You are on page 1of 22

01 CJE 26/5 pp.

539-559 FINAL

23/7/02 4:10 pm

Page 539

Cambridge Journal of Economics 2002, 26, 539–559

Breaking the mould: an institutionalist
political economy alternative to the
neo-liberal theory of the market and
the state
Ha-Joon Chang*
The paper criticises the currently dominant neo-liberal discourse on the role of the
state and proposes an alternative approach that will allow us to overcome its
shortcomings, especially its inadequate analyses of the role of institutions and
politics. It argues that the central problem with the neo-liberal framework lies not in
its excessively anti-interventionist policy conclusions, as some of its critics believe,
but in the very ways it envisages the modus operandi of the market, the state,
institutions and their interrelationships. The paper then discusses how we may
construct the alternative approach of ‘institutionalist political economy’.
Key words: State, Market, Political economy, Neo-liberalism, Institutionalism
JEL classifications: H0, P1

1. Introduction
This paper critically examines the neo-liberal discourse that currently dominates the
debate on the role of the state and suggests an alternative theoretical framework to
overcome its limitations. After tracing the evolution of the debate on the role of the state
during the post-World War II period that has led to the current dominance of neoliberalism (Section 2), I question some fundamental assumptions underlying the neoliberal discourse on the role of the state and point out the theoretical and practical
problems that arise from these assumptions (Section 3). I argue that, if we are to
overcome these problems, a marginal tinkering with the neo-liberal framework is not
enough and that we need to develop an altogether different framework, which I propose to
call institutionalist political economy. In Section 4, I outline this alternative framework and
show how its adoption will improve our understanding of the role of the state. A short
section of concluding remarks follows (Section 5).
Manuscript received 8 December 1997; final version received 4 October 2000.
Address for correspondence: Faculty of Economics and Politics, University of Cambridge, Sidgwick Avenue,
Cambridge CB3 9DD, UK; email:
*Faculty of Economics and Politics, University of Cambridge. I thank Peter Evans and Bob Rowthorn for the
discussions that they provided in the process of writing the paper. I also benefited from the comments on earlier
drafts by Shailaja Fennell, Jayati Ghosh, Jonathan di John, Grazia Ietto-Gillies, Joseph Lim, James Putzel, Shara
Razavi and Alfredo Saad Filho. I also thank one of the anonymous referees for helpful suggestions.

© Cambridge Political Economy Society 2002

01 CJE 26/5 pp. 539-559 FINAL


23/7/02 4:10 pm

Page 540

H.-J. Chang

2. The evolution of the debate: from ‘Golden Age Economics’ to neoliberalism
The end of World War II witnessed the world-wide rejection of the laissez faire doctrine
which had failed so spectacularly during the interwar period. During the following quarter
century or so, which is commonly known as the Golden Age of Capitalism, a variety of
interventionist economic theories such as welfare economics, Keynesianism and the early
‘development economics’ set the agenda for the debate on the role of the state (Chang
and Rowthorn, 1995A; also see Deane, 1989). These interventionist theories, which I
collectively call ‘Golden Age Economics’ (GAE), identified a horde of ‘market failures’
and argued that active state involvement was necessary in order to correct these failures.
Although the exact types and forms of policies recommended by different branches of
GAE were different from each other, it was widely agree that a ‘mixed economy’ of one
sort or another was necessary and desirable.
From the 1970s, however, following the economic and political changes that the GAE
had brought about both nationally and internationally, there were marked changes in the
terms of debate on the role of the state (on the rise and the decline of the Golden Age, see
Marglin and Schor, 1990). The new terms of debate was set by neo-liberal economists
such as Milton Friedman, Friedrich von Hayek, George Stigler, James Buchanan,
Gordon Tullock, Anne Krueger, Ian Little and Alan Peacock (for critical reviews, see
Mueller, 1979; Cullis and Jones, 1987; Chang, 1994A; Stretton and Orchard, 1994).
Neo-liberalism emerged out of an ‘unholy alliance’ between neoclassical economics,
which provided most of the analytical tools, and what we may call the Austrian-libertarian
tradition, which provided the political and moral philosophy.1 The central plank of its
argument regarding state intervention is that we cannot assume the state to be an
impartial and omnipotent social guardian, as is assumed in GAE. Instead, it is argued, we
should see the state as an organisation run by self-seeking politicians and bureaucrats,
who are not only limited in their ability to collect information and execute policies but are
also under pressures from interest groups. Neo-liberal economists argue that this
imperfect nature of the state results in ‘government failures’ in the forms of regulatory
capture, rent-seeking, corruption, and so on. They argue that the costs from these
government failures are typically greater than the costs of market failures, and therefore
that it is usually better for the state not to try to correct market failures, because it may
make the outcome even worse.
This attack was partly unfair, because many practitioners of GAE did not actually
believe a real-life state to be the modern equivalent of Plato’s Philosopher King, but used
such a state only as an ideal benchmark (Toye, 1991). However, it is also true that most of
them did not have a clear theory of the state, and therefore made themselves vulnerable to
the attack that their view of the state was ‘unrealistic’ and ‘naïve’.2
I say an ‘unholy alliance’, because the gap between these two intellectual traditions is not a minor one, as
those who are familiar with, for example, Hayek’s scathing criticism of neoclassical economics would know
(e.g., see essays in Hayek, 1949).
Interestingly, around the same time, a very similar criticism was made by many Marxists, who emphasised
the ‘class’ nature of the state. They argued that, thanks to their control over state revenue, political funding
and ideological machinery, the economically dominant class in a society (the capitalist class in the capitalist
society) is able to determine state policies in its favour, subject to the need to maintain some degree of
legitimacy among the dominated classes (see Jessop, 1982, for a review of Marxist theories of the state of the

ironically meant that. The same ‘non-committal’ nature of welfare economics regarding the appropriate role of the state. but policy-makers may justify their lax anti-trust policy in terms of some other logic which has no place in neoclassical economics—say. Therefore. to how high a standing they had in the Western academic hierarchy. political and institutional factors that define that particular market (see Section 3. the logic of market failure has been used in order to justify anything from the minimal state to full-blown socialist planning (Pagano. Accepting the analytical tools of neoclassical economics. and the provision of some large-scale physical infrastructure—and therefore that only a ‘minimal state’ is necessary. ways had to be found to ensure that any endorsement of state intervention had to be kept within a boundary acceptable to the neo-liberal political agenda. it could be absorbed by neo-liberalism. in fact. . which it had not been before World War II. All that welfare economics says. once the political consensus that lay behind various models of mixed economy that emerged during the GAE was undermined.2). Therefore. One such way is to argue that market failures. However. one can draw any conclusion one likes on the appropriate boundary between the market and the state. but whether a particular real-world market actually fails depends on the technological. meant that the neo-liberals had somehow to tame the logic of market failure that had by then become a central element in neoclassical economics. which was basically determined by how good they were in handling the concepts and the tools of neoclassical economics. This was because. 539-559 FINAL 23/7/02 4:10 pm Page 541 Breaking the mould 541 Once this attack was unleashed. however. however. it was according. What captured people’s imagination in those days was the Austrian-libertarian language of freedom and entrepreneurship. largely. In other words. 1985). Indeed. institutions and politics. Given that the Austrian-libertarian tradition had been on the margin of intellectual respectability until the 1970s. depending on various assumptions one makes about human motivation and psychology. for example. in return for which the Austrian-libertarian tradition supplied the popular appeal that neoclassical economics could never dream of supplying itself (whoever died in the name of Pareto Optimality or General Equilibrium?). does not actually have a determined position on the issue. unlike Keynesianism. which provided most of the tools used in drawing such a boundary at the time. and not the arid neoclassical language of Pareto Optimality and General Equilibrium.01 CJE 26/5 pp. albeit with some difficulty (see below). by citing the need ‘not to discourage entrepreneurship’. welfare economics. is that markets can fail. but later downplay the relevance of such models on the ground that real-life states cannot possibly be entrusted with such policies that are technically difficult (due to 1 The point is also poignantly illustrated by the experiences during the early days of ‘reform’ in the former communist countries. technology. the neoliberals could not afford to do without the ‘scientific’ respectability that neoclassical economics carried. neoclassical economists in universities may be doing research justifying stringent anti-trust policy. contrary to what many people had believed. So. it became impossible to defend them using the tools of welfare economics.1 The third way of taming the logic of market failure is to accept it fully and build models that may have strong interventionist policy conclusions. while logically possible anywhere. law and order. when the post-communist governments in these countries chose their foreign economic advisers. in reality exist only in a few limited areas—such as defence. The second way is to limit the spillage of the logic of market failure into policy actions by separating ‘serious’ academic discourse from ‘popular’ policy discourse. what had been regarded as a pretty robust theoretical consensus on the appropriate boundary between the market and the state proved fragile.

whereas many Third World capitalists (and indeed the capitalists in the now-advanced countries up until the early twentieth century) regard it as just that. an economist may build models that recommends state intervention as far as they are ‘technically competent’. As I shall argue in the next section.1 Defining the free market (and state intervention) The neo-liberal discourse on the role of the state. considered an ‘intervention’ in one society but not in another (and has been in the same society at different points of time). 3. in this neo-liberal age. Why is this? Let me answer this question with some examples. neo-liberal discourse on the role of the state contains some serious internal tensions and therefore can be sustained only through some intellectual contortions and political compromise. 539-559 FINAL 542 23/7/02 4:10 pm Page 542 H.-J. asked why they had written the book in the first place if they were going to say in the end that the interventionist policies that follow from their models cannot be recommended because of the political dangers that they carry (see Lucas. I discuss the limitations of the neo-liberal analysis of the role of the state by questioning four aspects of the neo-liberal doctrine that are considered so basic that they are rarely discussed. first of all. and. 3. As can be seen in the enthusiasm that the more interventionist conclusions of new growth theory or strategic trade theory have generated among some critics of neo-liberalism. and indeed the welfare economics discourse that it dethroned. is a highly complicated exercise. a few paragraphs of ‘pop political economy’ analysis dismissing the integrity and the ability of the state are deployed to discredit his own elaborate strategic trade theory models endorsing state intervention that make up the bulk of the article. what I propose to call ‘institutionalist political economy’. This example shows that. I argue that the mode of neo-liberal discourse itself is problematic. Through this discussion. First. the way it theorises about their interrelationships. is about whether state intervention can improve upon the workings of the free market. Even many of those who do not agree with the conclusions that are drawn from this discourse seem to regard the mode of discourse itself as unproblematic.01 CJE 26/5 pp. but requires an approach that takes the role of institutions and politics seriously. The limits of the neo-liberal analysis on the role of the state In this section. I show why overcoming these limitations cannot be done by marginal tinkering with the neo-liberal framework. the more serious problems of the neo-liberal discourse on the role of the state are to do with. despite its pretence of intellectual coherence and clear-cut messages. the state and institutions. but he/she has to prove his/her political credentials by rubbishing his/her own models on political grounds. the same state action can be. and has been. This is because in the advanced countries the right of children not to 1 Works of the American trade economist Paul Krugman provide some of the best examples. In many of his articles. secondly. Robert Lucas. Few people in the advanced countries at present would consider the ban on child labour as a state ‘intervention’ artificially restricting entry into the labour market. 1990). since defining the free market.1 These examples show that. reviewing Krugman’s book with Helpmann. But that is probably the least of its problems. . and therefore defining what counts as state intervention. let us take the case of child labour. As will become clearer in what follows. A leading neo-liberal economist. if he/she is to remain in the mainstream. these critics believe that the limitations of neo-liberalism can be overcome by building more models that justify state intervention. Chang informational demands) and politically dangerous (due to bureaucratic abuse and/or interest group capture). namely. However. the very way it conceptualises the market.

the nineteenth-century US). there are few people in these countries who would say that their country’s automobile market is not a ‘free’ market simply because of these regulations. In contrast. . for example. beyond what are dictated by what these economists themselves regard as ‘basic human rights’. although even then many people found the attempt distasteful. but once such a right is accepted as one of the fundamental rights of all members of the society. do not regard the heavy restrictions on immigration that exist in these countries as a state intervention (and indeed many of them will support stringent immigration control). while rejecting the right of the same citizens to contest the rights of the employers to offer the wages and working conditions they see fit. types of automobile).. as their citizens now regard the right to a clean environment as having priority over the right to choose technologies involved in production and consumption (e. the same state action could be considered an ‘intervention’ in one society and not in another.2 This is most clearly revealed in the current 1 This is also manifested in the existence of many institutions that support this particular hierarchy of rights (e. factory pollution standards. However.. depending on which rights and obligations are regarded as legitimate and what kind of hierarchy between these rights and obligations is (explicitly and implicitly) accepted by the members of that society. However. during the late twentieth century when slavery had become a distant memory and therefore less politically sensitive. in these countries. some American economic historians started a debate on the ‘efficiency’ of slavery. many environmental regulations were widely criticised as unwarranted intrusion on business and personal freedom (e. child benefits). whose impact on economic efficiency is still a legitimate subject of policy debate. Therefore.g. such a right for children is not totally accepted. some developing country exporters who do not accept the hierarchy of rights underlying such regulations as legitimate may consider them as ‘invisible trade barriers’ that ‘distort’ the workings of the ‘free’ market. This contradictory attitude is possible only because these economists (at least implicitly) accept the right of the existing citizens of a country to dictate the terms of non-citizens’ participation in ‘their’ labour market. In contrast.. the ban on child labour or on slavery in the advanced countries of today). The same argument can be applied to the case of slavery.g. To take another example.g. automobile emission standards) when they were first introduced in the advanced countries not so long ago. and therefore the state ban on child labour is considered an ‘intervention’. in these countries such regulations are these days rarely regarded as ‘interventions’. In societies where the right to selfownership is not universally accepted (say. an attempt by the state to ban slavery can be disputed as an efficiency-reducing intervention.01 CJE 26/5 pp. 539-559 FINAL 23/7/02 4:10 pm Page 543 Breaking the mould 543 toil is more or less universally regarded as having precedent over the right of the producers to employ whoever they find most profitable. in the developing countries (of today and yesterday).. many neo-liberal economists. And once a state action stops being considered an ‘intervention’ in a particular society at a given time (e. The examples can go on. but the point is that. the ban will not be considered an ‘intervention’ any more. debating their ‘efficiency’ becomes politically unacceptable—although there is no Godgiven reason why this should be the case. immigration control sets up an ‘artificial’ entry barrier into the labour market as much as do the other labour market ‘interventions’ they criticise. 2 Indeed.1 As a result. the ban on child labour is no longer even a legitimate subject of policy debate any more. who criticise minimum wages and ‘excessively’ high labour standards in the advanced countries as unwarranted state interventions that set up artificial entry barriers into the labour market. universal education. To take yet another example. production technology.g.

therefore. 1993. when there exist externalities). Chang disputes regarding the attempts to incorporate labour and environmental standards in the World Trade Organisation (WTO) negotiation agenda. Let us illustrate this point with some examples. we need to take a position on the legitimacy of the underlying rights–obligations structure for the participants in the relevant market (and indeed certain non-participants. as the whole point of the neo-liberal alliance was to combine the political and moral appeals of the Austrianlibertarian tradition with the ‘scientific’ respectability of neoclassical economics (see Section 2). if we want to decide whether a particular market is ‘free’ or not. 3. which denies the very notion of perfection competition. For example. in neoclassical economics. our discussion on the role of the state will be conducted with the pretence that our own opinions are based on ‘objective’ analyses while those of our opponents are not and therefore are largely ‘politically motivated’. as all markets have some state regulations about who can participate in which markets and on what terms (see Sections 3. see Trebilcock.-J. 1982). However. the ideal market is equated with the ‘perfectly competitive market’ of neoclassical economics. Many people think that one of the biggest ‘failures’ of the market is its tendency to generate an unacceptable level of income inequality (whatever the criteria for acceptability may be). to borrow Hirschman’s phrase. . could have been the neo-liberal theory of the market.1. 2 Given its intellectual composition. ‘market failure’. Thus. or even create) can be so totally accepted (by those who are making the observation as well as by the participants in the market) that some markets appear to have no ‘intervention’ at all and therefore to be ‘free’. there are still many Austrian economists who reject the neoclassical model of perfect competition. the apparently simple exercise of defining the free market (and therefore state intervention) becomes not so simple any more—and this is even before we can discuss whether some markets are ‘failing’ and therefore state intervention may make them ‘more efficient’. I would even go as far as saying that defining a free market is at the deepest level a pointless exercise. the neoclassical theory of the market is only one of many legitimate theories of market and not a particularly good one at that.1 Unless we recognise the ultimately political determination of the rights–obligation structure that underlies market relationships. However. this did not happen. Needless to say. because no market is in the end ‘free’. So. 1982A). this is not considered a market 1 The same reasoning applies to the judgement on how interventionist a particular state is. which has strongly influenced the country’s industrial evolution through defence procurement programmes and defence-related R&D contracts—especially in industries such as computers. telecommunications and aviation (Johnson.01 CJE 26/5 pp. with one party (the developing countries) arguing that these are hidden protectionist measures that go against the very principle of free trade that the WTO is supposed to represent and the other (the developed countries) arguing that these are ‘universal’ standards that are perfectly compatible with free trade. refers to a situation when the market does not work as is expected of the ideal market. many ‘rival views of market society’ (Hirschman.2 Defining market failure The term. for a discussion from a legal perspective). There are. the Austrian theory of the market.4 and 4. 539-559 FINAL 544 23/7/02 4:10 pm Page 544 H. depending on their respective theories of the market.2 However. it is because of the political consensus that defence is one of the absolutely necessary functions of the state that many people underestimate the interventionism of the US federal government. It is only because some state regulations (and the rights and the obligations that they support. But what is the ideal market supposed to do? In the neo-liberal framework. And therefore the same market could be seen as failing by some people while others regard it as normally functioning.

at its core. does not exclude the possibility (which is often realised) that an economy may be full of monopolies but undynamic. The point that I have tried to illustrate with the above examples is that. given the possibility of government failure. say. especially institutionalist economists.01 CJE 26/5 pp.. will lead to an instantaneous diffusion of new technology and thus to an instantaneous dissipation of monopoly rents.g. Brazil. if the market fails. Therefore. it is usually better to live with failing markets than to attempt state intervention (see Section 2). and may support some ‘non-distortionary’ lump-sum income transfers to reduce inequality. even the firm exists only as a production function. Of course. more precisely. an economics about the market or. as in the same market where one person sees a perfection another person can see a miserable failure. 1992. perfect information. we need to make it clear what we expect from the ideal market. p. However. formal producer associations. while it may not matter so much for other types of economists. against which only can the failures of the existing markets be defined. a market which is perfect in the neoclassical sense (e. . and vice versa (the above example about non-competitive markets illustrates this point very well). Neoclassical economics is. Other institutions that make up the modern capitalist economy (e. with. needless to say. which is necessary for a perfectly competitive market to exist. the economy fails. However. a classic example of market failure in the neoclassical framework. because the ideal neoclassical market (or at least in the Paretian version of it) is not supposed to generate equitable income distribution in the first place. Otherwise. while Marx and Schumpeter would have argued that the existence of non-competitive (in the neoclassical sense) markets is an inevitable. to borrow Coase’s analogy. trade unions) figure basically as ‘rigidities’ that prevent the proper functioning of markets (for a criticism of the view of non-market institutions as ‘rigidities’. about the barter exchange economy. informal enterprise networks. 1995). 1987. the concept of market failure becomes empty. 539-559 FINAL 23/7/02 4:10 pm Page 545 Breaking the mould 545 failure. even these economists would argue that an equitable income distribution is not what we should expect from the ideal market and therefore that there is no market failure in Brazil in this sense. can we make our notion of market failure clear. ‘lone individuals exchanging nuts and berries on the edge of the forest’ (Coase. if a secondary. a non-competitive market is one of the most obvious examples of a failing market for neoclassical economics. 718). To take another example. the non-competitive market. is regarded as an inevitable feature of a successful dynamic economy from Schumpeter’s or Marx’s point of view. as far as they acknowledge 1 Recall Schumpeter’s famous metaphor that the relationship between the efficiency gains from competition through innovation and that from (neoclassical) price competition was ‘as a bombardment is in comparison with forcing a door’ (Schumpeter. see Chang. 84) 2 This.g. how much does market failure matter. This is not to deny that many well-intentioned neoclassical economists may dislike the income distribution prevailing in.2 Or to put it differently. namely. many neoclassical economists with neo-liberal leanings would argue that market failures do not occur often and that. for whom the market is essentially the economy. Now.1 feature of a dynamic economy driven by technological innovation. no market power) may look like an absolute failure to Schumpeter because perfect information. Only when we make our own theory of the market clear. for neoclassical economists.. and not as an ‘institution of production’. however we define it? The short answer is that it would matter greatly for neoclassical economists. In this world. when we talk about market failures. Thus. which means that there will be no incentive for entrepreneurs to innovate and generate new knowledge and new wealth. p.

which enabled the coordination of a most complex division of labour—so. may not be very necessary. our economic activities. when most economic interactions in the modern capitalist economy are actually conducted within organisations and not between them through the market (Simon. 20). 1991). and the state as the creator and regulator of the institutions governing their relationships (while itself being a political institution).01 CJE 26/5 pp. who regard the market as only one of the many institutions that make up the capitalist economic system. 20–1). 1975. as well as other informal institutions such as social convention. Such an outcome is due to the success of modern business organisations. including the markets as institutions of exchange. state intervention in these markets. the only alternative they will seriously contemplate (and ultimately reject) is state intervention. in many modern industries where there are high incidences of monopoly and oligopoly. as neoclassical economics does. The real point is that the market is only one of the many institutions that make up what many people call the ‘market economy’. In contrast. In other words. there was central planning’ (pp. Chang the existence of market failure. but these industries are often very successful in more commonsensical terms because they generate high productivity growth and consequently high standards of living. 3. especially interventions of the neoclassical anti-trust variety. and have organised. (Arrow (1974) is the most sophisticated example of this view. really gives us a wrong perspective in the sense that we lose sight of a large chunk of the economic system and concentrate on one. where neoclassical economists see a ‘market failure’. However. markets are failing all the time according to the neoclassical criterion.3 The market primacy assumption One fundamental assumption about the nature of the market and the state in neo-liberal economics. And if this is indeed the case. Thus. and indeed under certain circumstances may be harmful. there were the markets’ (Williamson. . which is also shared even by the neoclassical economists without neo-liberal leanings. albeit important. and not simply as add-ons. p. For example. as well as other non-market institutions. non-state institutions as integral elements. or what I think is better called capitalism. is what I call the market primacy assumption—or the assumption that ‘in the beginning. The capitalist system is made up of a range of institutions. The point that I am trying to make here is not that market failures do not exist or that they do not matter at all—on the contrary. the real world is full of market failures and they do matter.) 1 Williamson defends this starting assumption on the ground of ‘expositional convenience’. the state. part only. the neoclassical) of many possible criteria may not really make much difference for the performance of the economy as a whole. arguing that the logic of his analysis would be the same even if the starting assumption was that ‘in the beginning. because they know that there are many institutions other than markets and state intervention through which we can organise. the firms as institutions of production.1 In this view. focusing on the market (and market failure). he never explains why and how one assumption makes the exposition more convenient than the other. 539-559 FINAL 546 23/7/02 4:10 pm Page 546 H. 1991). institutionalist economists may see an ‘organisational success’ (Lazonick. market failures may not matter as much. This suggests that we badly need an explicitly ‘institutionalist’ perspective that incorporates nonmarket. for institutionalist economists. the fact that some (or even many) markets are ‘failing’ according to one (that is. is seen as a man-made substitute which emerged only after market failures became unbearable.-J. because no intermediate institutions or organisations have a place in their scheme.

the fact that it is still taken so seriously among neo-liberal thinkers is symptomatic of their adherence to the market primacy assumption. as well as a range of other ‘institutional’ solutions (e. state intervention played a critical role in the emergence of individual markets and of the market system. Arrow. part of human economic life until the rise of capitalism.01 CJE 26/5 pp. Thus even those who wished most ardently to free the state from all unnecessary duties. Even in the UK. for example. is the classic work making this point. the amount of bureaucratic control involved in the administration of the New Poor Laws which for the first time since Queen Elizabeth’s reign were effectively supervised by central authority. and whose whole philosophy demanded the restriction of state activities.. this view explains even the very existence of the state itself as a market-like (contractual) reaction to market failure. as even many of its proponents acknowledge. which the Austrian-libertarian wing of neo-liberalism has used with great political effect. which is seen as necessary (and often sufficient) for markets to function (Nozick. in the beginning. p. ’ (Polanyi. centrally organised and controlled interventionism [italics added]. Thus. Buchanan. especially in the early stage of capitalist development (Polanyi. 1992. Economic historians have repeatedly shown us that. 1974. Witness the complexity of the provisions in the innumerable enclosure laws. could not but entrust the self-same state with the new powers. and instruments required for the establishment of laissez-faire [italics original]’ (p. In fact. there were not markets.g. 1957. The plain truth is that. organs. once argued that ‘markets develop naturally’ (Stiglitz. . To make Adam Smith’s “simple and natural liberty” compatible with the needs of a human society was a most complicated affair. It is well known. . In Polanyi’s words. one of the most enlightened neoclassical economists of our generation. . these economists would still see state intervention. except at the very local level (in supplying basic necessities) or at the very international level (in luxury trade). the state had emerged as a ‘contractual’ solution to the collective action problem of providing the public good of law and order. the firm) as man-made substitutes for the ‘natural’ institution called the market. See. There are many economists who start their analyses (at least implicitly) from the market supremacy assumption but willingly endorse a relatively wide range of state intervention.. Stiglitz (1999). especially the security of private property. ch. of course.2 1 However. Schotter. let alone the dominant. 140). although even Joseph Stiglitz. it must be emphasised that the fact that someone attributes institutional primacy to the market does not necessarily mean that he/she endorses a minimal state view. 1999). that this explanation is obviously at odds with the historical evidence. where the ‘state of nature’ is the state of ‘free’ market to the extreme degree (including in the provision of law and order) and that the ‘natural’ reaction of ‘free’ individuals to this undesirable state of affairs is to engage in a market-type behaviour of voluntarily signing a social ‘contract’ to set up the state as the provider of law and order (for more detailed criticisms of the contractarian argument. In this view. 1957. or the increase in governmental administration entailed in the meritorious task of municipal reform . 1). also see Block. the market was not an important. At this point. 1994A. where the market economy is supposed to have emerged ‘spontaneously’. 539-559 FINAL 23/7/02 4:10 pm Page 547 Breaking the mould 547 The most obvious example of the market primacy assumption is the contractarian explanation of the origin of the state. more recently Stiglitz has moved away from this view and embraced a more (if not completely) institutionalist position. p. However. 75).g. 140). ‘[t]he road to the free market was opened and kept open by an enormous increase in continuous. non-state institutions (e. However. and the other non-market. 1974. see Chang. 1986). 2 And he continues: ‘Administrators had to be constantly on the watch to ensure the free working of the system.1 the emergence of markets was almost always deliberately engineered by the state. 1985).

1995. The exact focus of intervention has certainly varied across time and space.2) and Kozul-Wright (1995. post-World War II French industrial policy. p. Table 4. Japan got tariff autonomy only in 1911 when all its unequal treaties. What we have just discussed is not simply of historical interest. For details. p. whether or not we accord institutional primacy to the market makes a critical difference to how we design developmental policies for countries which are yet to set up a fully developed market system. and modifying the existing rights–obligations structure in order to accommodate them. the state is constantly involved in creating new markets and thus setting up the new rights and obligations necessary for their functioning. and usually around 40%. reflecting what I have elsewhere called the ‘institutional diversity of capitalism’ (Chang. on the one hand. were rarely over 20%. Chang Also in the case of the US. Belgium. expired. 1995). the developmental crises that many developing countries have gone through during the last two decades or so also show how dangerous it is to assume the primacy of the market and believe that it will naturally develop so long as the state does not interfere with its evolution. 1995. see World Bank (1991. these countries would not be in such trouble now. the two supposed models of marketbased development. the US was the birthplace of the idea of infant industry protection (Freeman. Its average tariff rates since the 1820s was never below 25%.2). In fact.1 Once we accept that even the UK and the US. to take issue with the market primacy assumption in neo-liberal theory is not merely a theoretical quibbling nor is it a quest for historical ‘truth’. were key factors in its success in early industrialisation (Kozul-Wright. Box Table 5. if markets evolved as ‘naturally’ as neo-liberal economists believe. even the World Bank now recognises this—see the World Bank.8). on the other hand. The most prominent recent examples include the creation and the restructuring of markets by the state in mobile telecommunications. the funding of agricultural research. France. which it signed following its opening-up in 1853. the well-known stateled developments of the East Asian countries. state intervention in establishing property rights. Reinert. early Swedish state support of research and development. when those in other countries for which the data are available. 1989. the post-World War II transformation of the Austrian manufacturing sector through dynamic public enterprises.01 CJE 26/5 pp. it is much easier to see that there is virtually no country (except probably Hong Kong) that has achieved the status of an industrialised country without at least some periods of heavy state involvement. For one thing. Of the countries with tariff autonomy. also see Albert. did not develop through a spontaneous emergence of markets. Stiglitz. . which on the whole already have welldeveloped market systems. 1999). such as Austria. 1997. Berger and Dore 1996): ‘pre-emptive’ welfare state in Bismarckian Germany. p. and was indeed the most heavily protected economy in the world for about a century until World War II. 1991. and so on. 21. Therefore.-J. few countries had tariff autonomy either because of colonial rule or unequal treaties— for example. the US had by far the highest tariff rates. 97. Italy and Sweden. even in the most advanced capitalist economies of today. the fact remains that all successful developmental efforts involved substantial state intervention. Box 1. facilitating the provision of critical physical infrastructure (especially railways and telegraphy). Nevertheless. But perhaps more importantly. For example. the severe economic crisis that many former Communist countries that have opted for a ‘big bang’ reform have experienced during the last several years is one striking example showing how the establishment of a well-functioning market economy is impossible without a wellfunctioning state (see Chang and Nolan. electricity and Internet service provision. This assumption deeply affects the very way in which we understand the nature and the development of the 1 During this period. 539-559 FINAL 548 23/7/02 4:10 pm Page 548 H. 1997. 97. computer software. Likewise. Most importantly.

King (1987). Unless we abandon this assumption and develop a theory that deals with the market. in this section I should like to criticise the neo-liberal view on politics from another angle. This. Gamble (1988). is a highly political exercise. for example. according to their view. politics opens the door for sectional interests to ‘distort’ the ‘rationality’ of the market system.4 Market. Stretton and Orchard (1994). 539-559 FINAL 23/7/02 4:10 pm Page 549 Breaking the mould 549 market. Weiss (1998). the anti-statist conclusions of neo-liberalism need to be seriously modified. These studies argue that. contrary to the neo-liberal assumption. Once this assumption of pure self-seeking is dropped. the state and other institutions on a more equal footing. and Woo-Cumings (1999). The most extreme example will be the various episodes of ‘original accumulation’ in which property rights were redistributed through the most naked forms of politics. is to be achieved by restricting the scope of the state (through deregulation and privatisation) and by reducing the room for policy discretion in those few areas where it is allowed to operate. Chang. The neo-liberal solution to this problem is that we ‘depoliticise’ the economy.1). independent regulatory agencies). state and politics As we mentioned earlier. the establishment and the distribution of property rights and other entitlements that define the ‘endowments’ that market participants have. 1994A. As this point is already well known and as I shall develop this theme further later (Section 4. 1994).g. theft and even violence—such as the Great Plunder or the Enclosures in the early days of British capitalism or the shady deals that dominate the privatisation process in many developing and ex-communist countries today. Even a basic knowledge of the history of the advanced countries over the last two centuries reveals how many of those rights that are now regarded as so ‘fundamental’ that very few. There are many studies that take issue with the neo-liberal view of human motivation that underlies this political economy analysis (Cullis and Jones. . as the moral views and social norms held by individuals may restrain the extent to which they advance their interests by finding ways to ‘distort’ market outcomes through political means—that is. as well as its relationship with the state and other institutions. by strengthening the rules on bureaucratic conduct or by setting up ‘politically independent’ policy agencies bound by rigid rules (e. in chronological order. 1994B). if any. see.01 CJE 26/5 pp.2). My point here is that the market itself is a political construct and therefore the neo-liberal proposal for its de-politicisation is at best self-contradictory and at worst dishonest. 1987. Toye (1991). of their citizens will question them were perfectly contestable and often fiercely contested in the past—examples include the right to self-ownership (denied to 1 For further criticisms of neo-liberal political economy. which neoliberal economists take as given. Chang (1994A. even if all political modifications of existing rights and obligations can be interpreted as ‘distortion’ of the market through political means (I showed why this cannot be the case in Section 3.1 But what does it exactly mean to say that markets are political constructs? To begin with. the neo-liberal world of politics is one that is populated by selfseeking bureaucrats and politicians with limited capabilities and operating under the influence of interest groups. Stretton and Orchard. our understanding of the role of the state will remain severely incomplete and biased. Evans (1995). independent central bank. self-seeking is not the only human motivation even in the ‘private’ domain of the market.. 3. and that people do not operate with the same degree of selfishness in the public domain as in the private domain. involving corruption. In this view.

pollution. Interest rates are also highly political prices. at least de facto. which was prompted by the European Monetary Union. Chang slaves). sustenance and modification of the rights–obligations structures underlying markets will never end. The neo-liberal call for de-politicisation is often justified in a populist rhetoric as an attempt to defend the ‘silent majority’ from corrupt politicians. Wages are politically modified not simply by minimum wage legislation but also by various regulations regarding union activities. even when they are determined by a ‘politically independent’ central bank (for further discussion. which itself is a product of politics (see Vira. in calling for de-politicisation of the economy. what the neo-liberals are really doing when they talk of de-politicisation of the market is to assume that the particular boundary between market and the state they wish to draw is the ‘correct’ one. 2000). fiefdom-building bureaucrats and powerful interest groups. it is only because they contest the right of the people to influence monetary policy through their elected representatives. and so on. However. shows this very clearly. willy nilly. the right to organise. To begin with.-J. two critical prices which affect almost every sector. and the right not to be subject to physical abuse in workplaces. see Grabel. So.01 CJE 26/5 pp. where the British coal-miners were told to accept the logic of the ‘world market’ and face mine closures with grace. which the then British government argued to be beyond political negotiation. equal treatments across sexes or ethnicities. even when we accept the existing rights–obligations structure as uncontestable. there is virtually no price which is free from politics. the ‘market rationality’ that the neo-liberals want to rescue from the ‘corrupting’ influences of politics can only be meaningfully defined with reference to the existing institutional structure. turned out to be determined by the ‘political’ decisions of the German government to give subsidies to their coal.1 In other words. including those that are not perceived as such even by many neo-liberals. welfare entitlements. However. and that any attempt to contest that boundary is a ‘politically minded’ one. the world market prices. namely. are politically determined to a very large degree. labour standards. import contents. the diminution of the legitimate domain of politics that the neo-liberal proposal for de-politicisation will bring about only serves to diminish further what little 1 We were reminded of this clearly in the British coal crisis under the Conservative Government in the early 1990s. and not because there is some ‘rational’ reason that monetary policy should not be politically influenced. The recent debate in Europe on the relationship between political sovereignty and autonomy in monetary policy. but are also undermining. of the French government to allow the export of their subsidised nuclear electricity. child labour in their coal mines. the neo-liberals are not only dressing up their own political view as ‘objective’ ones that are ‘above politics’. More recent struggles regarding rights in areas such as the environment. 1998. . for a further exposition of this point). the right to vote (and thus to have a say in the political modification of market outcomes). And if this is the case. and consumer protection are reminders that the political struggles surrounding the establishment. wages and interest rates. Moreover. 539-559 FINAL 550 23/7/02 4:10 pm Page 550 H. and most importantly immigration control. if some people feel that central banks should be politically independent. as we argued in Section 3. the right to minimum working hours. and of the many developing country governments to allow. there is no one ‘correct’ way to draw such boundary. the principle of democratic control. When we add to them those numerous regulations in product markets regarding safety. it is only because those who are concerned do not even realise that the rights–obligations structure which underpins that boundary is potentially contestable. there are practically no prices which are not in reality subject to ‘political’ influences. If there appears to be a solid boundary between the two in certain instances. Moreover.1. However.

1997) and Chang and Rowthorn (1995B). 1995. Hodgson. high transaction costs would be incurred in search and bargaining activities if every allocative decision was regarded as potentially contestable. the state and politics. and therefore a full depoliticisation of the market is not only an impossibility but also has a dangerous antidemocratic undertone. 1999. 1 Thus seen. Joseph Schumpeter. as the neo-liberals do. 1995. Andrew Shonfield and Herbert Simon (recent developments of this tradition can be found in Hodgson. that no market under any circumstance should be subject to political modifications. unless the resource allocation process is at least to a degree accepted as ‘objective’ by the members of society. 1982B. 539-559 FINAL 23/7/02 4:10 pm Page 551 Breaking the mould 551 political influence the so-called ‘silent majority’ have in modifying market outcomes. are heavily influenced by politically determined institutional parameters in the first place. Moreover. but a development of the tradition found in the classic works of authors such as Karl Marx. and suggest how these analyses may be developed. I argue these problems can be overcome only by adopting an alternative approach that incorporates politics and institutions into its analytical core. because in the final analysis. 1988.2 As it is beyond the scope of this paper to develop this new approach fully. unlike the old Liberals. Karl Polanyi.01 CJE 26/5 pp. 2000). The way forward: towards an institutionalist political economy My discussion so far has revealed some important limitations of the currently dominant neo-liberal discourse on the role of the state. the political legitimacy of the economic system itself may be threatened. however. As the reader may have noticed already and as it will become clearer later. differs from the NIE in a number of important respects (see Rutherford. But before we proceed further. there is no market which can be really free from politics. See Chang (1994B. This tradition. Block. 1996. see Bobbio. we repeat. but also as 1 We should also note that political activities are often ends in themselves. Thorstein Veblen. this is not the same as arguing. Evans. the neoliberals cannot openly oppose democracy. the market is ultimately a political construct. 2000. However. 1993. 2000). which is sometimes called the ‘old Institutional Economics’. so they try to do it by discrediting politics in general and by making proposals which ostensibly seek to reduce the influence of those ‘untrustworthy’ politicians and bureaucrats but ultimately diminish democratic control itself. Lazonick. 1990). as was the case in the ex-communist countries. that by saying this we are not denying that a certain degree of de-politicisation of the resource allocation process may be necessary. which. Chang and Evans. in the rest of this section I shall attempt to describe the central theoretical features that distinguish it from the neo-liberal approach in the analyses of the market. which I propose to call an institutionalist political economy (IPE). In this section. pp. the neo-liberals deep down believe that allowing political power to those who ‘do not have a stake’ in the existing system will inevitably result in the ‘irrational’ modifications of status quo (for a critical exposition of old Liberalism along this line. one thing needs to be made clear. 2 I have attempted to develop elements of this theory in a number of my previous works.. 4. 1991. For one thing. Burlamaqui et al. 2000. Note. when I refer to an ‘institutionalist’ approach. However. but most importantly in seeing institutions not simply as constraints on the behaviours of the pre-formed and unchanging individuals as in the NIE. Like the old Liberals. and people may derive value from the activities per se as well as from the products of such activities (see Hirschman. 85–6). I do not mean it to be of the New Institutionalist Economics (NIE) kind. .

slaves. Who can participate in which labour market will be affected not only by formal regulation of the state and by private sector agents (e. we need to understand a wide range of institutions that affect and are affected by it. decisions by professional associations or social conventions may.01 CJE 26/5 pp.g. when pressed. and often implicit. by implication.g. some simplified notion of private property rights is all that exists in the neo-liberal analysis of the market. professional associations. we need institutions that define what exactly each agent’s rights and .g. It argues that.. of course. Laws on slavery. be enforceable through the legal system).. They also include private-sector selfregulatory institutions (e. Many of these institutions that need to be incorporated into the analysis of the market are often ‘invisible’ because the rights–obligations structure that underlies them is taken so much for granted that they are seen as inalienable components of naturally ordered free markets (see Section 3. ‘indecent’ publications. respectively. that human beings. While. in the neo-liberal discourse. contract law). no institution.. producer associations) and informal institutions such as social conventions. Banking laws or pension laws may limit the range of assets that banks or pension funds can own and therefore limit the range of asset markets that they can enter. Second. In contrast. These institutions are not. In most countries. gender and ethnicity.. IPE highlights the institutional complexity of the market.1). institutional specification. Company laws and industrial licensing rules will decide who can participate in the product market. 4.g. 539-559 FINAL 552 23/7/02 4:10 pm Page 552 H. rules of unions and professional associations) but also by social conventions regarding caste. although some of them may also consider those institutions that are needed for effective exercise and modification of property rights (e. women) cannot own property.1 Analysis of the market As I have argued above. Usually. But. addictive drugs or indecent publications).. say. in order to understand the workings of the market. laws may stipulate that certain types of individuals (e. ownership). Third. even when the legitimate participants in and the legitimate objects of exchange have been stipulated. however ‘natural’ it may look. there are laws outlawing transactions in things such as addictive drugs. foreigners. laws regulating professional qualifications.g. all markets are based on institutions that regulate who can participate. their analysis of the market itself involves only a minimal.-J. For example. Chang shaping the individuals themselves. human organs or firearms (although different societies have different views on what count as. while stock market listing rules and brokerage regulations determine who can participate in the stock market. can be regarded as such. there are institutions which determine the legitimate objects of market exchange (and. when it comes to the crunch. With this in mind. simply formal institutions such as law and state regulation. let me now sketch out what I think needs to be the distinguishing features of IPE. in the final analysis we should be willing and able to subject all institutions that support markets to analytical and political scrutiny. child labour and immigration will stipulate.3). the labour service of children and the labour service of illegal immigrants may not be legitimate objects of exchange. and although in many cases we may choose to accept many institutions as given. the market is seen as a ‘natural’ economic phenomenon that grows spontaneously out of the universal human nature to exploit gains from trading (see Section 3. most neo-liberal economists would admit that the market itself is an economic institution and while. the court system. To begin with. although many of these institutions are supported by formal institutions (e. given the recent influence of the NIE. many of them may even talk about some (albeit not all) non-market institutions such as the firm.

which the neo-liberals focus on.4). what acceptable conducts in the exchange process are. as many critics in the institutionalist tradition have pointed out. if we are to understand them properly. breach of contract. Consumer laws and liability laws. 1984. asserts that self-seeking is the only ‘genuine’ human motivation. McPherson. for example. safety and grievance resolution in workplaces will define the rights and the obligations of workers and employers. 1983.g. fire regulations.. 1988.. define how property rights in land can be exercised (e. In addition to property rights and the legal infrastructure that help their exercise and modification. zoning laws. In other words. 4. bankruptcy and other disruptions in the exchange process. And this criticism applies even more to the analysis of the state and other aspects of public life. which cannot be explained without admitting a range of non-selfish motivations and assuming a complex interaction between them (Simon. and we need a fuller institutional specification of markets. which are backed up by the police. and there are just too many non-selfish human behaviours. default. those regarding fairness and probity) or codes of conduct issued by trade associations (e. IPE adopts a ‘political economy’ approach not only in the analysis of the state but also in the analysis of the market. and for that matter neo-liberalism as a whole. 1999. and so on. Ellerman. 539-559 FINAL 23/7/02 4:10 pm Page 553 Breaking the mould 553 obligations are in which areas. and can be. whose legitimacy (and therefore whose contestability) is ultimately determined in the realm of politics. Social conventions (e. the laws regarding health. to take another example. except perhaps vis-à-vis family members (Williamson. what the legitimate objects of exchange are. 1997. Frey. neo-liberal markets are institutionally very under-specified. For example.01 CJE 26/5 pp. party loyalty. and in what terms different types of agent may participate in which markets. regarding pollution or noise).g. So. nationalism) . Markets are in the end political constructs in the sense that they are defined by a range of formal and informal institutions that embody certain rights and obligations. the court system and other legal institutions. Basu. ‘de-politicised’. there are numerous institutions that regulate the process of exchange itself. For another example.. will stipulate when and how buyers of unsatisfactory or faulty products may annul the act of purchase and/or claim compensation from the sellers. Etzioni. 1993. understanding the market requires consideration of a much wider range of institutions than what is normally discussed by the neoliberals. human motivations are multi-faceted. However. also see Section 3.g. To sum up the discussion in this section. is a recent and very passionate advocate of this view). what kinds of building can be built where). Consequently. liberalism. there are rules regarding fraud. Finally.g. social reform.g. This is not only because individuals often join public life with commitments to certain non-selfish values (e.2 Analysis of the state The neo-liberal analysis of the state starts by questioning the ‘public’ nature of the motivations of the agents that make up the state such as politicians and state bureaucrats. 1983. a public service ethic. and so on.. environmental regulations (e.. The theory of human motivation and behaviour underlying this analysis. bankers’ association) may also influence the way economic agents behave in economic transactions. Emphasising the institutional nature of the market in the above-discussed way also requires that we have to bring politics explicitly into the analysis of the market (and not just into the analysis of the state) and to stop pretending that markets need to be. we also need to consider all the other formal and informal institutions that define who can hold which kinds of property and participate in which kinds of exchange.

in turn. IPE does not see these motivations as given but as being fundamentally shaped by the institutions surrounding the individuals. emphasising the public service ethic in bureaucratic training). Hodgson. IPE acknowledges the usefulness of these institutions that target behaviour directly. In addition to accepting the variety and the complexity of human motivations. but would argue that behavioural standards can be also improved. but they are not able to change the motivation itself (Ellerman. 1999. Chang but also because. can happen through direct ideological exhortation (e. In neoliberal theories (including the NIE models).. we also need to acknowledge that human beings are fundamentally shaped by institutions. IPE’s emphasis on the constitutive nature of institutions should not be interpreted as meaning that people’s motivations are more or less determined by the institutional structure. and is a central hallmark of a truly ‘institutionalist’ approach. social norms. I would go a step further and argue that some of the neo-liberal recommendations that are intended to improve the behavioural standards of public personages may be downright counter-productive. diligently documenting their . This is because institutions embody certain ‘values’ (worldviews. Indeed. operating in an explicitly ‘public’ sphere of life. institutions may be able to shape individual behaviours by punishing or rewarding particular types of behaviour. 2000). In these theories.. tougher punishments). even if they are subject to the same institutions involving individualistically designed sanctions and rewards of the kinds that the neo-liberals recommend (e. different from the neo-liberal institutionalism of the NIE. However. if they undermine the non-selfish motivations that had previously motivated the public personages in question—that is. Of course. an increase in the monitoring of public figures may be able to make them behave in a more ‘moral’ way in areas where monitoring is easier (e. 2000) or what Hodgson (2000) calls the ‘downward reconstitutive causation from institutions to individuals’. they end up internalising many ‘publicly oriented’ values. Therefore. 539-559 FINAL 554 23/7/02 4:10 pm Page 554 H.g. by changing the motivation of public personages. moral codes. higher relative salaries. IPE differs from NIE in that it postulates a two-way causation between individual motivation and social institutions. individuals inevitably internalise some of these values and thereby have their selves changed. devising incentive systems that reward teamwork in the bureaucracy in order to boost esprit de corps). although IPE would agree that in the final analysis a truly institutionalist analysis should see institutions as at least ‘temporally’ prior to individuals (Hodgson. 2000). and in some cases more effectively improved. let me illustrate how an ‘institutionalist’ analysis of the relationship between motivation. If IPE is not to lapse into an unwarranted structural determinism. but perhaps more indirectly (given the constitutive role of institutions) through changing the institutions that surround them (e.g..-J. government officials may act with more probity compared with their counterparts in other societies without such a norm. behaviour and institutions may improve our thinking on the role of the state.01 CJE 26/5 pp. or whatever one may choose to call them) and.g.. This. by operating under these institutions.g. In societies where high standards of behaviour in public life have long been established. we need to accept that individuals also influence the way institutions are formed and run. as is typically done in the NIE models. In contrast. rather than a one-way causation from individuals to institutions. This is what I elsewhere proposed to call the ‘constitutive role of institutions’ (Chang and Evans. individual motivations (which they usually call ‘preferences’) are treated as the ultimate data. if they cause what Ellerman (1999) calls the ‘atrophy of intrinsic motivation’. Now. more thorough monitoring.

However. IPE argues that behaviour may be changed not only through changing institutions that define incentives for individuals. . they are claiming that markets should be. as neo-liberals wish (see Section 3. the neo-liberals are claiming an objectivity that no theory can claim. This is because it will make them feel that they are not trusted as ‘moral’ agents any more and therefore that they are under no moral obligations to behave morally unless they are forced to do so. by portraying the particular boundary of the market that they are advocating (into which. and adopt a more complex view of the inter-relationship between motivation. Moreover. rather than as a process in which interest groups try to change the ‘natural’ order of ‘free markets’ according to their own sectional interests. However. In contrast. In doing so. IPE argues that we need to see politics as a process through which people with different. Second. it must be emphasised that even in the private sphere the importance of self-seeking motivation is much less than what the neo-liberals believe. IPE argues that markets are fundamentally political constructs and therefore that it is not possible. Moreover. I accept that this myth may be useful.. it is a myth that markets can be free from politics. the usefulness of this myth does not change the fact that it is a myth.3 Analysis of politics Neo-liberalism has made an important contribution to the debate on the role of the state by bringing politics back into the analysis of state action. individual motivations are fundamentally formed by institutions that surround the individuals. as one’s political view will deeply influence whether one sees a particular boundary as a legitimate (or ‘rational’ in their language) one. IPE proposes that we accept. taking intellectual initiatives without material compensation). that human motivations are varied and interact with each other in complex ways. and can be. we should acknowledge that there is no need for selfish motivations to dominate our behaviour in the public sphere of the state. behaviour and institutions than exists in the neo-liberal discourse. However.g. However. 4. views on the contestability of the existing rights–obligations structure vie with each other. Since I am advocating an institutionalist ‘political economy’. they are effectively making two claims. First. where non-selfish values are institutionally emphasised and therefore the actors internalise many such values. free from politics.4). it argues. we can see that there is no ‘objective’ way to decide the ‘correct’ boundary between the market and the state. I am naturally sympathetic to the neo-liberal attempt to emphasise the role of politics. it may make them less motivated to behave in a moral way and take initiatives in areas where monitoring is difficult (e. as I have argued. we need to abandon its arguably most crucial assumption. they argue. In order to overcome the limitations of the neo-liberal analysis of the state. the neo-liberals claim that politics inevitably generates state actions that go against market ‘rationality’. but also through ideological and institutional changes that influence individual motivations themselves. once we accept the political nature of the market. namely the assumption that individuals have pre-formed motivations (or ‘preferences’ in the neo-liberal language) that are selfish. in containing the potentially disruptive effects of a very high degree of contestation of the rights–obligation structure underlying existing markets. and equally legitimate. both of which are highly problematic. 539-559 FINAL 23/7/02 4:10 pm Page 555 Breaking the mould 555 expenses for business trips). to try to rid markets completely of politics.01 CJE 26/5 pp. or even necessary. from the start. Thus seen. However. political influences should not be allowed) as the ‘rational’ one. or even desirable. Let me summarise the argument in this section.

Second. not simply because they fear some formal or informal sanctions but. considered to be of ‘poor taste’ than in societies where it is not.g. rent-seeking is likely to be less widespread in societies where open lobbying is. 539-559 FINAL 556 23/7/02 4:10 pm Page 556 H. Moreover. for example. not only because markets themselves are political constructs but also because the notion of an ‘uncorrupted’ market that the neoliberals have is based on a particular set of political beliefs that cannot claim superiority to other sets of political beliefs. I would go even further and criticise the neo-liberal analysis of politics for its failure to recognise the extent to which politics itself is an institutionally structured process (see Chang and Evans. institutions influence people’s views on what kinds of issues are legitimate targets of political action. because they do not even think the issue to be a legitimate item on the agenda for political action by any group (see Goodin. Unless we break the neo-liberal mould and see institutions as both constraining people’s behaviour . institutions influence the very perception of their interests by individuals. neo-liberals fail to see politics as an institutionally structured process in the deepest sense. although it acknowledges the harm that excessive politicisation can do. So. rules regulating the behaviour of public figures. And third. and of the appropriate standards of behaviour in politics. but different.. also see Section 4. mechanisms that are involved here. but because they influence people’s perception of their own interests. not only because institutions shape people’s political actions. However. given their motivations and perceptions. like the other neo-liberal analyses involving institutions. not even people who would potentially benefit from such a practice will start lobbying for its re-introduction. On the contrary. 1989. formal affiliation of political parties with trade unions or employers’ associations).. IPE argues that politics is an institutionally structured process. rules on agenda formation and voting in parliamentary committees). To summarise the argument in this section. even if both societies have the same scopes for rent-seeking.g. namely that they can influence politics not only by affecting human actions but also by influencing individual motivations and worldviews (Chang and Evans. many more voters will vote along ‘class lines’ than in societies without such parties. the neo-liberal claim that politics inevitably corrupts the market is problematic. argue along this line from a political science point of view). in societies where child labour is no longer a legitimate policy issue. electoral rules. 2000.-J. for example. First. even if legal. institutions influence how individuals perceive the legitimacy of particular types of political action.2). operation and change. So. often with success. more importantly. 2000. So. for further discussions of the issue of ‘public agenda formation’). 1986. Chang IPE treats politics not as something alien and damaging to the market but as an integral part of its construction. it has not gone beyond seeing institutions as ‘constraining’ factors on human behaviour and fails to see that institutions are also ‘constitutive’. I am not saying here that institutions do not feature in the neo-liberal analysis of politics. March and Olsen. it has tried to analyse. They see institutions as constraining political actions but fail to see that institutions also affect people’s motivations and perceptions. We can talk about three related. in societies with political parties with more class-conscious organisations (e. for example. Of course.01 CJE 26/5 pp. It also emphasises that there is no such thing as a ‘correct’ political view and therefore that no one should be able to claim the boundary between the market and the state that he/she believes in to be the ‘correct’ one. of the legitimate boundary of politics. how the formal and informal institutions that govern the way in which interests are organised and power exercised affect political actions (e.

New York and London. no. W. this paper is only the first step on a potentially long and laborious road to developing a full-blown institutionalist political economy. Concluding remarks In this paper. and the analysis of politics. 1974. Economic and Political Weekly. My main criticism is that the very way in which it envisages the market. as well as their mutual relationships. it emphasises the role of institutions in affecting human actions. the political economy of IPE goes much further than its neo-liberal counterpart in that it emphasises the fundamentally political nature of the market and applies the political economy logic to the analysis of the market. after pointing out the internal contradictions of the neo-liberal view on the role of the state that arises from the tensions between its neoclassical and Austrianlibertarian components. 2011–12 . institutions and politics. K. is highly problematic. it puts emphasis on the role of political factors in determining state policy. IPE is an ‘institutionalist’ approach because. On why we do not try to walk off without paying after a taxi-ride. I proposed calling it institutionalist political economy (IPE) and sketched out how its analyses of the market. the definition and the implications of market failure. like the neo-liberal analysis. The Limits of Organisation. Admittedly. especially given that the broader institutionalist framework that should provide a background to this approach is still not fully developed. including the state). including those within and surrounding the state. M. 5. Four Wall Eight Windows Arrow. However. and not just to the analysis of the state. At the same time. Capitalism. 539-559 FINAL 23/7/02 4:10 pm Page 557 Breaking the mould 557 and being constitutive of their motivations and perceptions. Capitalism vs. like the New Institutionalist branch of neo-liberal economics. Norton Basu. 1991. My main criticism of the neo-liberal analysis of the role of the state is not that its policy recommendations are too anti-interventionist. our understanding of politics will remain biased and incomplete. Four main points were raised: the definition of the free market. Therefore. but only by breaking this mould and developing an alternative framework that brings institutions and politics to its analytical core. IPE is a ‘political economy’ approach because. K. 1983. W. the state. as it is done in NIE) and that it sees institutions as not simply ‘constraining’ individual behaviour (as in NIE) but also as being ‘constitutive’ of individual motivations. the market primacy assumption (namely the view that the market is logically and temporally prior to other institutions. as some of its critics argue. we examined critically some of its basic concepts and assumptions from the institutionalist point of view. However. 48. it is hoped that the paper serves a useful role by proposing a new research agenda that will allow us to break the mould of current debate on the role of the state set by the very powerful and informative but fundamentally flawed and misleading discourse of neo-liberalism.01 CJE 26/5 pp. Bibliography Albert. I suggested that overcoming the limitations of the neo-liberal discourse on the role of the state cannot be done by looking for more interventionist models within the neo-liberal mould. However. the state and politics differ from the ones that are offered by the neo-liberal discourse. the institutionalism of IPE goes much further than that of NIE in that it emphasises the ‘temporal priority’ of institutions over individual (rather than the temporal priority of individuals over institutions. New York.

(eds). G. 2000. 1982. National Diversity and Global Capitalism. in Chang. Brighton. Oxford University Press Coase. Structuration Theory and Critical Realism’. Polity Press Hodgson. London. P. J. New York. I. mimeo. Davis Bobbio. H-J. Chang Berger. Markets and Citizenship. Oxford. no. no. NJ. (eds) 2000. Macmillan Chang. Princeton. Not Just for the Money—An Economic Theory of Personal Motivation. no. 1995. 1 Frey. in Chang. Wheatsheaf Books Burlamaqui. Palma. H-J. H-J. G. madness. A. H-J. Polity Press Hodgson. P. Foundations of Social Choice Theory. R. Oxford. The institutional structure of production. State. DC. Unemployment. Oxford. 1. Europe versus Asia—contrasting paths to the reform of centrally planned systems of political economy. ‘Helping Others to Help Themselves: The Challenge of Autonomy-Compatible Development Assistance’. 1 Hayek. in Killick. Role of the state in economic change—entrepreneurship and conflict management. and Rowthorn. 1997. H-J. 4 Cullis. A. S. vol. 1986. 1995. Edward Elgar Gamble. P. H-J. Macmillan Chang. in Chang. Department of Sociology. and Economic Policy. 82. and Nolan. mimeo. Cheltenham. and Rowthorn. and Jones. 24. N. 1982.). The State and the Economic System. 1987. J. Embedded Autonomy—States and Industrial Transformation. (eds). and Dore. R. Macmillan . Cambridge University Press Grabel. London and Basingstoke. G. F. The Capitalist State. 539-559 FINAL 558 23/7/02 4:10 pm Page 558 H. 1987. Oxford. ‘Structures and Institutions: Reflections on Institutionalism. 2000. H-J. London. R. Cheltenham. Cambridge Journal of Economics. in Elster. P. F. Essays in Hour of Geoff Harcourt—Volume 2: Markets. vol.). Liberalism and Democracy.-J. (eds). Basil Blackwell Johnson. 1995. (eds). Liberty. and structural change. Washington. London. H-J. 1990. Institutions and the Role of the State. vol. Market and State. Cambridge. 1999. 1993. 1982A. Rival views of market society. 1995A.. (ed. 1997. and Rowthorn. Economics and Evolution. T. Routledge & Kegan Paul Hirschman. MITI and the Japanese Miracle. no. B. The role of institutions in economic change. (eds) 1996. Routledge Chang. 1994B. A. G. Laundering preferences. Ithaca and London. D. Oxford University Press Ellerman. M. Explaining ‘flexible rigidities’ in East Asia. and Sawyer. 1995B. Basil Blackwell Deane. H-J. Stanford. NJ. Princeton University Press Freeman. and Hylland. 1988. P. The New Right: Politics. Princeton. The Free Economy and the Strong State: The Politics of Thatcherism. The political economy of ‘policy credibility’: the new-classical macroeconomics and the remaking of emerging economies.. J. 1982B.01 CJE 26/5 pp. R. New technology and catching-up. Journal of Economic Literature. H-J. 1988. and Evans. The Business School. The Transformation of the Communist Economies—Against the Mainstream. 1949. 1994A. 1992. Cambridge. institutions. no. Stanford University Press King. L. 1989. 1988. Microeconomics and the Public Economy: A Defence of Leviathan. C. Contractarianism and democracy. 48. H-J. Structural Change and Economic Dynamics. London. Introduction. Role of the State in Economic Change. 1999. paper presented at the workshop on ‘The Other Canon in Economics’. R. London. American Economic Review. Oxford University Press Chang. 2 Chang. B. (eds). Oslo. The Moral Dimension. mimeo. and Nolan. Markets. University of California. 1989. Castro. (ed. Routledge Chang. The Political Economy of Industrial Policy. J. Oxford. A. 4 Hirschman. Role of the State in Economic Change. Verso Buchanan. vol. The Free Press Evans. vol. and Chang. and Rowthorn. A. Norway Chang. H-J. in Arestis. R. Princeton University Press Hodgson. The World Bank Etzioni. R. Shifting Involvements. A. and many middle ways: some reflections on the institutional diversity of capitalism. 15–16 August. ‘Deconstructing Capitalism as a System’. 1986. Edward Elgar Chang. 5. Cambridge. P. D. The Flexible Economy. European Journal of Development Research. P. Economics and Institution. in Buchanan. University of Hertfordshire Jessop. London and Basingstoke. C. London and Basingstoke. Cornell University Press Block. Individualism and Economic Order. 2000. Macmillan Goodin.

Cornell University Press World Bank 1991. Reasons in Human Behaviour. Public Enterprise and Public Choice. Utopia and the State. in Chang. 1985. Institutions in Economics—The Old and the New Institutionalism. H-J. Cambridge. no. no. Krugman (1989. 1975. M. R. The Myth of the Powerless State. 36. Anarchy. Ithaca. trust. 1991. The Free Press Marglin. Basil Blackwell Pagano. The Limits of Freedom of Contract. Oxford. 1999. New York. Calculativeness. W. 1995. New York. H. Cambridge. Oxford. 31. 1985. Business Organisation and the Myth of the Market Economy. The Great Transformation. New York. and Manor. Oxford. Aldershot. 28–30 April. (eds) 1990. Helpman and P. Cambridge. 1979. J. vol. O. Public Choice. Macmillan Toye. Rediscovering Institutions—The Organizational Basis of Politics. Wither reform?—Ten years of the transition. World Development Report 1997. M. 1987. (eds). Oxford University Press Trebilcock. Cambridge MA. vol. in Dutt. E. Cambridge University Press Lucas. Journal of Political Economy. The myth of Anglo-Saxon capitalism: reconstructing the history of the American state. The Free Press Williamson. Oxford. Balliger Mueller. K. K. Free Market Economics – A Critical Appraisal.). R. Journal of Economic Issues. 1983. Organisations and markets. DC Stretton. Oxford University Press World Bank 1997. J. A. 453–86 Woo-Cumings. Capitalism. J. Polity Press Williamson. Clarendon Press McPherson. 3 March. 1990. 1974. Competitiveness and its predecessors—a 500-year cross-national perspective. New York. (eds). Public Goods. 1996. (eds). (eds) 1999. 98. Oxford University Press . 1991. no. Oxford University Press Lazonick. H. New Directions in Development Economics. Massachusetts. States or Markets?: Neo-liberalism and the Development of Policy Debate. 1984. The Golden Age of Capitalism. World Development Report 1991. paper presented at the Annual Bank Conference on Development Economics. Neoclassical Political Economy. K. Journal of Economic Perspectives. Unwin Paperbacks Simon. and Jameson. Alternative tactics and strategies in economic development. Cambridge. The Developmental State. Basil Blackwell Simon. 5. B. 1992. J. H. R. Basingstoke. and Rowthorn. U. O. 6th edition. Harvard University Press Vira. New York.01 CJE 26/5 pp. D. Cambridge. M. vol. 6. J. vol. MA. 1994. (ed. Basil Blackwell Polanyi. Limits of self-seeking: the role of morality in economic life. J. The political Coase theorem: identifying differences between neoclassical and critical institutionalism. in Colander. 1993. 1 Rutherford. 3. C. S. Washington. Beacon Press Reinert. Oxford. Socialism and Democracy. and Olsen. 1998. Oxford. Markets and Hierarchies. and economic organisation. 1993. Structural Change and Economic Dynamics. 1957. 539-559 FINAL 23/7/02 4:10 pm Page 559 Breaking the mould 559 Kozul-Wright. Cambridge University Press Schotter. Cambridge University Press Nozick. New York. Boston. 1995. Work and Welfare in Economic Theory. no. L. M. 761–79 Weiss. R. MIT Press). Role of the State in Economic Change. J. and Orchard. Is there a neo political economy of development? in Colclough. St Martin’s Press Schumpeter. J. and Schor. A. 2 Stiglitz. vol. 1989. Journal of Law and Economics. Review of Trade Policy and Market Structure by E. L. Edward Elgar Stiglitz. London. 1991. D.

539-559 FINAL 23/7/02 4:10 pm Page 560 .01 CJE 26/5 pp.