One More Time: What are Institutions and How Do They Change?

This paper defines institutions, presents their basic characteristics, and discusses forces for institutional change. It draws on a wealth of research and study by scholars in different fields and follows from the flagship 2002 World Development Report on this topic, including relevant new research since then that illuminates key issues. The definition of institutions includes rules and organizations (specifically rules guiding peoples? actions). The paper emphasizes the diversity of institutions that can effectively perform similar functions, the arguments for and against standardization of institutions across specific transactions or groups, and the implications of these choices. It highlights the relationship between informal and formal institutions. It discusses forces for institutional change influencing all economies, ranging from external shocks (whether they be economic or weather-related shocks or driven by wars) to internal dynamics, such as those following population growth or compositional changes, and those following technological innovations. Many of these forces have a two-way relationship with institutions: they are affected by them and, in turn, influence whether and when they change. A special section is devoted to institutional transplants and their effectiveness. The paper concludes with a discussion of whether and under what conditions institutional change can be externally directed and the lessons for aid donors.

This paper defines institutions, presents their basic characteristics, and discusses forces for institutional change. It draws on a wealth of research and study by scholars in different fields and follows from the flagship 2002 World Development Report on this topic, including relevant new research since then that illuminates key issues. The definition of institutions includes rules and organizations (specifically rules guiding peoples' actions). The paper emphasizes the diversity of institutions that can effectively perform similar functions, the arguments for and against standardization of institutions across specific transactions or groups, and the implications of these choices. It highlights the relationship between informal and formal institutions.
It discusses forces for institutional change influencing all economies, ranging from external shocks (whether they be economic or weather-related shocks or driven by wars) to internal dynamics, such as those following population growth or compositional changes, and those following technological innovations. Many of these forces have a two-way relationship with institutions: they are affected by them and, in turn, influence whether and when they change. A special section is devoted to institutional transplants and their effectiveness. The paper concludes with a discussion of whether and under what conditions institutional change can be externally directed and the lessons for aid donors. 2

What they are (redux!) and why markets need them
In 2002, the World Bank published its flagship publication, the World Development Report, with the title "Building Institutions for Markets", a report drawing on decades of research by scholars from the economics, legal, history and political science professions, and the experience and knowledge of practitioners around the world. It also assimilated the vast experience of the Bank in supporting institutional development over several decades. This paper, on institutions in economic development, draws on the principal messages of the WDR and representative pieces of related research that followed, to highlight some important lessons for donors supporting development.

Defining Institutions
What are they? Institutions are the rules and enforcement mechanisms that govern 1 economic, social and political interactions. Organizations or networks are a special type of institution: they have a set of rules guiding the interaction of the group and group members with each other and non-group members. While much of this note refers to institutions broadly, the focus is on markets where people buy and sell things and the institutions that influence transactions in markets. 2 1 Here "govern" reflects the dictionary definitions of (1) to exercise a directing or restraining influence over; (2) guide; (3) control or hold in check. The Oxford Dictionary definition: to control, influence or regulate a person, action or course of events. 2 In a later paper, (Hodgson, 2006) also draws together the views of scholars (North, Weingast, Greif, Hayek, Veblen, Williamson, and others) to provide a unified perspective. Not all scholars have treated organizations in their definition of institutions, viewing them as only "actors". Hodgson explains differences among the views. By treating organizations in the category of institutions, some may argue that nations should be seen as large organizations and thus institutions too, as are supranational organizations, such as the WTO. While, this may be true, this note will not deal with the nation state as an organization, but rather, its constituent institutions.
Formal institutions focus on the third incentive and thus, can be ineffective if this economic sanction is weak. Norm-based institutions can supplement or supplant laws and formal rules. They may substitute for formal institutions where the latter do not exist or are not accessible or where they fail to facilitate business transactions. In these cases, informal institutions allow those sharing norms or culture to behave predictably, lowering the risks in a transaction (chapter 1). Corrupt environments, for example, are often the result of ineffective formal institutions that coexist with weak social deterrence, sometimes called a "culture of corruption"(chapter 5).5 In such situations incentives for corruption rise as peers also become corrupt, leading to a vicious cycle of socially undesirable behavior. For geographically isolated and poor market participants, formal institutions are not easily accessible. These groups are more likely to use informal mechanisms to improve information flows and enforce contractual arrangements. For much of the world's poor, informal institutions play a primary role in making business easier.But informal mechanisms are used not only by those in poorer environments. Social networks grounded in class, caste, tribe, and neighborhoods-as well as school background and membership in clubs-can be as important for cementing deals in corporate towers as in rural markets. In these markets informal institutions tend to complement formal ones. 4

Formal institutions are needed as markets become complex-and partly replace informal ones.
Peoples' actions are governed simultaneously by formal and informal institutions. The more complex that transactions become, the more separated in time and space, and the greater the share of transactions between counterparts that are not known to each other through social or business networks, the higher will be the relative importance of formal institutions with third party enforcement mechanisms (see Box 2). For example, competitive, small scale, village markets where you buy daily provisions with cash have little need for formal rules to regulate behavior. Problems encountered with this simple transaction (e.g. poor quality or use of counterfeit money, for example) are generally dealt with through informal codes of conduct. However, engaging a financial institution or advisor to support your retirement accounts, or exporting goods to a country where the firm cannot use its connections to gain customers, will require more formal rules to facilitate transactions. An example of how formal institutions may replace informal ones is when court enforcement of contracts is preferred towards the use of arbitration through community leaders. It is easier to hold a firm to account through a court than through a community enforcement mechanism when the transaction relationship is "arms-length" that is, outside the community. The WDR 2002 provides many examples of how increasing complexity in transactions (as described above) raises the demand for more formal institutions to complement or substitute for more informal ones.

Formal and informal institutions also work together. Examples of formal and informal
institutions complementing each other are many. For example, (i) when formal laws and regulations regulate property transfer, but, in cases of conflict, adjudicated decision rules favor one party over another due to social expectations as determined by religious or traditional practices; (ii) when public services, such as electricity connections are only delivered once informal systems of payment (a system of baksheesh or bribes) are added to the regulated formal system or when connections are prioritized 6 Following the institutional economics literature, institutions are needed to reduce the transactions costs of participating in markets. 8 They do this by performing three important functions: (a) regulating information asymmetries (who knows what); (b) defining property rights; property rights is used in a broad sense here. It means the "right" to use or own assets, rights to income, (or other gains from a transaction) or the right to take an action that will constrain another's gain from a transaction. In this sense, if the law allows the factory owner to pollute a river, it gives him a property right from which he gains, but constrains the value of the water to those who fish in or otherwise use the river; (c) determining the degree of competition (or its converse-cooperation) in any activity. An institution may simultaneously do more than one function. For example, by directing information flows, they can affect the degree of competition. 9 A corollary of the above, is that these rules provide structure to societal interactions, for example, by influencing peoples' incentives and ability to behave a certain way.

Information Asymmetries
Whenever a party to a transaction knows something relevant to the transaction that another party does not, there is an information asymmetry. For example, when you take out a loan, you have the most information on your assets and your willingness to repay the loan. The bank or moneylender uses the information at his/her disposal (credit registries and bureaus, your income level, title to your assets etc.,) to make an informed decision about your creditworthiness and whether to lend. There is asymmetric information in this transaction. At the same time, when you deposit your savings in the bank, you do not know whether the bank is truly safe or likely to fail. When you buy stock in a company, you do not have all the information on the company's financial health. You make an informed judgement based on available information. Institutions determine how much information is available to each party: accounting standards, regulations on auditing, publication of accounts, credit bureaus and 7 registries, property ownership registration requirements, the bank's reputation, news media, and so on.
When you buy an airplane ticket, you do not know how safe the airplane is to fly or what rights you have if the airline cancels your ticket. The government has regulations requiring airlines to provide this information. When one purchases a toy, clothing or food, the seller knows what are the constituent elements and the quality of the good or service. Today, government regulations and consumer preferences expressed in buying trends, govern the degree of information provided. News producing and distributing institutions are very important in this regard (WB 2002, WB 2008. Freedom of information laws govern the information asymmetry existing between people and the public sector. By influencing the stock and flow of information available to people, institutions influence interactions.
Clearly, economic outcomes are affected as institutions are used to restrict or to enhance information flows.

Property Rights
Laws, regulations and customs of all kinds determine who has what "ownership" rights in markets. Some common examples might be the right or title to a piece of land, or a car, the right to pollute a river, the rights of workers versus equity holders in a company, 10 the right to build commercial property on a piece of land, or the government's right to your property through taxation or direct appropriation. Without formal and informal rules that determine the limits of individual ownership/rights to gain from a particular economic transaction, there could be no community as each would aim to maximize his/her own self-interest. 11 Government's main role is to capture resources 8 from individuals/businesses in order to use/redistribute them for the benefit of the economy and society. Thus, rules delineating government responsibilities and rights over resources and the mechanisms by which it enforces or conducts these responsibilities (such as taxes and customs duties) fall in this category. 12 Degrees of Competition (and therefore, its flip-side, cooperation) Institutions also define the degree of competition in markets. For example, predatory pricing, capture of market share and mergers and acquisitions that give firms monopoly power in markets are actions governed by institutions. Institutions may granting a company monopoly rights to explore national oil resources, or may require competition. Collusion among firms (or people or public officials in their professional capacity) is also regulated. Banks face rules against collusion in setting lending rates. 13 Use of common property is regulated by various rules as is evident from anyone visiting public parks, rivers or the oceans so that communities can collectively benefit from a resource, rather than engaging in potentially destructive (and in the long run, self-defeating) competition aimed at maximizing own gain relative to others'. Rules may be designed to either prevent or facilitate competition.

What do good institutions look like?
Each set of institutions, has specific functions with the design varying from community to community, though their ultimate objectives and functions may be similar. The specific design or form of the institution that is effective in a given place depends upon societal preferences (historical and current), and political and economic conditions. Accepting that one size does not fit all, means that there is no "best practice", and the form an institution takes matters only to the extent that it is effective in achieving the intended function. As soon as entities from different groups (communities, 9 states, countries) engage in transactions there is a question about whose institutions (or which mix of institutions) are relevant for the transaction. Sometimes, altogether new institutions are created for specific types of interactions. For example, there are international arbitration courts or the Basel financial standards. In order to simplify transacting, some argue for standardization in institutional design. All other things equal, it is easier to do business in a country if its legal system and customs are similar to one's own. Proponents may argue for a "standard" institutional design to be followed everywhere. Increasing cross border flows of finance and knowledge have called for rule standardization in various areas, for example in the design or strength of Intellectual Property Regimes. Trade agreements of various kinds aim to standardize at-the-border as well as behind-the-border rules on goods and services production and processes. Another area where calls for regulatory standardization have increased is in cross-border banking regulations; clearly applying the same financial regulation may produce better regulation in many cases or for certain kinds of transactions. However, countries differ in their capacities to supervise, in the nature of their financial systems, in the risks they are exposed to and other factors that will support more diversity in regulation.
Standardization can carry benefits in transactions between groups, but it is not necessary to standardize in order to promote market transactions, or a group's welfare. 14 As discussed above, institutions that are "best practice" for one context, may not be so in others and the acceptable standardized solution, may be "best" for no party. 15 In addition, other barriers may be far more important than institutional standardization in promoting efficient transactions. For example, labor mobility may be hindered by language, education, growth in output, or housing scarcity rather than labor market institutions.
Countries have advanced to the technological frontier with institutional diversity within themas witnessed by the experience of the USA with its constituent states. The institutional differences of the states within the federal union reflect their varied histories and endowments (rural, agricultural America versus high-tech America for example). This diversity has been critical in ensuring federal stability and economic growth and prosperity for decades. For example, the US unemployment support program today allows for substantial differences in state rules and eligibility requirements (Kirkegaard and Posen, 2018). Other countries, such as China and India, to pick two large ones, may similarly thrive with institutional diversity. The EU countries have a common goal of economic prosperity and stability, yet Then, can rules adopted from other contexts work in a given country? Generally, to be effective, their design or form will need to be modified because one size does not fit all. Direct transplants can often fail. 17 Changes in legal structure that are adopted with a view to modernizing economies need to 16 Uber is facing much more stringent regulations in Europe than it faces in the USA today. complement existing conditions in a country. For example, the Federal Trade Commission's Antitrust Division uses a complex regulatory interface and rules to determine anti-competitive behavior. This institutional framework may not be the best-suited to enhance competition in a small, open, developing country with limited government capacity, less sophisticated economic structure, lower literacy, higher initial mis-governance (including corruption), different societal rules and political histories, or widespread poverty. As another example, an excessive emphasis on written procedures in courts and requiring representation by fully trained lawyers will limit access to justice in poor countries. Since one size does not fit all, institutions doing similar things look different across groups, will look different, or have different forms.
In conclusion, as long as institutions perform their functions well, the objectives and the rules are known to all parties and predictably enforced, diversity in form is consistent with market development and growth, and even desirable. Even when rules are standardized, for example to facilitate trade, there will be a constant push for change and innovation because countries develop at different rates and from different initial conditions. Forcing all groups to adopt the same rules may hinder institutional innovation in a dynamic context; how will better ways to regulate an activity be found if there is no change or experimentation with which to compare the status quo? To some extent, there will always be some tension between the two forces of standardization and innovation. The WDR 2002 provides an in-depth discussion of these issues.

Policies and Institutions
Very often, the terms "policies" and "institutions" are used interchangeably and sometimes together, without there being clarity on their differences. For the WDR 2002, and in this note, policies are defined as goals or desired outcomes/directions a country wants to take and institutions the rules

Introduction Two
As discussed in Part One, institutions exhibit substantial diversity in form, even though they may support similar transactions in different countries. Part One also discusses the complex, two-way relationship between economic, social and political factors on the one hand, and institutions, on the other.
As a result of these influences, a defining characteristic of institutions is their persistence over time. They generally tend to change slowly. For example, it takes a long while to design a new law, have community agreement on it, change it formally, and enforce it effectively. Related to the idea of persistence is the concept of threshold effects or tipping points. For example, institutions governing social benefits for the poor may only change when inequality or poverty rise above a certain level at which the peoples' support for change is stronger than is their support for maintenance of the status quo. Other threshold effects may be political-such as the degree of political repression in a country, or economic, such as the magnitude of a price change or change in information. The implication of a threshold effect is that people will seek change only at intervals or be able to effect change only at (sometimes long) intervals. Another fact of institutional change is that it usually requires collective action of some sort even if the initial impetus for change is due to individual initiative. The rest of the note discusses a variety of factors that affect institutions and promote change. The forces for change discussed below are (a) not necessarily independent of each other implying that (b) more than one force may be present simultaneously and (c) many are themselves influenced by existing institutions: for example, existing institutions may prevent or discourage innovation, which is the first force for change discussed below. 21

Institutional Change Factors
Change factors are those events or developments that change the distribution and size of economic gains (or losses) or expectations of such for individuals and communities. 22 These factors may influence government, political leaders, business leaders/groups or households/communities to lead or support institutional change.

Innovation
Discoveries -new products, new ways of producing etc., bring economic gains to innovators and others in society. With each such invention, new rules demanded by inventors and producers, consumers, government or others in society, develop to oversee the use and spread of the invention. Inventors usually wish to capture as much as possible of the gains associated with their discoveries or creations. 23 It is this desire, combined with the desire of rulers to support innovation and capture some of the gains associated with it that, historically, led to the development of patent law in its various forms. As Abraham Lincoln once put it, "The Patent System added the fuel of interest to the fire of genius." 24 The first such law was adopted in 500 BCE where chefs in Sybaris, Greece had the monopolized right (and associated profits) to produce a unique dish they had created. The Republic of Venice was the first to grant a patent for a technological invention in 1416. 25 Over time, in different contexts, and as innovations proliferated, different variations of these laws were adopted, reflecting the preferences and situations of countries.
Even in the United States, which receives the second largest number of patents per year, the value and "philosophical basis" underpinning the patent system has been debated since its inception, leading to various changes in the system over time. 26 While inventors have sought to guard their profits, societal preferences reflected in public policy, have also effected change. However, many other types of legislation, not just patent law are affected by innovation: rules on drug testing and provision, rules on electricity pricing and provision, regulations restricting materials that can be used in home construction or airplane construction, regulations in financial markets, e-commerce and e-banking laws and so on. 23 By inventors, I am referring to the person or persons that discover a thing and also develop it for commercial use. 24 https://ladas.com/education-center/a-brief-history-of-the-patent-law-of-the-united-states-2/ 25 At least according to Western history. https://onlinellm.usc.edu/resources/infographics/history-of-patent-law/ 26 https://ladas.com/education-center/a-brief-history-of-the-patent-law-of-the-united-states-2/

Economic shocks and crises
Economic shocks and crises can lead countries to undertake institutional change by dramatically changing the incentives and endowments of those with existing political or economic power. Both Whether or not crises have an impact on institutional development depends to a large extent on how "deep' they are, depth being a function of both the magnitude of the economic and social change, and its duration. Cohen-Sotten and Vallee (2018) discuss how the Great Depression in the wake of the banking crisis in the US, led to the Federal Reserve being transformed. While the Great Depression provided the impetus for a substantial transformation, it takes time, and sometimes, experimentation to get the "right" system, Gelperon and Veron (2018) recount how the Federal Reserve and the federal deposit insurance institution required repeated adaptation and even "radical reconstruction" at times to account for "unforeseen problems and account for market and political pressures and arbitrage". Koo and Kiser (2001) find that Korean labor markets were not flexible at the onset of the Asian financial crisis, but they became more flexible and efficient due to legal and institutional changes.
Institutional changes wrought during and in the wake of economic shocks may or may not have durable and broad effects. The impact will depend on the depth and duration of the shock, as these will determine the magnitude of the impact on economic, social or political gains. Institutions adopted in response to crises will tend to be adjusted over time.

Domestic political and social turmoil/ Change in leadership
Throughout history, large institutional changes have been preceded by indigenous social and political turmoil and/or the emergence of "leaders". The communist, and then "transition" countries of Eastern Europe and Asia are examples of countries that have undergone this type of institutional transformation. The move into communism and subsequent move out of it were essentially social and political responses to economic and political inequality. In other words, a threshold or tipping point had been reached and people were ready for change. At the same time, "leaders" were available to organize the movement and establish vast policy and institutional changes. On a more recent note, Brexit reflects a threshold effect in social and economic effects reflecting changes that had been occurring over a long period of time. Disaffection with income and job status over the years enhanced support for change in trading and immigration rules among others. Knowledge and information flow between countries affects these movements-by either moving people to form coalitions in support of or against similar movements in their own countries.
Sometimes, countries get rulers whose goal is to create their personal futures through change; and sometimes they get visionary leaders for their country, who are benevolent, and charismatic. The rise of certain rulers is sometimes obviously rooted in social and economic distress, and sometimes the reasons are not so obvious (though history books may make it look like they were all obvious occurrences).
Leaders are not necessarily in public office or in political positions. In any case, often leaders emerge that change how people think and organize collectively and therefore have an impact on institutions. Various "leaders" have sought to "modernize", "socialize", "Westernize", establish religious governing institutions or establish more secular institutional frameworks, among other things.
Even when domestic economic, social and political conditions are such that wide-ranging institutional changes are imminent and desired by citizens, the pace at which these occur and the degree of simultaneity of the changes have varied substantially. For example, countries in Eastern Europe chose different transition paths to market economies (Wolf, 1999). Poland's process was a big-bang approach to institutional reform: along with price liberalization, several rules were changed at once, with the philosophy that "radical" reforms created "safeguards" so that the process became irreversible (Balcerowiz, 1993). Proponents of gradual reform contended that because institutions needed time to be established, "effective" reform tended to lag behind de jure legal reform strengthening the case for a stepwise, rather than radical approach. For example, the legal abolition of entry restrictions for new firms (Wolf, 1999) would not increase competition if new firms could not enter due to lack of bank credit, effective protections, competitive distribution networks and so on. The Czech Republic, Estonia and Poland with fast privatization and liberalization are identified as more radical reformers; Bulgaria, Moldova and Romania as slower reformers. There are even slower reformers among the transition economies of Eastern Europe and those where change has been so slow that it has yielded no positive results.

Many institutional changes occur in response to domestic social evolution. These forces develop over time, though their manifestation may be perceived as "sudden". These may be characterized as taking place when a tipping point is reached, for example, a degree of inequality, or poverty, religious repression or political repression, which when reached, generate forces for change. Once this point is reached, institutional change will happen; it may be either all-encompassing and fast or step-by-step
and slower.

Knowledge, information (transparency), demonstration effects, and preferences
People's preferences are influenced by knowledge and information-and these changes in preferences are another source of institutional change. Sharing knowledge with stakeholders, is often referred to as increasing transparency. 27 When people have more information about the effects of particular institutional states, their support for a given structure changes. 28 For example, knowledge that unregulated food labeling can lead to mislabeling of food content has induced consumer groups to demand better regulation of food (and drugs) labeling. Knowledge that countries with a better regulatory framework will be likely to receive more foreign investment has led some countries to change their regulations. Knowledge that producers have mistreated workers or used child labor has led to calls for more regulation of labor markets and restraints on employers. Knowledge about the effects of climate change is slowly encouraging countries to adopt laws and regulations governing carbon emission and citizens to support them.
There is a large literature on how "demonstration" produces change, including institutional change. Peoples' preferences and behavior may be influenced when there is demonstration that a specific set of institutions has achieved desirable outcomes in another context. Demonstration effects have influenced the behavior and expectations of women around the world and (admittedly, after a long period of time) has raised support for regulatory change supporting women's role in many spheres. Showing how it can be done is as important as showing that it works. Bangladesh's Grameen Bank is a good example.
As it demonstrated its methods and successes in bringing poor women into financial activities, demand for similar ventures developed in other countries, including parts of the USA. Lankes et al (1999) considering institutional transformation of transition economies, contend that demonstrating successful processes and institutions to support corporate restructuring was important in motivating countries to adapt. Abir (1996) discusses how the success of the United States in curbing anti-competitive behavior led the Republic of Korea to adopt antitrust laws similar to those of the United States. However, he also notes that while consistent enforcement of antitrust laws in the United States has been effective in breaking existing monopolies and preventing their creation, Korea's preference was to have a selective approach to enforcement of these antitrust laws in order to facilitate the growth of Korean industries and increase their international competitiveness.
An important source of all knowledge is the media industry (Djankov et al., 2002, Islam 2006, Islam 2008). Therefore, not only are they important players in effecting institutional change, but also, industry and regulatory structure in the media industry are of critical importance for how subsequent institutions develop. 29 Technological innovation has created the need for new institutions to support/regulate the media industry; regulating the industry in the digital age is complicated. For example, whether Facebook is accepted as simply a "platform" or whether it is considered a news provider, will determine how it should be regulated. In this sphere as in others, innovation constantly puts pressure on institutions. For example, even while certain countries restrict access to certain types of news, new technology and knowledge allows circumvention of regulation and fosters competition in the news market.

Conflict and conquest: Forced transplants
History is replete with instances of conquests, following which, new rulers have imposed new rules on the conquered. Several historical studies (legal, economic, political or social) examine the impact of transplants by conquest. Some of these insights are summarized later in the paper. Many of the new institutions are direct transplants of existing institutions in the home country of the colonizer, while others are unique to the conquered country and aimed at regulating transactions there. British colonial rule in Asia, America and Africa, Spanish and French rule in South America and Africa, Japanese rule in Korea, Russian rule in much of Asia and Europe, reconstruction after World Wars I and II, Ottoman conquests, other Islamic and Christian conquests, Jewish settlements… the list is seemingly endless. What is striking is that institutional frameworks brought in by the same ruler after conquest, have since developed very differently across countries, their paths being influenced by existing conditions at the time, the degree of conversion by rulers, economic developments and other factors.
Conquests often involve quite radical institutional changes. In a study that has relevance for many conquests, Acemoglu et al (2011)

Often these rules were "direct transplants"-replicas of laws in the rulers' countries applied to the conquered community and usually modified over time. When new rulers systematically erode the old distribution of power reflected in property rights, or set conditions in motion that will do so over time, the new institutions tend to be more effective and durable. All countries today, retain aspects of their historical conquests in their institutional frameworks, reflecting the persistence of institutions.
21

Competition (or cooperation)
The desire to out-do others provides a strong motivation to support institutional change (see below on trade). A good example of the impact of competition on institutional change is the improvements in institutions wrought by governments seeking to increase a country's performance on global rankings. (e.g., Worldwide Governance Indicators, Doing Business Indicators). Competition among local governments to get more resources from the federal government has been a motivation for improving institutions at subnational levels. Competition to retain skilled labor has led to change in institutions at the firm and country levels. The flip-side of competition is cooperation. Institutions supporting cooperation may evolve when groups (or individuals) are exposed to competition and they find that they can capture gains that are greater or more sustainable, by cooperation (thus the rise of oligopolies), and the cost of cooperation is low. 32 Usually, increasing scarcity of a resource, rents or profits will lead to institutional change.

Trade in goods, services, information, capital
Trade between communities has been a strong catalyst for institutional change. Sometimes institutional change is encouraged by the possibility of doing more trade and sometimes trading erodes the effectiveness of old institutions (for example, by creating new sources of income or wealth or new knowledge and preferences). The pressure for change may come from any of a variety of interest groups.
Businesses seeking to reduce risk and raise profits will support the development of financial markets, or corporate law, or producers seeking foreign direct investment will support legislation making FDI easier.
Change may be promoted by governments seeking to protect the national (or their) interest. Trade in information (know-how, discoveries) influences institutional design, trade in technology accompanied by rules governing its use (e.g., IPRs) being an obvious example. The EU custom union's rules were developed to encourage trade; and in turn the formation of the union spawned a further set of legislative initiatives aiming to foster collaboration among the member nations. Examples are common product and process standards, regulations on food safety, and the use of additives in food, to name a few.
It is not only institutions that govern cross-border movement that are changed. For example, the export of pharmaceuticals requires both exporting and importing countries to consider implications for domestic consumers and accordingly, to consider new legislation protecting their interests. Apart from the obvious effect of trade agreements on labor, environmental, and patent laws (among other things), the simple act of engaging in commerce with partners separated across space, time, language and abilities-to name a few factors-creates pressure for institutional change. To support firms to enter global value chains governments have their own legislative agendas on the domestic front. Most recently, cross border operations of multinationals have spurred governments to devise new legislation aimed at maximizing own-country benefits (for example, tax legislation), whereas, the information industry's expansion-for example, from their US bases to China or the EU-have launched a flurry of activity regarding new legislation and regulation governing privacy or data-sharing. To take a historical example, Acemoglu et al (2002), document the substantial institutional changes catalyzed by international trade, that occurred in Western Europe during three centuries (1500-1850). World Bank (2002) discusses this factor of change in depth.

Population: Growth and demographic change
As human populations grow and fall, and change in composition, policies and institutions will change (McMillan et al 2016, Cincotta andEngelman, 1997). 33 Total changes in population will affect the strain on resources and exacerbate conflicts or require cooperative behavior. Access and rights to property are affected. Population composition also affects preferences. As an obvious example, older populations will tend to demand different services to those demanded by younger ones (such as better and more extensive child care versus cheaper elder care, or health insurance for the elderly). Policies and institutions will generally tend to change gradually as this happens. Similarly, changes in ethnic composition may lead to new legislation that address ethnicity-specific issues (for example, whether businesses must provide bi-lingual services or not, or anti-discrimination legislation, or legislation affecting temporary foreign workers). As populations increase, resources become scarcer per person. At some point, people will demand change in the rules overseeing the management and use of these resources. Natural resources are a good case in point: increasingly scarce water resources will cause conflict, competition for resources or cooperation. Immigration/migration is an important source of population size and composition and will be a strong force for institutional change.

Changes in population size and in composition (demographic, ethnic, religious, education-related) will affect the relative supply of and/or demand for change in the rules governing the economy.
Finally, institutional change involves either redoing rules governing transactions between private parties, within the public sector (for example, government and public enterprises), or between private and public parties. For example, governments may decide to privatize and then regulate firms, rather than run them as state-owned enterprises, in the hope that it will increase efficiency of the sector. How the new system functions will depend on whether the government is a better regulator than it is a producer in that specific activity. 34

Aid: Transplants, knowledge transfer, information provision, demonstration, change in endowments
There is some research on the impact of aid in effecting institutional change, yet this literature remains to be developed and the conclusions to date differ between studies. Research has discussed several ways in which aid may either improve or weaken institutional quality either by affecting formal rules or informal ones, or by supporting incomplete institutional reform (for example, eliminating old institutions, but not replacing them with new ones). 35 However, it has been difficult to find systematic evidence of overall improvements in institutions as a result of aid flows. The difficulty with translating the impact of aid onto changes in institutions is manifold, not the least of which being to precisely identify the impact of aid versus other factors on institutional change in empirical analysis. There are also issues with the timeframe over which the impact of aid is assessed. The time for a particular institutional change to become effective often extends far beyond the timeframe of the aid intervention and the related assessment. For example, when laws are changed and expected to be enforced through the courts, they may be enforced only slowly at first. There may be a period during which traditional rules and norms dominate outcomes. Over time, traditional rules are also affected and formal laws are better understood and capacity better developed. In addition, indicators of institutional change that are available to assess 34 The famous phrase "who monitors the monitor?" comes to mind. 35 For example, aid may render governments less accountable to their citizens, as they will rely less on taxation for resources. Aid changes the implicit rules (and sometimes formal ones) governing public expenditure patterns and resource use.
impact may not be appropriate measures of the intended effect. Finally, effectiveness of a particular institutional change may depend on other supporting institutional changes having been made (see Part One). With these caveats in mind, I present some conclusions. The literature that examines institutional development over long periods is the most promising because institutions exhibit persistence. Also, the more specific is the change reviewed, the greater is the likelihood of learning for policy purposes.
Multi-and bilateral donors of foreign aid often attempt to influence institutions of the country to which they donate funds or other assistance. These may be aid projects explicitly designed to improve "governance" in which particular institutions are directly changed or transplanted (e.g. changing the legal framework for company registration or investments), or they may fund investments or other support in various sectors of the economy (for example, agriculture sector investments) where governments undertake to reform policies and institutions over a period of time. In addition, there are projects where the goal is to induce institutional change by affecting preferences (for example in projects enhancing information flows-transparency-between market players, training programs, technology provision, infrastructure provision and policy advice, or projects that build skills or education).
Often aid donors support some sort of transplant of institutions from developed or aid-giving countries. Laws, rules, or regulations may or may not be adapted in the aid-receiving country. As explained in this review, the impact of the new rules will depend on how well they are enforced and how well complementary rules function. For example, not only may an effective judiciary (or arbitration system) and/or police be necessary for rules on collateral to take effect, but institutions allowing for ownership verification (property registration) and transfer, institutions allowing for verification of the value of the property, etc., will also need to be in place and to "work". At the same time, these rules and regulations must be compatible with social norms and capacity. Competition law will only have its The impact of each of these interventions has not been thoroughly studied.
A number of existing studies that link aid to broad institutional change generally do not find a significant relationship. Svennsson (2000), Knack (2001) and Brautigam and Knack (2004) find that aid may have a negative impact on how institutions function (by increasing corruption and reducing accountability). Another conclusion is that if the people for whom the legal change is intended, prefer the traditional way of doing business, then it is likely that a variety of measures will exist to delay the law from taking effect in a given country. Shirley (2005) finds foreign aid to be a poor tool for institutional reform, citing the "fundamental mismatch" between the characteristics of institutions and those of aid, the former exhibiting durability or persistence and reflecting existing power structures. She also points to the multiple objectives of aid agencies. However, she acknowledges that change is possible over time in response to market competition and when there is adaptation to local circumstances. Islam and Coviello (2006), find that, on average, aid has had no impact on the quality of economic institutions-in other words, even if institutions were transplanted, overall, they were not effective. Young et al (2014), using a panel of up to 116 countries from 1970 to 2010 estimate the effects of foreign aid flows on a variety of measures of institutional quality and find that aid flows are associated with the deterioration of both political and economic institutions. Askarov et al (2015) find that aid has no effect on governance in transitional economies. Deaton (2013) finds that aid can affect institutions negatively when foreign aid is a large proportion of the budget relative to domestic resources, while Jones and Tarp (2015) do not find such evidence, finding that aid directed towards improving governance improves institutions in recipient countries. World Bank (2008) finds that support to the public sector did not always result in successful institutional reform with "success" varying across different types of initiatives. For example, institutional change supported by the Bank was more successful in reforming tax administration, an area where there was a strong "leader" (the finance minister) and a specific and well-defined outcome, and with a clear distribution of gains (and losses). Support to anticorruption laws and commissions rarely succeeded, mainly because they lacked support from "political elites" and the judicial system. Complementary institutions (for example, an independent and effective judiciary, independent media, data and evidence and so on) 36 are critical to the success of these initiatives. The report also recommends focusing on basic reforms and accepting that significant results will only show after a long time.  examines the judicial system reform for commercial transactions and concludes that rules within and external to the judiciary matter for effective courts, that different types of dispute resolution will co-exist in any system so that it is important to understand their interactions, and that many legal scholars have found comparative legal study to be an indispensable tool to design effective courts. Acemoglu and Jackson (2016) show that laws that are in strong conflict with prevailing social norms may backfire; they contend that gradual tightening of laws can be more effective than large changes in influencing social norms and behavior. Aldashev et al (2009), examining the judicial system, argue that even where the "modern" law and court system is little used, there may be an institutions, principles and terminologies in the contemporary legal system in China were transplanted by the Chinese from foreign countries in the name of modernization, though they were modified in order to implement them according to local needs. He and Li (2003) contend that foreign laws are now an indispensable part of the main body of modern Chinese law. However, enforcement may not happen as in the original scenario without appropriate modification. In China for example, local legislators responsible for enforcement may have different motivations from federal ones. Transplanting legislation that works in more centralized systems would be an incomplete institutional change, unless the transplanted legislation is adapted for the supporting domestic institutional system. Modification may also entail redefining the situations under which the legislation is binding. Chen (2014) finds that both reasons explain "incomplete" transplantation. Berkowitz et al (2001) studying 49 countries find that countries where populations are more "familiar" with the adopted legal system, are those where laws are transplanted more effectively. 37 Xu et al (2016) looking at the impact of Chinese aid to Africa in the agriculture sector discusses how the transplantation of Chinese systems with their own rules -a system where state-backed science, a close link between the state and business united by a common aim, seeking technological solutions to development-did not transfer well to the African context (Mozambique and Zimbabwe, for example). The Chinese approach was to demonstrate their success, and transfer their technology and systems intact. In the African context, with different history and norms, and expectations of behavior the approach did not translate well.
As the literature above indicates, the story of aid and institutional development is a complex one and much remains to be studied. This is partly because it is difficult to precisely identify the effect of aid versus other factors, such as host country initial conditions. It is also because these studies have often tended to look at broad measures of institutions and institutional change which may not have been 37 Familiarity may derive from similar religious backgrounds, for example. 29 appropriate to the context, the desired outcome may not have been sufficiently clarified, the complete set of institutional reforms may not have been adopted, and change takes a long time to happen.

What does this mean for development policy: Can institutions be changed as policies can?
Institutional change is happening all the time; but an important question is whether it can be externally directed, that is influenced by donors (working with government). The short answer is that history shows that leaders (external/internal), governments and people working together can change institutions. Donors, if they are humble, learn from history, and understand the forces for change, may help move the process in the right direction, remembering that all change takes time and the time it takes is often underestimated.
What are some of the lessons?

Is the institutional change, especially if it is an addition to existing institutions, really needed?
Countries are sometimes asked to adopt certain institutions, without due consideration of whether such change is needed. For example, when market competition is low, the right answer may not be to establish a competition authority (but to reduce land use restrictions, develop financial markets or change trade policy), or when corruption is high, establishing an anti-corruption authority may not be a useful move.
2. If the change is necessary, the first actions are to identify the forces for change in the economy (as discussed in this paper), assess how strong they are, form an informed judgement regarding whether these forces for change will pick up momentum and how they might be strengthened (and in some cases, created) and thus, design interventions that are most likely to bear fruit in the given context.

Related to this is the next step, identifying who benefits and who loses from the institutional change
and in what timeframe? How will the forces for change influence their incentives? Answering this question will not automatically provide a solution, but will at least inject realism regarding the magnitude/speed of change possible.
Clearly if a consideration of (2) and (3) leads to the judgement that the original reform will not be effective or sustainable, then a modification is recommended.

Design a multi-pronged strategy/ employ more than one source of change: Any endeavor to change
institutions directly (such as by transplants) should be accompanied by knowledge transfer, data and estimates of the distribution of gains, benchmarking against relevant countries, and demonstration.
Sometimes transparency is promoted with a view to imposing reputational penalties that enforce change in debtor behavior. 38 Attempts to change institutions by increasing information (transparency) or knowledge, should be designed with an understanding of how they may affect people's incentives, and the demand for change ((for example, will citizens care enough to act) but also, what else is needed for actions to change. For example, transparency regarding connected lending in the banking sector will have no effect unless supervisory and regulatory institutions are changed with a view to incentivizing supervisors and supervisors can enforce appropriate behavior in bank managers without undue penalties.
In the same vein, transparency regarding wealth and conflict of interest in government may not reduce conflict of interest unless there are penalties, such as being voted out of power or the payment of punitive fines, in case of inappropriate behavior. 39 Moreover, the timeframe for change may be very long.

There is no "best" solution for all cases; one size does not fit all. Direct transplants do not generally
work, particularly where the recipient country differs dramatically in economic, social or political conditions from the transplant country. Adjustment should be made accounting for differences in factors such as capacity and norms. For example, enforcement of intellectual property rights requires considerable knowledge of the relevant industry, the economics of the issue, the distributional consequences of IPR protection, of the state of technology, among other things. Systems, when adopted, need to be modified accordingly. At the same time, it is important to note that prevailing views of what degree of market power is "acceptable" will differ across countries, influencing how the systems are designed and implemented. 38 Unfortunately, reputational penalties have been very weak enforcement mechanisms in this regard. 39 These penalties may be in many forms. For example, fines, relinquishing of assets, loss in elections or in position and so on.

Have measures of success and make sure there is some evidence of impact before leaving the field:
It is important to form clear and shared (with stakeholders) expectations of the timeframe required to achieve effective institutional change -i.e., a change in desired outcomes. Intermediate goals should be set and institutional design adjusted as needed. It is often hard to get the right design in the first try, but it is also hard to change it once adopted, unless the goal is directly linked to the institutional change adopted. Establishing objective measures of impact will make it easier to adjust design as needed and help maintain momentum for the reform.

Value of experimentation:
In order to attain effective institutional design, experimentation may be very useful. For example, different health care provision systems, or different types of social assistance systems may be designed in the same country in order to "test" the best design. Experiments that yield desired outcomes can be expanded. For this process of institutional reform to be effective, there must be clear (observable, measurable) evaluation criteria and the ability to end the pilots that do not yield results.
Experimentation allows countries to learn from their own past experience and other-country experience at relatively low cost. Failing designs are easy to drop when the scope is smaller and innovations may be easier to handle at a small scale (many NGOs, such as Grameen Bank start this way).

Identify the set of complementary institutions:
More than one rule may need to be changed to achieve a certain outcome. Changing a rule in isolation of others that influence the same transaction may have unintended consequences (or none at all). For example, regulating the use of a pesticide will not be effective unless there are systems of enforcement. Improper use of medicinal drugs through the adoption of a legal framework will not be effective unless there are penalties for non-compliance, and systems to check compliance and collect penalties. However, overly burdensome systems to check-compliance could have side effects in terms of reducing drug use, or raising mortality rates due to delays, or facilitating corruption opportunities when regulation is complex or cumbersome and regulatory agents are hard to monitor. Unless the set of complementary institutions can be changed and adapted together, it may be better to rethink the approach. This consideration leads to the next "lesson".

Systemic versus narrow institutional reform.
Sometimes countries will undertake systemic reform of the economy and sometimes they will choose particular areas to reform. Categorizing broadly, governments may reform institutions that (a) directly impinge on a number of sectors (for example, business entry and exit regulations, judicial sector reform), while simultaneously doing sector-specific change (for example, agricultural land use regulations); (b) are broad (such as adopting a competition law), but do not have the necessary complementarities in place (such as enforcement agencies and qualified evaluation officers) or sectoral depth; (c) that are "narrower" in scope (such as energy production or distribution reform), but are not accompanied by economy-wide reforms and (d) that are "narrow" and do not require immediate change in performance (such as requiring publication of accounts of firms in a country without requiring additional measures to change performance). By choosing which type of reform they are willing to adopt and how fast they will adopt each one, governments influence whether institutional reform is sustained or effective.
History has several examples of different types of reform as discussed in the pages above. The evidence indicates that whether directed institutional change takes root may depend on how substantial the change is and what forces are put in motion to ensure that the initial momentum for change is picked up as time progresses. In cases where the change is narrower in scope, but substantial within the scope selected, the eventual impact will depend on whether the initial change fosters further reform. Change is unlikely to be sustained unless these forces are in motion and success outcomes defined. The short run negative consequences of economy-wide and deep changes may be large and sudden on the economy.
For example, fearing a spike in unemployment, reform may be undertaken in small steps over a longer period of time. However, a slower or narrower reform process may not be sustained, may be too small to have any impact on outcomes, or may simply create so much uncertainty (or disruption) that economic processes are stalled. Many countries have unfortunately exhibited these outcomes, when a more deliberate and substantial approach may have borne fruit. The diversity in initial conditions, such as the political economy, the extent of distortions prevailing, growth and income distribution, may argue for diversity in approach, Yet, even if the reform process is limited to a limited segment of the economy (for example, the energy sector), it should be complete, with a well-defined sequence of steps, force for change, conditions for sustainability and desired internal and external outcomes clearly defined.
Piecemeal, incomplete reform in numerous sectors, particularly if critical constraints are not modified, while giving the impression that large steps are being taken, will result in no change in outcomes. Under either type of reform, change imposes costs on the economy, some of which may be persistent.
Embarking on reform without due consideration of these costs could jeopardize the viability of institutional reform? 40

Conclusions
This paper covers two questions: (1) what are institutions and (2) how do they change? This paper presents a widely accepted and workable (for policy purposes) definition drawn from the academic and policy literature, and focuses mostly on institutions that are relevant for market transactions. It then discusses the many forces influencing institutional change. Some important features of institutions are highlighted: (a) they exhibit diversity in form for similar functions and objectives; (b) any market transaction is supported by more than one (formal and informal) institution; (c) they exhibit persistence, but are dynamic. They are persistent because those that benefit from extant systems have an interest in preserving them. They are dynamic because these benefits change as a result of external shocks, directed interventions or endogenous domestic forces, or a combination of all three; (d) there are several sources of change, and these are discussed in this note. However, the direction of influence is not one-way. For example, trade changes institutions, but institutions also influence how much you can trade; (e) when institutional change is to be influenced, the important thing is to harness the forces for change discussed in this paper.

34
Finally, a policy maker or policy advisor seeking to improve economic or social outcomes, will necessarily aim to reform certain institutions. This note builds on studies conducted across a variety of fields, and points to some important lessons in this area. It also highlights the need to do much more detailed research and study on institutional reform and how external agents may be drivers of such reforms.