Understanding institutions: different paradigms, different conclusions

There is widespread agreement across the social sciences that institutions matter. Although there is widespread agreement that institutions shape the behavior of actors, there are disagreements over the extent to which they actually influence individuals and the degree to which people can shape them. This article argues that these disagreements derive from the fact that scholars are studying different institutions at different levels. Many misunderstandings within the social sciences result from the failure to understand that scholars are often studying different phenomena. This problem would appear to be particularly stark in the discipline of Business Administration because it draws on a variety of areas of study in the social sciences.

formal and informal institutions, whereby the former are devised rules and the latter are conventions and codes of behavior. They constrain actors through sets of incentives and disincentives that channel human behavior in a particular direction, thereby creating stable structures that promote efficiency in human interactions by reducing uncertainty and transaction costs. They provide structure and order by aligning the actions and expectations of individuals in a society. Interactions are more efficient because perceptions and understandings are implicit and do not have to be explained or negotiated. According to Paul (1994) institutions help coordinate actions between diverse actors in society without the need for centralization, albeit by limiting what actors can do. Coordination is possible because actors share histories that provide them consistent expectations. Pierson (2000) argues that the increasing returns that derive from the existence of an institution over time makes it difficult to dislodge even if another institution may be more efficient in the abstract.
Some scholars argue that institutions practically determine the actions of individuals. They claim that institutions define preferences and power in society (Powell & DiMaggio, 1991;Thelen & Steinmo, 1992), while also providing the shared meanings and cognitive frames that shape how human interpret the behavior of others (Fligstein, 2001;Hall & Taylor, 1996). They shape preferences and even determine what people can imagine themselves doing (Hall & Taylor, 1996). Institutions are difficult to change because they shape the very choices individuals make when attempting to change them (Hall & Taylor, 1996 Although North (1990) asserts that humans consciously design institutions to help them efficiently meet their goals, Powell and DiMaggio (1991) argue that institutions are not human designs but rather evolve out of the particularities of a given historical and cultural context. Scott (2007) has a less deterministic interpretation of institutions, focusing more on intersubjectivity and individual interpretations rather than larger social processes. Cognitive institutions, he argues, reflect the manner in which actors understand their environment, whereby meaning arises through interactions and involve subjective interpretations and the social construction of individual and collective actors. However, he also draws attention to what he terms are regulative and normative institutions. Regulative institutions are the rules of the game and consist of written and unwritten codes with enforcement mechanisms. Normative institutions are norms and values that structure choices, emphasizing how things should be done and defining legitimate means to accomplish them. Even though they provide a stabilizing force in society, actors internalize them to different degrees. Schmidt (2010) contends that there four types of institutional approaches, namely rational choice, historical, sociological and discursive institutionalisms. Rational choice institutionalism presumes that actors have fixed preferences and act rationally to maximize their preferences. Institutions only influence actors by incentivising actions and reducing uncertainties. Work using this approach assumes that actors can understand the effects of the institutions they create and often relies on functionalist explanations for the existence of institutions. Sociological institutionalism examines how actors follow rules and norms, assuming that identities and culture are the sources of interests for individuals. Human action is examined in terms of its symbolic and ceremonial purposes not utilitarian ones. This approach emphasizes how institutions shape actors and not how actors shape institutions. Historical institutionalism conceives of institutions as being sets of regularized practices. It focuses on how they develop over time, often comparing institutions across countries.
Discursive institutionalism examines how actors generate and legitimize ideas through a logic of communication, focusing on the interactive process through which ideas are generated. Institutions provide the framework for discourse by defining the range of ideas that are more or less acceptable in discursive interactions. Schmidt (2010) argues that these four types of institutionalisms should be considered complementarity to each other rather than different ways of understanding the same phenomena. Williamson (2000) contends that institutions exist at four levels in any society. The first, and most general, are informal and consist of customs, traditions, norms, religion, etc. Although they are slow to change, they "have a lasting grip on the way a society conducts itself" (Williamson, 2000: 597). The second level, the institutional environment, consists of formal rules such as constitutions and laws. They emerge partially out of an evolutionary process and partially out of design. Governance, the third level, is the creation of contracts and agreements between a limited number of groups in society. They are designed to reduce conflicts and set the basis for mutual benefits. The fourth level is resource allocation. For companies, it is the level of the production function and consists of institutions that guide the operation of an organization. By contrast, governance describes the regulation of operations between independent organizations. Lower levels can have only limited influence on those levels immediately above them, whereas upper levels impose clear limits on lower ones. Williamson (2000) points out that the work of economists focuses on levels two and three in his framework.
Although many scholars across the social sciences claim to use a neoinstitutional approach, they often fail to specify exactly which type of neoinstitutionalist approach they are following. Consequently, scholars frequently misinterpret the work of others even within their own discipline. Moreover, some scholars even seem to confuse the different approaches within their work leading them to make misguided conclusions. This article now turns to an analyze both of these points by examining the type of institutionalist approachs used by Khanna and Palepu (1997) and Kostova, Roth, and Dacin (2008). These two works represent seminal pieces in the areas of strategy and international business respectively. This article suggests ways that both of these works could be improved by incorporating insights from some of the neoinstitutionalist approaches described above. Khanna and Palepu (1997) argue that emerging markets have a number of "institutional voids" in product, capital and labor markets. The lack of a reliable business press and missing regulations cause firms to have to reliable on their own financial resources. The dearth of reliable sources of information in general cause firms with strong brand names to gain more clients than their rivals. The absence of adequate schools and training institutes requires firms to train their own workers. In short, the lack of institutions supporting the market cause firms to have to vertical integrate activities that normally can be acquired on the market. In emerging markets business groups are successful because they have the capabilities to fill these voids.

Examples of the application of neoinstitutional theory
This article by Khanna and Palepu focuses only on formal institutions, overlooking the possible role that informal institutions could play in facilitating the operation of the market. Mair and Martib (2009) contend that the term institutional void should be used to describe only the lack of institutions supporting markets, not the lack of institutions per se. Emerging markets have a number of institutions, many of which actually impede the operation of markets. In terms of Williamson's framework level 2 institutions may be absent or deficient but level one institutions are always present. Furthermore, firms may actually be able to address deficiencies in institutions by working with a limited number of other firms and/or organizations at level three in Williamson's framework. In terms of Scott's approach, this article by Khanna and Palepu focuses only on regulative institutions, not on normative or cognitive ones. Kostova et al. (2008) contend that scholars in international business exaggerate the degree to which local environments shape the behavior of multinational companies in foreign countries. They contend that these companies operate in three Table 1 The authors and the theories of institutions. distant organizational fields, namely the ones of their home country, their host country, and the international community. By being in three fields at the same time, the strategies of these firms are said not to be limited any one of them, enabling them to largely ignore the isomorphic pressures of their host environments.
By deploying terms such as isomorphism and organizational fields, this article by Kostova, Roth and Dacin seems to draw on sociological institutionalism because these terms are developed in this area of research. As mentioned above, this type of institutionalism emphasizes symbolic and ceremonial purposes focusing on norms, identity and culture and the way these institutions shape actors. According to this form of institutionalism firms cannot simply avoid or ignore institutions, particularly informal ones. These authors seem to confuse the sociological approach with the rational choice one. Moreover, they contend that governments in host countries do not expect these companies to adapt their practices to their countries' contexts.
Given that isomorphic pressures result from general institutional settings and not from the decisions of policy makers to consciously enforce them, it would be difficult to imagine how MNCs could actually be excluded from isomorphic pressures. Consequently, the choices available to subsidiaries of multinational corporations would seem more limited than those available to local firms. The behavior of the former is limited by isomorphic pressures from their headquarters and from the local setting in which they operate, while the decisions of the latter are only shaped by the isomorphic pressure of their local context. Although these authors draw on the sociological approach they reach conclusions from a rational choice institutionalist perspective that focuses on formal institutions and regulations. In terms of Williamson these scholars claim to be studying level one institutions but they are using assumptions derived from level two institutions. Although these authors appear to study institutions at level one and two in Williamson's framework, they seem to attribute a level of flexibility in choice that is really only available to firms a levels three and four in it. Table 1 provides a summary of the concepts on which each of these articles draws within the four major approaches to institutions discussed above.

Conclusion
The neoinstitutionalist approach across the disciplines in the social sciences has played an important role in drawing the attention of scholars to the important role played by institutions in coordinating activities between individuals as well as those between groups. Although there are significant differences in the approaches outlined above in regards to the degree to which institutions determine the behavior of actors, all of them agree that they are critical to the functioning of any society. However, in order for progress to be made in this field scholars have to be careful not only to properly understand the approach they are using but also the potential insights that could be derived from other theories addressing levels and types of institutions. Work of this kind would help stimulate better discussions that could eventually lead to more fruitful dialogs across approaches, thereby potentially advancing theory even further.

Conflicts of Interests
This author declares no conflict of interest.