433-444.

Yotopoulos, Pan A. (1989) The (Rip) Tide of Privatization: Lessons from Chile. World Development 17:5 683-702.

SYED NAWAB HAIDER NAQVI, Professor Naqvi is President of the Pakistan Society of Development Economists and Director, Pakistan Institute of Development Economics, Islamabad.

This is a revised version of the Presidential Address which Professor Naqvi delivered at the Sixth Annual General Meeting of the Pakistan Society of Development Economists. Thanks are due to Prof. Alamgir Hashmi for making useful editorial changes, and to Mr M. Aslam for typing the manuscript of the Address.