Economic policy for development and the new institutional economics

Motivation: Economists have increasingly emphasized the importance of institutions as a fundamental factor for economic development. However, too frequently institutional approach used in economics is simplified, both from the starting point in defining an institution and during the analysis itself. Aim: The aim of this article is to summarise the current state of knowledge on the use of the institutional perspective in the research on the economic development of developing countries. Materials and methods: The analysis is conducted from the perspective of the new institutional economics. The article uses the qualitative analysis method that includes a literature review and descriptive analysis. Results: The first part briefly describes the contribution of the new institutional economics to the research on the economic development. The second part demonstrates how the institutional perspective is used in the mainstream economics, while the third, how it is used in the development economics. Although the contribution of the new institutional economics to the development theory is significant, in practice, it is very difficult to provide clear guidelines for development policy. This is the reason why there is a huge difference between researching economic development in the spirit of the new institutional economics and how institutions are implemented in the mainstream and develop-


Introduction
The growing popularity of the new institutional economics resulted in the institutional approach gaining more and more proponents, consequently leading to an increasing influence on the mainstream economics. In the case of research on the economic development, the influence of the new institutional economics began to be noticeable when an increasing number of problems concerning the application of economic policy, which was the embodiment of the so-called Washington Consensus, came to light. Since then, there have been numerous publications devoted directly to the issue of the importance of institutions for the economic development. However, the too frequently used approach was (and still is) simplified, both from the starting point in defining an institution and during the anal-cby CATALLAXY Volume 5 Issue 2 December 2020 e-ISSN 2544-090X  www.catallaxy.pl Original article received: 10.11.2020Original article received: 10.11. / accepted: 15.12.2020Original article received: 10.11. / published online: 31.12.2020 ysis itself. This approach can be summarised as follows: (1) institutions are an important factor for the economic development; however, (2) institutions primarily constitute the formal rules created by the government; therefore, (3a) the state should actively work to stimulate the development processes, or (3b) the state will create property rights and ensure their enforcement, and the market will continue to function by itself. In this manner, institutions are a part of a relatively pointless discourse between the opponents and proponents of the government intervention in the market economy. It appears that ideological prejudices may pose an obstacle in the economic research. Economists and policymakers sometimes show a tendency to unilateral adoption of the development visions led by the state, other times they uncritically believe in market forces. Thus, when they try to incorporate institutions into their research, they make many simplifications and, therefore, their analysis has little in common with what the representatives of the new institutional economics write about the economic development. The aim of this article is to summarise the current state of knowledge on the use of the institutional perspective in the research on the economic development of developing countries. Section 2 briefly describes the contribution of the new institutional economics to the research on the economic development. Section 3 demonstrates how the institutional perspective is used in mainstream economics, while Section 4, how it is used in development economics. The analysis is conducted from the perspective of the new institutional economics. The article uses the qualitative analysis method that includes a literature review and descriptive analysis.

The new institutional economics and economic development
The new institutional economics seeks to incorporate institutions into economics either by creating a set of new theories from scratch or by complementing the existing ones, which is easier and more common (North, 1997, p. 17). Although there is still no complete institutional theory of development, which would fully explain the functioning of the economy, the already existing concepts allow to provide many answers that bother economists dealing with the problems of the economic underdevelopment. The contribution of the new institutional economics (NIE) to the analysis of the economic development can be divided into two main parts: (1) the grand theory of the role of institutions in development (macro-level institutional analysis, Section 2.1) and (2) the application of certain elements of NIE to analyse and solve specific problems of underdeveloped countries (micro-level institutional analysis, Section 2.2) (Toye, 1997, p. 56). It would be a mistake to state that either of these research perspectives is better suited to delving into the situation of developing countries. The macro-level institutional analysis, which focuses on explaining the manner and direction in which institutions evolve, makes it possible to clarify what factors shape the legal, political, and economic spheres. The subject of institutional research on a micro-level is the way that institutions ensure the effective operation of an organisation. Both approaches complement each other, providing a more complete picture of the real significance of institutions in economic systems.

The micro-level institutional analysis
At present, a great emphasis in development economics is placed on solving problems associated with the so-called new market failure. Institutions play a pivotal role in resolving issues such as informational problems, commitment problems as well as cooperation and coordination problems (Roland, 2016). Transaction cost economics dominates as a part of the micro-level institutional analysis. Transaction cost economics, explaining why some transactions take place and others do not as well as why agreements take a given form under certain conditions, provides knowledge that is particularly important from the view-point of the economies of underdeveloped countries. The reason for their underdevelopment to a great extent lies in the difficulty of concluding transactions (North, 2000, p. 37). At the same time, along with economic development, transaction costs are increasing 1 . The increase in the transaction costs is certainly a consequence of the trade development as well as the progressive specialisation and division of labour. In addition to the undeniable increase in productivity resulting from the economic development, carrying out transactions has also become more complicated, and consequently, more expensive. Thus, the costs of concluding and enforcing contracts have increased to enable impersonal transactions (Wallis & North, 1992, p. 122). Specialisation entails costs associated with information asymmetry.
Then how the paradox of high transaction costs in developed countries (calculated as the share of the transaction sector in the GDP) resulting from economic development as well as high transaction costs inhibiting the growth in less developed countries should be explained? Firstly, it should be noted that although the methodology for calculating these costs proposed by Wallis and North is one of the most popular ones, it is still an imperfect approach that does not fully encompass the concept of transaction costs. The trading sector estimated by Wallis and North covers only those transaction costs that are officially registered in economic statistics. This data does not include many of the costs that oc-1 This is confirmed by the empirical attempts to calculate their size. Wallis & North (1992, p. 121) estimated that the size of the transaction sector of the United States economy grew from a quarter of the GDP in 1870 to about half in 1970. Using a similar methodology, Dollery & Leong (1998, p. 225) calculated that in Australia this sector accounted for 32.3% of the GDP in 1911, and 1991, as much as 59.5%. On the other hand, Chobanov & Egbert (2007) estimated that as a result of the socioeconomic transformation, within six years, the transaction costs in Bulgaria increased from 37% of the GDP in 1997 to 52% in 2003. Similarly to the case of Bulgaria, in Poland, the creation of the market was accompanied by an increase in the transaction costs from 49.68% of the GDP in 1996 to 67.25% in 2002 (Sulejewicz & Graca-Gelert, 2009, p. 247). The situation was somewhat different in Argentina, where in the years 1930-1970, the share of the transaction sector was about 25% of the GDP and only in the 1980s, it reached 35% (Wang, 2003, p. 4). Measuring transaction costs is a big challenge for economists and there is still no standard methodology. cur but are not included directly in the profit and loss balance sheet of economic entities. This method enables, for example, observation of an increase in insurance costs resulting from a growing risk; however, it will not register the costs arising in situations when an entrepreneur has to spend resources on meeting many legal requirements of an inefficient bureaucratic system in order to start a new economic activity or conduct an existing one. It also does not register the costs of bribes, which in many developing countries are often necessary for conducting an economic activity 2 . Hence, using this method, the estimated transaction costs in Argentina are much lower than those in the United States.
Secondly, it is impossible to register costs if there is no transaction. As Coase (1992, p. 716) writes: "If the costs of making an exchange are greater than the gains which that exchange would bring, that exchange would not take place". There are no rules of behaviour that would reduce these costs to the necessary minimum, thus contributing to the extension of exchange and the deepening of the division of labour. According to North (2005, p. 120), developing countries are facing a great challenge: "To survive and grow in the context of the competition from the already developed world, they must deliberately construct an effective price system and supplement it by creating the institutions and organizations to integrate that knowledge at low costs of transacting. Standard economic theory is no help as a guide". Obviously, this is not an easy process. As noted by Shirley (2008, p. 20), next to or among institutions lowering transaction costs, there is also a need for those that will support the market by limiting the government and the strongest market actors so that they do not take advantage of their dominant position. A strong country has a monopoly on the use of force, thanks to which it can ensure stability, protect property rights and enforce contracts, and thus reduce transaction costs. However, the same country can use its position to expropriate property or oppress its citizens. The inclusion of transaction costs in the economic analysis is tantamount to rejecting Walras's view of markets. Hence, what is important in the analysis of the economies of developing countries, the market is not necessarily the optimal place to distribute all goods and services. Nonetheless, the existence of market failures does not imply that state intervention in the economy will always be effective. Governments are also imperfect and very often, through overly bureaucratic administration and opportunistic actions of officials and politicians, instead of eliminating them, they increase transaction costs. Thus, a question arises: is it the imperfect government that should correct defective markets, or perhaps the state should be limited by deregulation and liberalisation? Toye (1997, p. 58) stresses that the conclusions of the research on the new institutional economics do not allow a clear answer to this question. Each specific case requires separate analysis, which will enable the examination of whether a given economic policy reduces transaction costs or not. Hence, Coase's call to explore the world of positive transaction costs is still valid.
The analysis of transaction costs allows for the formulation of some guidelines for economic policy in developing countries. Firstly, research shows that economic activity is considerably easier in places, where small producer organisations are simpler to set up. Collective actions enable reaching more consumers and lead to the expansion of businesses and reduction of risk. The most common obstacles to cooperation are poorly protected property rights and excessive bureaucracy, which are the causes of opportunistic actions by some individuals. This situation discourages people from working together to achieve common goals. Secondly, under specific conditions prevailing in the markets of developing countries, alternative forms of contracts may turn out to be more effective than the traditional forms dominating in the West. This means, for example, that privatisation is not always the optimal solution, which will ensure the greatest efficiency of a given sector of the economy. Solutions, such as leasing for a part of agricultural crops or contracting production, guarantee security and stabilisation in countries with a high risk of conducting economic activities. Thirdly, information asymmetry is a significant problem that limits economic activity. There is a lack of institutions, which would enable both sellers and buyers to obtain the knowledge necessary to complete a transaction. Such a situation favours the gain of a dominant position on the market by stronger players (most often those having a closer relationship with bureaucrats), and consequently leads to an unfavourable redistribution of income and the disappearance of entrepreneurship among the poorest. Small agricultural producers or artisans cannot expand or even start a business because they are not aware, for example, that they could sell their product for a higher price or purchase the necessary capital goods cheaper. Easier access to information contributes to the formation of collective action organisations and the adoption of alternative forms of contracting. Greater transparency of the markets can in some cases also be achieved through partial deregulation and liberalisation. Fourthly, the low level of social capital is all too often the reason for high transaction costs in underdeveloped countries. Societies with a low level of social capital face high administrative costs due to the need to limit the negative effects of a low level of mutual trust between counterparties. Lack of trust is often the reason for limiting trade only to a group of relatives, a given ethnic group or family, i.e. a group of people with a closer social relationship (Araral, 2014;Bhatt & Tang, 1998, pp. 633-634;Buckley, 1989, p. 24;Dixit, 2003, pp. 130-131;Jack & Suri, 2014;Key et al., 2000, p. 258;Kherallah & Kirsten, 2002, pp. 25-35;Ouma et al., 2010;Sara & Newhouse, 1995, pp. 324-325;Staal et al., 1997, p. 792;Winter-Nelson & Temu, 2002, p. 572).

The theory of institutional changes
The macro-level institutional analysis is predominantly a theory of institutional changes. According to the NIE representatives, the key to understanding why some states are developing and others are not are institutions, and specifically the process of institutional changes. Based on the works of North (1990North ( , 2005 and Greif (2006), we can briefly describe the course of institutional changes as follows. The entire process is non-ergodic, i.e. historically conditioned, and unpredictable. The impetus for changes in the existing institutional matrix may be a change in the price relation, e.g., factors of production, costs of information or technology, or a change in the preferences of individuals. These sources of institutional transformation are both exogenous and endogenous; however, it is the internal factors, which reflect the actions of people trying to maximise benefits, that are the main reasons for the transformation of the existing system of institutions. The direction of institutional changes is a result of mutual interactions between the existing rules and people usually associated with organisations. Institutions shape people's actions, while organisations try to influence institutions to achieve the intended benefits. The created institutional system consists of both formal and informal institutions. The rules and formal principles created intentionally by people are based on informal institutions. The beliefs, customs and traditions that make up informal institutions are an essential part of the cognitive models of individuals. Since the informal rules included in customs, traditions and codes of conduct are relatively insensitive to intentional human actions, the entire institutional system is a product of these intentions and the inadvertent effect of the matrix transformation. The main cause of economic and social underdevelopment is the problem with the transformation of the system of institutions so that it can enable impersonal economic or political transactions to be carried out at relatively low transaction costs.
For the economy to develop, it needs an institutional environment, which will enable and support the activities of the market. Shirley (2005, p. 611) divides institutions forming this respective institutional structure into two not necessarily complementary sets: (1) those that facilitate market exchange by reducing transaction costs and increasing confidence, and (2) those that shape the system of state power in the direction of strengthening private property and individual freedom, and not in the direction of using and subordinating units of central power. For the process of economic development to be permanent, political institutions must transform along with the socio-economic evolution. As Greif (1993) demonstrated on the example of the development of trade in medieval Europe, the most successful were the merchants and states, in which traditional institutions, such as the bilateral reputation mechanism or trade norms regulated by social networks (religion), were replaced by increasingly complex ones, e.g., the commercial code. On the other hand, institutions supporting trade were supplemented by those that limited power and protected property. Rules and regulations supported transactions (by lowering transaction costs) and expanded the scope of transactions (more people could trade on a larger area), while institutions reducing the abuse of power and protecting property made people want to trade at all.
Since both types of institutions are found in developed countries, the exchange is significantly simplified, and economic activity becomes profitable. In the case of poor countries, the existing institutions supporting exchange are at a lower level of development, which makes trade difficult or even impossible. Even if there are rules created to resemble those found in rich countries, such as modern trade codes and organisations supporting the economy, there are still no institutions limiting the power and its abuse, which makes economic institutions limiting transaction costs ineffective. In developed countries, the progress of institutions supporting transactions was accompanied by the evo-lution of institutions regulating operations of the state. The rules and regulations existing in rich countries limit the abuse of power most often through the existing mechanisms of democratic control. These institutions are both formal (the constitution, judicial system, separation of powers) and informal (social responsibility norms). Setting up such a system requires time and is associated with high costs. Moreover, it never guarantees effective operation and avoidance of any problems. Even the most developed countries in the world are still struggling with the problem of abuse of power or the most powerful people, as exemplified by the recent economic crisis, which was the result of the failure of financial institutions. Nevertheless, in the long run, the cost-benefit balance is positive, enabling people to profit from their investment by having lower transaction costs and ensuring that their actions are not sabotaged by others (Shirley, 2008, pp. 22-24). North et al. (2009), attempted to present a theory of development based on the integrated use of the economic, political and social behaviour theories. Such an approach, being an expression of the consistent application of the assumptions of the institutional analysis, allowed for the development of an interesting theory of violence as a phenomenon combining the spheres of politics, economy, and interpersonal relations with social life, which exerts an overwhelming influence on all these spheres. The starting point of their theory is the assumption that solving the problem of violence is a fundamental condition for the occurrence of the economic development. In developing countries, individuals and organisations use or threaten to use violence to gain wealth and resources. The authors distinguished two types of social order: a limited access order -this is how societies of developing countries are organised, as well as an open access order, which exists in developed countries (North et al., 2009, p. 13). For a state to join the group of the richest countries and for the growth and economic development to be permanent, the order of limited access must transform into open access. The theo-retical concept of North, Wallis and Weingast was used to explain the causes of development or its lack by examining the cases of several developing and developed countries (Legiedz, 2019a(Legiedz, , 2019bNorth et al., 2013).

From Washington to post-Washington consensus
At present, economists and officials operating within international organisations, such as the World Bank or the International Monetary Fund, have a significant impact on the economic policy in the poorest countries and the general discourse of the economic development. The 1980s and 1990s were dominated by the approach focusing on state failures. The negative effects of the government intervention into the price system were emphasised, attention was paid to the phenomenon of rent-seeking, and the operation of state-owned enterprises was unequivocally negatively assessed. The development paradigm at the time was based on liberalisation, privatisation, and deregulation. Mainly through the operation of the International Monetary Fund and the World Bank, there were attempts to impose a policy of renouncing the state from direct control over the economy. The guidelines for neoliberal development policy are best known as the Washington Consensus. In 1989, during a conference on the development of Latin American countries, Williamson (1990) of the Institute of International Economics delivered a report, in which he included policy guidelines proposed by the Washington experts that were largely based on the principles of neoliberal economics.
The one-sidedness of the approach to development dominating in the 1990s was noticed when the reforms in Latin America did not bring the expected results, and problems arose in the transition countries. In many cases, the situation of the former socialist countries at the beginning of the transformations at the turn of the 1980s and 1990s resembled those of underdeveloped countries. Analysis of the experiences of these countries enabled the addition of many new aspects to the theory of development, especially that the type and effects of the undertaken reforms were remarkably diverse. The problems of the countries transitioning from planned to market economies, the Asian financial crisis as well as the situation in Latin America (e.g. Argentina) all resulted in greater attention being paid to numerous voices criticising the policy of the Washington Consensus (Rodrik, 1997;Stiglitz, 2002). Many development economists then began to point out the importance of institutions. Hence the attempt to improve the Washington Consensus, sometimes referred to as the "post-Washington Consensus", assumes that both the set of proposed actions and the intended aim should be more complex and dependent on the cultural and political context. The market is still considered as a dominant institution; however, at the same time, attention is paid to the importance of all factors determining its efficient functioning. The experience of developing countries around the world shows that the reforms of the 1980s and 1990s were too often carried out without appropriate facilities and institutional preparation. Rash privatisation sometimes became the source of several fortunes and millions of unemployed, and the deregulation of financial markets caused crises, in which global banking institutions made money, while ordinary citizens lost it. Consequently, the post-Washington Consensus predominantly focused on the preparation of a proper ground for future reforms (Stiglitz, 2005).
According to Rodrik (2007, p. 17), the supplemented Washington Consensus is a result of the economists becoming aware of three facts. Firstly, economic policy based solely on the market cannot be effective without deeper institutional transformations. Secondly, some regulation of the markets was considered necessary due to the increasing frequency of financial crises. Thirdly, the Washington Consensus lacked the social sphere. As Rodrik observes, the supplemented Washington Consensus is institutional in na-ture, which is consistent with the trends both in the mainstream and development economics. The World Bank (2005, p. 6) study, which summarises the conclusions of the development reforms of the 1990s, also emphasises the importance of institutions in the process of economic development (Cameron, 2004).
However, it should be noted that despite the institutional nature of the new consensus and frequent references to North, this approach has little to do with what economists associated with NIE wrote about economic development. A problem arises when, by the trend dominating in modern economics, the "institutions matter" hypothesis is empirically verified by the means of econometric models. The construction of econometric models requires the necessary simplification of socioeconomic phenomena, which is largely incompatible with the assumptions of the new institutional economics that attempts to study the real economy.
Also, just a confirmation of the hypothesis that institutions influence the economic growth and development does not mean that it is clear what must be done for the growth and development to actually occur. If better institutions ensure faster growth or a higher level of development, a question arises: what should be done to improve the quality of institutions? The empirical studies presented earlier do not provide any answers. Unfortunately, it can be observed that the operation of international organisations dealing with the problem of economic development comes down to recommending actions, which are supposed to improve the indicators "measuring" institutions that are used in empirical studies. For such studies, the World Bank (2020) also created other measures of institutions, such as the Ease of Doing Business Index or The Worldwide Governance Indicators (Kaufmann et al., 2009). The creation of such indicators can be useful in the analysis of the economic situation, especially that countries, which occupy the top of the "institutional" rankings are typically in a considerably better economic situation. Nevertheless, it cannot be concluded on this basis that a pol-icy, which would result in the improvement of these indices is a sufficient condition ensuring economic growth and development, nor that such a policy is a consequence of the NIE recommendations (Wallis, 2011, p. 591). Two NIE findings should be recalled. Firstly, the process of development is historically conditioned, and the change in the institutional system is slow. Secondly, the entire system is conditioned by the interaction between the formal and informal institutions, and institutional indicators predominantly present the state of formal institutions. Consequently, the mere improvement of certain components of formal institutions does not necessarily have a positive impact on the entire institutional system, especially in the short term.
However, how the World Bank and other international organisations utilise the institutional perspective in research concerning development is not merely a consequence of econometric analysis. As Wallis (2011, p. 591) rightly points out, political organisations that concern themselves with economic development expect simple prescriptions, which could be easily applicable to developing countries, not subtle nuances. When Washington bureaucrats asked the academic environment associated with NIE for guidelines on how to change institutions to ensure economic development, they did not get a simple and clear answer because such an answer did not and does not exist. In this situation, the officials began to pursue a policy aimed at creating "better" institutions, defining them in the simplest, easy to imagine way, i.e. as "better" formal rules. Hence, it can be concluded that while the Washington Consensus was a simplified and superficial interpretation of the neoclassical counter-revolution in the economics, the post-Washington Consensus is a simplified version of the institutional approach.
Institutions are also the key components of the concepts of Acemoglu et al. (2005;Acemoglu & Robinson, 2010, 2012, 2015, 2016. This theory is not directly embedded in the new institutional economics; however, it has recently become very popular, mainly thanks to the book by Acemoglu & Robinson (2012) titled Why nations fail. In this work, the authors introduced the concepts of inclusive and extractive institutions. Extractive institutions limit access to political and economic markets and, in the long run, hinder economic growth and development. While Acemoglu and Robinson believe that institutions are the key to understanding the causes of economic development, they use the concept of institutions in a rather narrow sense. Admittedly, they refer to the North's definition of the institution; however, one should agree with Dzionek-Kozłowska & Matera (2016, p. 4) that the authors focus on formal institutions and that the concept of informal institutions is not used in their work. Acemoglu & Robinson (2012) seem to downplay the importance of informal institutions, rejecting the "cultural hypothesis"; therefore, their concept seems closer to the attitude of Heydemann (2008), who criticises the exaggeration of culture as the determinant of economic growth and development 3 . The controversial component of the concept proposed by Acemoglu and Robinson is not only the underestimation of the importance of culture but also a clear suggestion of the domination of politics over economics.

Development economics and institutions
Development economics had a huge impact on the economic policy in developing countries until the 1980s when the pro-market concepts gained popularity in mainstream economics. One can speak of a clear crisis in the de-velopment economics at that time (Piasecki, 2007, pp. 18-19). There was no room for development policy in the free market doctrine. With the crisis of the Washington Consensus, the role of the development economics grew again. Its representatives never agreed with the market-oriented guidelines for the economic policy of developing countries. Development economists began to highlight that the reforms made so far had too often been carried out without appropriate facilities and institutional preparation. Among development researchers, the word "institutions", as well as the institutional perspective in general, quickly became very popular 4 . After all, also, in this case, the new institutional economics has little influence on how development economists integrate institutions into their research. Moreover, some development economists have raised objections to the new institutional economics.
While analysing the progressions of the Washington Consensus, Rodrik (2006), currently one of the most influential development economists, formulated warnings against excessive institutional fundamentalism, which would replace market fundamentalism. On the one hand, Rodrik's approach is characterised by an understanding that institutions are necessary for the proper functioning of the market, while on the other hand, by downplaying the actual difficulties associated with the process of emergence of institutions that favour economic development. Rodrik (2007, p. 51) cites North's definitions of institutions. He also points out that formal and informal institutions complement each other and categorically disagrees with the claim that there is a universal institutional system, which works effectively under all conditions. Concurrently, when he moves on to conclusions and attempts to formulate guidelines regarding economic policy, he assumes that institutions can be changed relatively easily if the local specific-ity is considered (Rodrik, 2007, p. 162). Based on welfare economics rather than institutional economics, he underestimates the problems associated with state limitations and creating institutions, which are pointed out by leading NIE representatives. It should be noted that a similar approach to institutional problems still dominates both in orthodox economics and development economics. Considering that institutions play an important role, economists mainly postulate greater state control over the economy; however, they forget that formal rules (i.e., those by which governments can influence the economy) are only one component of the entire institutional system.
An example of Rodrik's incorrect approach to the problem of institutions is the fact that he constantly highlights the non-market character of institutions in his works. According to him, all institutions that enable the operation of the economy are of non-market nature. Furthermore, the existence of legislators and a police force is necessary to establish these institutions (Rodrik, 2007, p. 155). "Markets require institutions because they are not self-creating, self-regulating, self-stabilising, or self-legitimising" (Rodrik, 2007, p. 156). Of course, markets never function in an institutional vacuum; however, the process of creating institutions does not come down to the work of the legislator and police force alone. Moreover, Rodrik (2007, pp. 156-161) distinguishes five types of institutions that support markets and thus contribute to the economic development: property rights, regulatory institutions, macroeconomic stabilisation institutions (fiscal and monetary), social insurance and conflict management institutions. Rodrik (2007, pp. 163-166) believes that there are two methods of creating these institutions. The first method, which can be observed on the example of the Washington Consensus, is the transfer of an institutional system from the outside as a proven technology. The second method is creating institutions locally, based on experience, local knowledge and experimentation. Both the proposed division of institutions and the presented methods of their creation prove that Rodrik's institutional approach has little to do with what economists associated with the new institutional economics have written about the problem of economic development. Rodrik is particularly interested in formal institutions and the creation of economic policy, which stimulates the process of economic development. He is not interested in the intricacies of the process of institutional change 5 .
A considerably more serious critique of the institutional perspective regarding development research was made by Chang (2011). In an article published in the Journal of Institutional Economics, he questioned the nature of the relationship between institutions and economic development. He claimed that it is not only institutional changes that stimulate development, but that development contributes even more to the emergence of better institutions. Chang also believes that institutional economists rashly assume that institutional liberalisation will ensure economic development as well as that the process of institutional change is much more complex than is commonly believed. In his opinion, the institutional discourse was dominated by two approaches. According to one of them, institutions can be easily changed, while the other assumes that it is virtually impossible to intentionally change institutions. According to Chang, both beliefs are incorrect, and their roots lie in the fact that institutional economists do not study the world as it is.
Chang's objections, however, make little reference to the new institutional economics. As Shirley (2011) writes, Chang created a "straw" opponent whose theories were simplified and some flaws of which were exaggerated to prove his point of view. In formulating his allegations, Chang did not refer to his opponents directly, but only provided a general outline of their views. Thus, even if he rightly points out the shortcomings of institutional analysis in certain areas, his argu-ments are generally of little value. As Wallis (2011, p. 592) correctly points out, the roots of Chang's criticism can be found in the fact that the new institutional economics is still unable to provide a complete theory of institutional changes. However, this does not justify the simplifications made by the Korean economist 6 . Chang's approach could be defended if one assumes that his criticism did not concern the representatives of the new institutional economics, but the post-Washington Consensus described earlier. Nevertheless, Chang did not make such a distinction in his article.

Conclusion
Although the contribution of the new institutional economics to the development theory is significant, in practice, it is very difficult to provide clear guidelines for development policy. Firstly, a complete institutional theory of economic development does not exist. Moreover, it seems that there is no possibility of creating such a complete, and at the same time, universal theory. Thus, one cannot formulate a universal prescription for development. Secondly, in nearly every case, the representatives of NIE highlight that the development process is historically conditioned, unpredictable, and relatively insensitive to intentional human actions. Hence, development policy will always have a limited significance for the economic development. Thirdly, the process of building formal institutions does not progress in a way often perceived by mainstream economists. It is remarkably difficult to find situations when the purpose of newly created rules is to improve the situation of all or most of the concerned parties. In practice, especially in developing countries, the introduced changes favour a specific group. Additionally, almost as often, the attempts to implement reforms that significantly undermine the status quo fail. Therefore, it is difficult to find accurate guidelines for an economic policy, which would ensure economic growth and develop-ment in underdeveloped countries. Guidelines do not go beyond generalities such as "history matters", development requires defined and enforced property rights, etc. We still know little about how to create appropriate institutional conditions, which positively impact economic development. We can even risk a hypothesis that we know that the intentional creation of such conditions is very difficult or even impossible. This is the reason why there is a huge difference between researching economic development in the spirit of the new institutional economics and how institutions are implemented in mainstream and development economics.