2. Concepts

I begin with a brief discussion of some conceptual issues regarding economic development and institutions.

On economic development, the ‘dominant discourse’ focuses on economic growth in empirical applications, but on efficiency (in terms of individual utility) in theoretical discussions, and Chang emphasizes economic growth. Although it is tempting to focus on single indicators such as economic growth (given that economic development in some broader sense is unlikely without sustained economic growth), there are sound arguments for adopting a broader perspective that incorporates not only growth and efficiency, but also income poverty and distribution, freedoms and rights of different kinds, participation in the life of communities, security and protection against undesirable fluctuations in conditions, functionings and capabilities in the sense of Amartya Sen (Reference Sen1999), especially in providing access to things that fulfill what societies consider to be basic needs, and the environment. Not only are these goals intrinsically important, but their achievement is likely to be interrelated in interesting and possibly different ways in different situations.

The term ‘institution’ has been defined in a variety of different ways, but for our purposes it is useful to define it as comprising of: (1) formal or written laws rules concerning what people and groups can and cannot do; (2) informal habits, social norms or conventions which affect how people and groups behave, especially with each other; and (3) more or less formal organizations of people or groups (this follows Hodgson, Reference Hodgson2006). Formal and written rules include statutes, bodies of case law and constitutions. Informal norms and conventions can be ingrained habits, though all habits are not norms and conventions, only those that are more or less widely shared. Organizations include the courts, the police, civil service bureaucracies, firms, trade unions, families and official international organizations. The need for combining these three aspects of society into one concept comes from the fact that it is impossible to understand each one and its role in the economy without understanding the others.

3. Theory

To examine how institutions affect the economy, I start with the theoretical underpinning of the dominant discourse. This approach can be viewed as drawing on the neoclassical general equilibrium model and the fundamental theorem of welfare economics. According to these, under certain conditions – including those relating to the absence of externalities, imperfect information, public goods and non-atomistic buyers and sellers – free market equilibrium outcomes, with self-interested maximizing agents, result in the full employment of resources and are Pareto efficient. However, this analysis requires also that people do not deprive others of their private property and that they fulfill contracts into which they enter. For instance, if property rights are not protected, economic agents will have little incentive to produce, exchange, save and invest, fearing that their gains from these activities will be appropriated by others. It follows that institutions that protect private property rights and enforce contracts and policies which remove government-imposed restrictions on the free operation of markets, given the other assumptions noted, result in socially desirable outcomes.

This approach suffers from several problems, however, which can be illustrated using the example of property rights. First, in moving from theory to policy regarding institutions, it is not clear what real-world institutions are required. Even if attention is confined to formal laws, there are many dimensions to property rights, including the right of access without subtractive benefits, the right to withdraw products, the right to manage and to improve, the right to exclude others from access and withdrawal and acceptable methods of so doing, and the right to sell exclusion rights; different configurations of these dimensions may have different effects depending on how they affect the degree of control rather than some definition of ownership of the appropriate decision makers (see Rodrik, Reference Rodrik2007). Second, the effects of institutions depend on how its three aspects interact. For property rights, formal rules are never completely specified but require interpretation by the courts, their compliance depends on both social norms and enforcement by organizations, and changes in rules may well result in changes in norms that reduce the control of relevant owners. Third, institutions can have different effects regarding different goals of development. Property rights strengthen a kind of rights, and their effect on efficiency and growth has been emphasized by the dominant discussion. But they also affect distribution and the rights to functionings such as being adequately nourished, which are in turn related to efficiency and growth. Finally, attempts to secure property rights and enforce contracts and to increase market freedom can have undesirable consequences because the real world is unlikely to satisfy the stringent conditions that market economies need to satisfy for the first welfare theorem to hold. Externalities, asymmetric information and various kinds of market power abound in many economies, especially in less-developed countries. Individuals may find it too costly to gather all relevant information and lack the computational ability to find optimal outcomes. In the presence of fundamental uncertainty (as opposed to objectively calculable risk) it is in fact impossible to even decide on what is ‘rational’ for actions involving future consequences. With these other departures from the assumptions of the model, it is by no means clear that changing some specific institution in some way will necessarily have desirable consequences.

It is therefore unsatisfactory to examine the implications of institutional change by relying on the role of institutions in the neoclassical general equilibrium model. Chang (Reference Chang2011) makes a plea for careful empirical analysis of the effects of institutions of different kinds, because they are likely to vary across place and time. While this is eminently sensible, a theoretical framework is also needed to know what aspects of the economy to look at, analyze mechanisms explaining the empirical observations and to generalize beyond specific cases. A number of institutional approaches are available (see Hodgson, Reference Hodgson1998). One is that of the ‘old’ institutional economists, such as Thorstein Veblen and John Commons, who provided a rich discussion of the different types of institutions and the role they have in the economy (in addition to an evolutionary theory of institutional change). However, a very sympathetic reading of this approach concedes that ‘there has been a failure of the “old” institutionalists to agree upon, let alone develop, a systematic theoretical core’ (Hodgson, Reference Hodgson1998: 168). Another is the ‘new’ institutionalist approach pioneered by Ronald Coase, Douglass North and Oliver Williamson, which invokes the concept of transactions costs and recognizes that there may be many problems with the economy even if property rights can be secured and contracts enforced, due to the existence of market imperfections such as externalities and imperfect information, and in some cases (see especially North's Nobel Prize lecture), deviations from ‘rational’ or optimizing behavior. Although this approach has provided many valuable insights on how institutions can affect various kinds of market ‘imperfections’, it is arguably not very specific about what exactly is involved in the portmanteau expression ‘transactions costs’ in a manner that can tell us about the effects of specific real-world institutions on development goals.

An alternative theoretical approach starts with a general framework involving accounting identities (such as income accounts and commodity balance relations) relating relevant flow and stock variables of the economy (including, for instance, the rate of growth of output, capital stock, and distributional variables), then introduces additional behavioral and equilibrating relations between these variables, choosing appropriate parameters, and then solves for the values of the variables to obtain an equilibrium solution (see Dutt, Reference Dutt1990). The aim of this is to show how the variables of interest in the model depend on the underlying structure of the economy in terms of equations embodying such things as the role and determinants of aggregate demand and supply forces, power relationships and policy parameters. The effects of institutional changes can be examined using a two-step procedure which involves, first, an analysis of how institutional changes affect the parameters of the model, and second, how these parametric changes affect the variables of the model. Not all economies will have the same relationships since in some aggregate demand may be more important in determining outcomes, while in others, particular aggregate supply forces may have a stronger influence. Such structural characteristics will have an impact on how institutional changes will affect growth and distribution and other variables. However, not all countries are sui generis; some generalizations based on suitable classifications may well be possible. Thus, making property rights more secure may increase saving rates, which may increase growth in economies constrained by the supply of capital, while it may have an adverse effect – due to the paradox of thrift – in demand-constrained economies. Institutional change which promotes labor market flexibility may reduce the wage share and increase growth in economies that have been called profit-led, while it may have the effect of reducing growth in output in employment in wage-led economies in which a reduction in aggregate demand due to the lower wage share reduces investment incentives.

4. Political economy

Institutions change slowly. Legislatures take time to change laws, constitutions require large majorities to change, and even though judges can change laws by setting precedents, this change follows lengthy legal processes. Habits, social norms and conventions exhibit rigidity, because of individual inertia and group reinforcement. Moreover, many norms may be rooted in religious and cultural values, which are likely to be slow to change.

Nevertheless, they do change over time due to forces endogenous to society and the economy. Formal laws depend on political processes, which at least in part reflect the interests of different groups and classes, and the relative ability of different groups and classes to influence political processes and outcomes, which in turn depends on their income and wealth (as stressed by Marx-inspired scholars) and on their numbers (as stressed by Mancur Olson) and on how they are organized. Norms and conventions depend on economic and other factors. For instance, norms regarding acceptable levels of corruption among government officials may depend on average income levels, and those regarding work habits are likely to depend on the effect of wages on motivations. These dynamics are complex and are difficult to capture in terms of law-like regularities, and seem less amenable to mathematical formalization than to careful, context-dependent political economy analysis.

These observations have several implications regarding policy. First, institutions may be difficult to change, not least because changes in one dimension of institutions (such as laws and organizations) can be negated by rigidities or even contradictory changes in other dimensions (such as norms). Second, despite this, they can be changed by policy. Laws may be changed by the stroke of a pen, and organizations can be created by the government fairly quickly. Even norms and conventions can be altered by government actions, for instance, as discussed by behavioral economists, by taking advantage of what is called the status quo effect. But such changes relate not only to changes in institutions, but also to other aspects of policy. All these changes, of course, respond to class and group pressures which are endogenous to the society and the economy, but the state, especially at certain junctures – when the democratically elected government has a strong majority, when (somewhat paradoxically) it has a weak majority and has little to worry about staying in power, when there are economic crises, if dominant classes have divisions within them which the state can use for dividing and ruling and due to international pressures – can be relatively more autonomous from societal factors. Third, prior institutional changes are not essential for improved development performance, since institutions are affected by other economic changes. There is, in fact, little evidence that major growth spurts are preceded by major institutional changes (Rodrik, Reference Rodrik2007). It is therefore unclear why institutions should be thought of as fundamental determinants of growth and development, and other changes in economic policy and behavior should be called proximate, as some growth empiricists do.