Understanding the Relationship between Institutions and Economic Development: Some Key Theoretical Issues

Date August 13, 2008
Speaker Ha-Joon CHANG(Reader, Political Economy of Development, Faculty of Economics, University of Cambridge)
Commentator Youn Hee CHOI( Visiting Scholar, RIETI)
Materials

Summary

Ha-Joon CHANGThis talk is about the framework for understanding the role of institutions in economic development. The issue of institutional development has recently come to prominence. In the 1980s, the World Bank and International Monetary Fund (IMF) were criticized for ignoring the economic roles of institutions, politics, and history; all of which were deemed by both organizations to be minor details. Times have changed though, and today even the World Bank and the IMF put great emphasis on institutional reform, to the extent that some people now worry that these organizations are interfering in areas where they do not have a political mandate.

Recently, a substantial amount of literature has been developed on the relationship between institutions and economic development, including relatively abstract theoretical work, historical studies of particular institutions, and econometric studies that attempt to measure the correlation between institutional quality and economic growth rates.

Some key theoretical issues have to be resolved in order to advance our understanding of institutions and economic development. Definitional problems must be addressed. One fundamental difficulty involved in the study of relationships between institutions and economic development is that, after years of debate, there remains no widely accepted definition of "institution."Thus, when the term is used its limitations must be kept in mind. This can make dialogue more difficult because people often mean different things when they use the term "institution."

At a very general level, we may say that there are certain functions that institutions have to serve if they are to promote economic development, and that there are certain forms of institutions that are best designed to carry out these functions. Most institutions, however, serve more than one function, and many institutions have functions that overlap with those of other institutions. What's more, the same function can be served by different institutions in different societies. This means that there is no straightforward way to map institutional functions with institutional forms.

Sometimes institutional forms and functions are confused, as when "governance" indexes mix up variables that capture the differences between the forms of institutions (e.g., the absence of state ownership) and their functions (e.g., respect for private property). As the principal interest in these institutions centers around what they do rather than in their particular forms, especially given the large number and diversity of institutional forms in existence, the correct basic approach is to prefer the function variables over the form variables. Still, the forms cannot be ignored and real life cases need to be examined very closely to identify the various kinds of different combinations that would otherwise never have been imagined.

The current orthodoxy, however, is overly fixated on particular forms; a view that is most clearly manifested in the literature on Global Standard Institutions. This is unfortunate, though, because the whole point of bringing institutions into the discourse was to modify the 1980s' one-size-fits-all approach, which has now shifted to another level. In the 1980s it was realized that sometimes policies did not work because they were built on assumptions about certain institutional configurations in a particular society. When attempts were made to apply those policies to other societies with different institutional configurations, they ended in failure. Contemporary orthodoxy holds that if policies do not work because of differences among institutions, then the institutions should be made the same. For example, if Brazil were to adopt Anglo-American institutions, then Anglo-American policies should work in Brazil.

In the orthodox literature, the institution that is considered by far the most important is the property rights system; the emphasis has been so great that Dani Rodrik once described this as "property rights reductionism." However, there are a number of problems in this area.

The first is the measurement problem. Some institutions, such as the central bank, are rather compact and well-defined. In some cases, even the building and the individuals are identifiable. In contrast, the property rights system is made up of a vast set of institutions including corporate law, contract law, inheritance law, tax law, land law, urban planning regulations, bankruptcy law, intellectual property rights law, and so on. Even if the composite elements of the property rights system could be identified, it would be impossible to aggregate them. This is why survey data are relied upon in empirical studies. Particularly popular are surveys of foreign businesspeople, yet these introduce certain biases. Moreover, the results of these surveys can be significantly affected by subjective considerations and should be used with great caution.

The second problem is the coverage problem. In the orthodox discourse on property rights, only three types of property rights are essentially recognized: pure private property, state ownership, and open access. In reality, there are many other forms of property rights, which include communal property rights and hybrid forms of property rights, such as the Chinese Township and Village Enterprises (TVEs).

The third problem is the assumption of the superiority of pure private property rights. In measurements of the quality of a country's property rights, there is a very clear bias against countries with a high proportion of state ownership, except for the cases of France and Singapore, where state-owned enterprises are well-managed.

The fourth problem is the notion that the stronger the protection of property rights, the better. It is a mistake to deem this relationship to be linear, despite regression analysis that suggests as much. More protection is not necessarily better. It could be the case, for example, that it is better to protect property rights strongly up to a certain point, beyond which protection could reduce social productivity. The current debate surrounding the granting of patents is relevant here: even though it is good to encourage invention by rewarding inventors with patent monopolies, if they are granted for too long they have a negative effect on society because further technological development can be blocked.

The weakest aspect of the orthodox literature on property rights is its stance on the phenomenon of institutional change. One flaw of mainstream theories is the assumption that once institutions are in place, they perpetuate certain patterns of human interaction. The institutions themselves are seen as being determined by immutable factors, or at least ones that are very difficult to change; such as the climate, resource endowments, and cultural traditions.

It is unproblematic to expect persistence from institutions insofar as they remain stable and provide stability to the patterns of human interaction. However, the trouble is that a country's institutional complex contains various elements and can be described in many different ways. A lot of people claim that Confucian culture is good for economic development. There are good reasons to think that certain aspects of Confucianism, such as its emphasis on education and hard work, encourage economic development. However, the negative aspects of Confucianism can yield arguments that are equally as effective. Until the 1950s, a lot of people claimed Confucianism was the reason why East Asia was failing to develop; pointing to the placement of merchants and artisans at the bottom of the traditional Confucian social hierarchy, and the stifled entrepreneurship in conformist societies. It is very difficult to define a particular cultural or institutional tradition, and for this reason explanations that rely on culture and institutions as the embodiments of cultural values can easily degenerate into ex post justifications.

There is almost always more than one tradition in a country's culture and institutions. For instance, between the Napoleonic era and World War II, France was arguably the most laissez-faire country in Europe, which belies its reputation for statism today. For those countries that have more than one tradition, deliberate political choices and the ideologies that influence those choices become important in determining the development path of the country.  

In the long run, cultures and institutions are not immutable, largely because cultural/institutional changes and economic developments influence one another. Mainstream theories hold that individuals are born with a predetermined preference, so causality always runs from culture and institutions to economic development. This view is too simplistic. Taking a broader view of institutions, they can be seen influencing individuals; so causality does run both ways. Economic development changes people: the Germans and the Japanese, for example, were not always the hardworking, efficient, cooperative and well-organized peoples that they are thought to be today. That a given nation holds to certain values does not mean that the people refuse to, or cannot develop economically. In short, attitudes change when there are changes in material conditions.

Culture and institutions change partly because of changes in material conditions, but also because of human agency. Fixed structures are privileged in mainstream theory, in which the role of human agency is overlooked: rational individuals have no choices to make and are merely the executors of the logic of the system. In reality, however, individuals make choices that are not totally determined by their objective interests. Consideration of human agency is necessary in order to understand how institutions change and how societies change as a result.

If institutions are taken as technologies for social management, there is a strong case for using Alexander Gerschenkron's catch-up framework in order to understand institutional development. Indeed, due to institutional imitation on the part of developing countries, their citizens are now enjoying higher standards of political democracy, human rights, and social development than did today's developed countries back when they were at similar levels of economic development. In nineteenth-century Europe, for example, democracy was widely regarded as a counter-productive idea and suffrage was not universal. Today, even the most anti-democratic military regimes would deem the introduction of a selective franchise to be unacceptable; either everyone can vote or nobody can.

Institutional imitation is rarely sufficient, as every country needs to adapt imported institutions and ideally, in the long run, add their own institutional innovations. There is no better example of this process than Japan in the Meiji era (1868-1912): when importing institutions from Europe, the authorities took a pick-and-mix approach based on what model they thought would be best for Japan (e.g., the Japanese modeled their navy on the British one, and their army on the Prussian one). Each of the selected models in this patchwork was adapted; for example, the Japanese initially implemented the American education system, but when it became clear that it was not working as well as had been hoped for, a shift was made to a hybrid of the French and German systems. Modifications such as these were made by the Japanese, who then added their own innovations, like lifetime employment and subcontracting networks.

This evolutionary dimension is missing from the mainstream literature, in which it is assumed that institutions from developed countries can simply be imported by less developed countries and made to work overnight.

Ha-Joon CHANG and Youn Hee CHOI

Questions and Answers

Q: Are there cases when the state is not an antagonist to the market, and should we discard the idea that statism spoils the market?

A: A German politician once remarked that "the market is too important to be left to the private sector," which is the attitude in a lot of continental European countries. This tradition is evident in China, too, where Deng Xiao Ping said, "It doesn't matter if a cat is black or white, so long as it catches mice." In China, whether an enterprise is a TVE, a state-owned enterprise, a joint venture, or a foreign firm has become less of an issue.

Q: What will be the next stage of your research?

A: The emphasis will continue to be on the time dimension. So much can be learned from history: countries have devised very unusual institutional combinations that one could never have imagined.

One theoretical preconception is that when a small group of families dominate an economy, inequality stems from that situation. Many Pakistanis assert that the "22 families" are blocking the development of their country. Sweden, though, has 15 dominant families, of which the Wallenberg family, which owns about 40 percent of the Swedish private sector, is the most powerful. In spite of this, Sweden has built one of the most egalitarian societies in the world. Another example is that of Denmark, with its Møller family and the A. P. Møller-Mærsk Group. Some Danes claim there is no clear distinction between that family and the state, yet Denmark is also one of the most egalitarian societies in the world.

Many people's lives have been ruined by supposedly brilliant ideas, such as Soviet central planning and the IMF's Structural Adjustment Programme that was cooked up by economists. Live experiments are to be avoided. Much can be learned from history because someone has already done the experiment for you, albeit not in controlled laboratory conditions, which means that it is not easy to draw conclusions.

In the long run, my plan is to develop a more generalized theory. Theoretical work, though, can only be elevated to another level when there is a lot of empirical material available to draw upon.

It is possible to generalize at some level on every subject. The important question is whether the level of generalization made is relevant for the question being addressed. Methodology should be judged according to the usefulness of that methodology in the context of the particular question that has been posed. One type of theory generalized at one level is not applicable to everything.

Q: What is the motive of your research?

A: My motive is to understand the world better. How can such institutional diversity be explained? Why are the interactions so complex? Only when these things are understood can meaningful changes be made; changes that serve the interests of the greatest number of people. Whether this is a defensible position from any particular theoretical point of view is not of concern to me.

Q: Have you identified a pattern of "import-adapt-innovate" for development?

A: This notion of importation, adaptation and innovation is not sequential; in reality, everything is going on at the same time. The configuration of importation, adaptation and innovation will keep changing. At the beginning, importation of foreign institutions predominates; there is a little bit of adaptation, but very little innovation. Over time, adaptation and innovation increase.

Q: What is your core definition of an institution?

A: The minimum definition: rules that regulate human interaction and also form individuals.

In the mainstream literature, institutions are seen only in terms of constraints, but there are two other aspects that need to be taken into account. The first aspect is the enabling dimension. A lot of institutions constrain individual behavior in order to enable the collective, and very often the individuals themselves, to do more things. Take traffic rules, for example. These rules constrain the way one can drive, but without them there would be gridlock.

The second aspect that ought to be emphasized is that institutions also "constitute" individuals, in that many individuals are at least partly the products of the institutions that have shaped them. In mainstream theory, institutions are seen as products of individual choices, but they can also shape individuals to such an extent that they can be said to constitute them.

Q: Is Masahiko Aoki's theory of institutional complementarity a useful method to disentangle values configuration?

A: Masahiko Aoki's theory of institutional complementarity, as a general idea, is a very important one, but before forming a strong view on this I need to undertake more empirical research. When looking at historical examples, so many unusual combinations can be seen. It is not clear how much we can theorize about this complementarity, except at the very general level undertaken by Aoki. Beyond that, we cannot yet say much.

Q: What do you think of Dani Rodrik's argument that we should study national trajectories and the specific factors that are limiting development?

A: National trajectory is important because the most relevant case for any country is its own. On the other hand, sometimes it helps to look at other cases because unusual combinations can be seen. An examination of the trajectory of only one's own country, whatever its historical scope, still leaves the trap of preconceptions about what is possible and what is not. A look at Sweden, for example, reveals that an egalitarian society is not necessarily incompatible with an economy dominated by large private sector companies. The case of Singapore shows that not everything needs to be owned by the private sector in order to have a free market economy. A study of France leaves no doubt that the boundary between the state and the private sector does not need to be, and should not be, rigid; nor should the state and the private sector be seen as dichotomous. There is great value in broadening one's horizons through comparative studies.

While I am not against Rodrik's point, the world is very complex and cannot be understood from just one kind of study. However good that study may be, it needs to be complemented. Econometrics can be very useful in terms of identifying broad patterns, but that is only the beginning of the solution to a problem. A lot of different approaches are needed, so one should not say that this or that approach is necessarily the best one.

Q: Could you give us some of your observations about rule of law?

A: The problem is with the substance. What does rule of law really mean? In different societies, law means different things and laws cover different areas. It is very often the case that debate is not about whether there has to be law or not, but rather about which areas should be regulated by law. It is doubtful that there is any society that says there should be no law. So the question is: which areas should be covered by law and in exactly what way?

Even with the existence of laws, there is the question of implementation. In the 1960s and 1970s Japan, and to a lesser extent Korea, were famous for "administrative guidance," which meant that laws were vague - sometimes they were designed to be vague and sometimes they were vague as a result of mistakes. It was left up to individual officials to interpret laws and provide more concrete guidance. In the Japan of that era, was there rule of law or not? Government officials at that time never violated any law; they merely interpreted the law to suit the particular situation. Is it better to have a law that specifies almost all contingencies? This is the real debate we should have. This abstract notion of rule of law is like human rights; what does "human rights" mean? The devil is in the details.

Q: What happens when monetary incentives are introduced?

A: My view on whether capitalist behavior is compatible with some cultures and not others is that any behavior is, in the long run, compatible with anything. The real world works in the way it does only because there are many other motives. If the monetary incentive was the only thing that mattered, why does one bother to read books written by Jack Welch and Bill Gates, when you could just hire McKinsey and Company to devise a rational scheme and implement it? If all problems could be solved using rational behavior models, the practice of management would be easy.

Q: In defining desired institutions only in relation to desired economic activities, in what ways and to what extent do you engage with the mainstream new institutional economics (NIE), and what are your main criticisms?

A: There is nothing in NIE I would fundamentally disagree with; it has shown us some very interesting things about one aspect of this very complex phenomenon, but it is clear that we need a broader canvas.

*This study was prepared within the UNU-WIDER project on Institutions for Economic Development: Theory, History, and Contemporary Experiences, directed by Ha-Joon Chang, and is reproduced with the kind permission of the author and UNU-WIDER.

*This summary was compiled by RIETI Editorial staff.