Process institutionalism: toward an action-centric approach to state extraction

What is the relationship between actions and institutions in state extraction? State extraction is the process whereby revenue is extracted from constituents to the state. Studies on state extraction in the early modern era mostly adopt the institution-centric approach, which perceives actions as manifestations of institutional and structural characteristics in a social context. However, it does not explain the varying actions and the resultant diversified institutional changes beyond the behavioral and institutional repertoires determined by these characteristics. This article proposes the process institutionalism model as a new paradigm for understanding the relationship between actions and institutions in state extraction. This model employs an action-centric approach, which maintains that actions lead to changes in institutions and the actors’ consciousness. It also demonstrates the qualitative contradictions among the incentives in the efficiency and legitimacy dimensions of an action and adopts an eventful explanation of actors’ understandings of and selections among the contradictory incentives during the temporal process of actions. Process institutionalism engages theoretical and empirical research on the relationship between actions and institutions by reviewing existing literature on state extraction in history, especially the history of the early modern period, the critical juncture whereby states and other related institutions experienced dramatic changes and displayed regional diversity.

as manifestations of objective forces embodied in institutional and structural characteristics. 1 Thus, this approach does not explain the varying actions and the resultant diversified institutional changes beyond the behavioral and institutional repertoires determined by initial institutional and structural characteristics. Consequently, it leads to disengagement between the theoretical and empirical research on the relationships between actions and institutions in state extraction.
In contrast to the institution-centric approach, there is an action-centric approach opened by Weber's (1978Weber's ( , 2012Weber's ( [1913) interpretive understanding of social action in theoretical sociology. In general, Weber (1978: 20-21;2012[1913: 284, 288) emphasizes qualitatively contradictory motives faced by an actor during the process of action and, thus, illustrates why actions are central to institutional changes and lay a solid foundation for the action-centric approach. Barbalet (2009a, b) suggests that actions lead to changes both in institutions and in the actors' consciousness, presenting the central proposition of the action-centric approach. Regarding state extraction, Schumpeter (1991Schumpeter ( [1918) and Campbell (1993) suggest that actions in state extraction are central to the changes in state institutions and other related institutions in a broader context and thus employ an action-centric approach to state extraction.
This study begins with the action-centric approach and proposes the process institutionalism model as a new paradigm for understanding the relationship between actions and institutions in state extraction. The focus is on the process by which actors, having different interests and values, initiate social actions. The model moves beyond institutions to adopt an action-centric approach, which holds that actions are fundamentally important to changes in institutions and actors' consciousness.
However, disengagement tends to appear between theoretical and empirical research on the relationship between actions and institutions in the existing studies of the actioncentric approach. On the one hand, Weber (1978Weber ( , 2012Weber ( [1913) and Barbalet (2009a, b) tend to develop an action-centric approach in general and demonstrate the relationship between actions and institutions at the theoretical level. Schumpeter (1991Schumpeter ( [1918) and Campbell (1993), on the other hand, focus on the empirical process of state extraction. However, they do not offer a coherent theoretical model for understanding the relationship between actions and institutions in state extraction.
The early modern period was an emblematic Schumpeterian turning point "during which existing forms begin to die off and to change into something new, and which always involve a crisis of the old fiscal methods" (Schumpeter 1991(Schumpeter [1918: 101). 2 In this period, fiscal and other related institutions varied across regions and changed over time, demonstrating characteristics that the institution-centric approach could not explain. Thus, the state extraction processes in this period are deviant cases for this particular approach. In contrast, analyses of the early modern period's divergent institutional changes and the crucial role of actions in the processes suggest the advantage of the action-centric approach and offer substantive content for the process institutionalism that adopts the action-centric approach.
Thus, my study builds on the existing literature on state extraction in history, especially the literature on the early modern period, 3 to develop the process institutionalism model, which combines theoretical and empirical research to interpret the relationship between actions and institutions. Specifically, process institutionalism proposes a theoretical framework that can be applied to the sociological analysis of the formation mechanisms of fiscal states in both Eastern and Western societies, especially in the early modern era, thereby contributing to the understanding of their formations and changes after the fifteenth century. Moreover, it contributes to the ongoing debate about the relationship between actions and institutions in state extraction and furthers the project of new fiscal sociology (Martin et al. 2009;Martin and Prasad 2014;O'Brien 2017).
First, the paper reviews previous studies on state extraction in the early modern period in both Eastern and Western societies to illustrate the theoretical inadequacies of the institution-centric approach. It then develops the process institutionalism model for understanding the relationship between actions and institutions and presents conclusions.

The institution-centric approach
The institution-centric approach emphasizes the conformity of action to either the efficiency properties of formal institutions or the ideas embodied in informal institutions, accompanied by structural characteristics in a broader social context. Specifically, studies following an institution-centric approach can be classified as those recognizing action as benefit-driven and those perceiving action as culturally formed (see Fig. 1). The former include Tilly's (1975) bellicist theory (Karaman and Pamuk 2013;Liu 2015;Saylor and Wheeler 2017), Kiser's (1994) principal-agent model (Kiser and Tong 1992;Kiser and Schneider 1994;Sng and Moriguchi 2014), Levi's (1988) predatory rule theory, and North's (1981) Fig. 1 The institution-centric approach 3 It is worth noting that although this study focuses on studies of state extraction in the early modern period, it also appreciates some studies of state extraction in other historical periods, for example, Whiting's (2000) study of state extraction in China in the reform period (1980s-1990s).
view actors as self-interested and their actions as benefit-driven and emphasize the conformity of action to the efficiency properties of formal institutions. Meanwhile, the cultural theory of Webber and Wildavsky (1986: 24-31) suggests that actors' preferences emerge from the ideas embodied in cultures and perceived by the actors and views the actions of actors and the resultant forms of institutions (such as budgets) as manifestations of a specific culture or combination of cultures. Both the models that recognize actions as benefit-driven and cultural theory are examined below.
First, models recognizing actions as benefit-driven tend to perceive actors' actions in state extraction as determined by interest-specific incentives and behavioral programs, embodied in the structural characteristics of the formal institutions in state extraction, other institutions, and structures in a broader context. Models recognizing actions as benefit-driven tend to assume that all the actors in state extraction are "rational and selfinterested. " They assume that the ruler is inclined to "maximize revenue to the state" (Levi 1988: 10;Kiser and Schneider 1994: 190).
However, studies adopting the new institutionalism model (North 1981: 20-32), including those on the variety of organizations in processes other than state extraction (March and Simon 1958;Simon 1972;Williamson 1975), recognize that various other institutional incentives may be imposed on an actor, thus constraining self-interest. Nevertheless, these studies imply that the fulfillment of various incentives through alternative behavioral programs can be measured in comparable units in an actor's preference frame, which remains relatively unchanged during an action. They further imply that an actor pursues the maximum number of utility units by selecting and adopting an appropriate behavioral program (North 1981;Levi 1988;Kiser 1994;Sng and Moriguchi 2014). Therefore, an actor's selection of an appropriate behavior essentially becomes a matter of utility arithmetic.
In addition, these models tend to recognize institutions as "humanly devised constraints that structure political, economic, and social interaction" (North 1991: 97). They also perceive such constraints as incentives and behavioral programs imposed on actors subject to these institutions (Williamson 1975;North 1981North , 1991. The models suggest that an actor deliberately searches for and adopts appropriate incentives and behavioral programs to minimize transaction costs under institutional constraints, thus maximizing net revenues in state extraction. Therefore, these models elucidate the causal connections between various initial institutional or structural characteristics, and the actions of authoritative actors and the resultant institutional changes. The bellicist theory emphasizes that fiscal needs arising from expensive wars, together with the specific characteristics embodied in economic structures and political regimes, lead rulers to set up state extraction institutions (Tilly 1975;Karaman and Pamuk 2013;Liu 2015). In contrast to Tilly's (1975: 42) suggestion that fiscal exigencies arising from expensive wars led to state building, Saylor and Wheeler (2017) argue that the structural characteristic of the credit market-the position of government creditors in the credit market-determines their economic interests and resultant actions and, in this way, conditions the connection between war and the state.
Additionally, the principal-agent model suggests that when a country is faced with a military threat, factors such as its geographical size, the number of officials, the complexity of the tasks delegated to agents, and the monitoring capacity of the ruler are factors that impose constraints on the ruler's attempt to expand taxation institutions (Kiser and Tong 1992;Kiser and Schneider 1994;Sng and Moriguchi 2014).
Second, cultural theory recognizes actions as entities that are culturally driven, and it argues that the actions in state extraction are determined by the ideas embodied in informal institutions as perceived by the actors. The cultural theory of Webber and Wildavsky (1986) suggests that the ideas embodied in informal institutions, or "political cultures, " provide a common understanding of the constraints imposed by formal institutions. First, in contrast to the models recognizing an actor as self-interested, cultural theory perceives actors' preferences as manifestations of a specific combination of cultures to which they are committed (Webber and Wildavsky 1986: 20). Such preferences are ultimately about what is valuable and worth having (Webber and Wildavsky 1986: 20). Cultural theory suggests that "the values people prefer, and their beliefs about the world are woven together through their cultures" (Webber and Wildavsky 1986: 24).
Furthermore, cultural theory suggests that, to understand the legitimacy of their selected behavior, actors rely on the cultural values they are accustomed to (Webber and Wildavsky 1986: 25). Legitimacy refers to a specific behavior's righteousness. Both the new institutionalism model and cultural theory understand the legitimacy of an action in the context of rationality and, thus, include this legitimacy in the scope of the study. Nevertheless, only cultural theory recognizes legitimacy as an independent dimension of rational action.
Specifically, the new institutionalism model tends to remain within the paradigm of instrumental rationality (Weber 1978: 24) and suggests that an action is rational when it maximizes self-interest by minimizing the sum of costs associated with the selected behavioral program and the selection itself (Williamson 1975). From the perspective of the new institutionalism model, the legitimacy of an action is associated with transaction costs and, in turn, with the efficiency calculation of the action-in this context, legitimacy is related to rationality (Williamson 1975;North 1981).
In contrast, cultural theory adopts the paradigm of value rationality and recognizes the legitimacy of an action as an independent dimension comparable with-and not subordinate to-the dimension of efficiency. The cultural theory further suggests not only that an actor recognizes the rationality of an action as worthwhile in the frame of preferences emerging from cultural commitment but also that it is "desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions" (Webber and Wildavsky 1986;Suchman 1995: 574).
Cultural theory suggests that the ideas embodied in informal institutions provide a common understanding upon which an actor relies to discover the meanings of actions and to select an action that is not only worthwhile but also morally right (Webber and Wildavsky 1986: 24-25). Therefore, from the perspective of cultural theory, an action is the manifestation of the specific prescriptions embodied in informal institutions. Hence, a change in political culture leads to a change in preferences and actions, rather than the other way around (Webber and Wildavsky 1986: 24-31). For instance, cultural theory suggests that the Confucian value of benevolent governance (ren-zheng) impelled China's rulers to under-tax the population and led to the small size of the formal state in traditional China (Elliott 2009;Rowe 2009).
In sum, both models-those recognizing actions as benefit-driven and those recognizing actions as culturally formed-suggest that preferences and behavioral programs remain exogenous to an actor (Chang and Evans 2005: 102-109, 131). These models assume that an actor is effectively an automaton, responding to structurally determined incentives, behavioral programs, and ideas embodied in institutions (Chang 2007a(Chang , 2007b. The institution-centric approach suggests that actions are determined by objective institutional and structural constraints in a broader social context, and they are unrelated to the psychological characteristics of actors.
Thus, the institution-centric approach recognizes action as the manifestation of specific institutional and structural characteristics, and it ascribes institutional changes to the initial institutional and structural characteristics in a broader context. The institution-centric approach argues that action plays no significant role in institutional change, except as a conduit for projecting the initial institutional and structural characteristics into the future.

The theoretical inadequacies of the institution-centric approach
The central proposition of the institution-centric approach is that institutional and structural characteristics determine that action. Action is, therefore, predictable and explainable by examining initial institutional and structural characteristics in a specific context. This section borrows from Weber's (1978Weber's ( , 2012Weber's ( [1913) interpretive understanding of social action and Barbalet's (2009a, b) action theoretic foundation of economic sociology to illustrate the qualitatively contradictory constraints imposed by institutional and structural characteristics. It also flags the inconsistency and diversity of an actor's preferences during the action. It, therefore, invalidates the institution-centric approach in the relationship between actions and institutions. The discussion shows that the institutioncentric approach inevitably leads to disengagement between theoretical and empirical research on the relationship between actions and institutions.
First, the studies employing the institution-centric approach, including those recognizing actions as benefit-driven and those recognizing actions as culturally formed, ignore the qualitative contradictions in an actor's subjective perceptions of the constraints embodied in the institutional forms and other characteristics in a broader context. Indeed, Weber (1978Weber ( : 20-21, 2012Weber ( [1913: 288) illuminates the qualitatively contradictory incentives imposed on action "in the majority of cases of action important to history or sociology. " Weber (1978: 20-21;2012[1913: 288) argues that, in the majority of cases, the motives imposed on action are "qualitatively heterogeneous, " and in most cases, in addition to economic motives, each of the "traditional restraints, affects, errors, and the intrusion of other than economic purposes or considerations, " may play a role in the process whereby an action goes on.
Regarding state extraction, institutional and structural characteristics impose economic benefits and costs and noneconomic consequences, such as a decrease in public support, on actions. When an actor considers whether an action is worth taking or not, the perceptions of economic and noneconomic consequences tend to be "qualitatively heterogeneous" (Weber 1978: 21). They cannot, therefore, be measured in comparable units. In addition to efficiency, legitimacy also factors into an actor's consideration, and an action that is thought to be worthwhile is not necessarily legitimate. Thus, qualitative contradictions in an actor's understanding may lead to qualitatively contradictory behavioral tendencies in state extraction.
Second, the studies adopting the institution-centric approach, including those recognizing actions as benefit-driven and those recognizing actions as culturally formed, tend to assume that the structure of an actor's consciousness is constant throughout the action process. However, as Barbalet (2009aBarbalet ( : 145, 2009b has argued, in general, as structural contradictions free an actor from the compulsion of objective forces embodied in the characteristics of institutions and structures, the actor is forced to creatively and volitionally select among the contradictory constraints imposed by institutions and structures during the process of action, thus facilitating changes in both the institutions and the consciousness of the actor. In other words, during the process of an action, an actor is forced to select among contradictory constraints and opportunities imposed by the institutions and structures, inevitably facilitating a change in both their incentive and idea structures. Therefore, a conscious action inevitably leads to a modification of consciousness itself (Barbalet 2009a, b). Likewise, Sewell (2005: 110) recognizes the temporality of an action as causally heterogeneous, which implies that action necessarily changes an actor's incentive structure and consciousness.
Third, among the studies following the institution-centric approach, those recognizing actions as benefit-driven presume that all actors in state extraction are "rational and self-interested" and that the ruler is inclined to "maximize revenue to the state" (Levi 1988: 10;Kiser and Schneider 1994: 190). The studies make an invalid assumption about actors' preferences in state extraction. Indeed, Weber (1978: 21) argues that it is "unrealistic or abstract" that an actor is assumed to be purely oriented to economic ends alone. Empirically, the studies' assumption about actors' preferences in state extraction contradicts some historical cases.
For example, a paternalist ruler in ancient China (such as the first Ming emperor) tended to extract sufficient-rather than maximum-revenue for state operations, except when faced with serious financial inadequacies such as those arising from the heavy expense of war or inefficient state machinery (Huang 1998). A paternalist ruler in ancient China considered not only revenue for the state but also the tax burden on common peasants and their welfare (Hung 2008).
As Wenkai He (2013: 1) correctly suggests, the interactions between the socioeconomic structure in which institutions are embedded and "actors with different ideas, interests, and institutional blueprints" will lead to multiple possible consequences rather than a determinant institutional outcome. The institution-centric approach, which perceives actions as determined by initial institutional and structural characteristics, ignores an actor's subjective understanding of constraints and opportunities embodied in a specific context and the creative selection of options available. The approach fails to explain heterogeneous actions and the resultant diversified institutional changes, which lie beyond the behavioral and institutional repertoires determined by the initial institutional and structural characteristics.
In sum, the institution-centric approach is ill-equipped to empirically explain why institutions change over time or why similar institutional and structural characteristics lead to divergent institutional forms at different times and places. This leads to disengagement between theoretical and empirical research on the relationship between actions and institutions in state extraction. Thus, it is necessary to go beyond the institution-centric approach to develop a new model that adopts a novel approach-an actioncentric approach-to understand state extraction in Eastern and Western countries.

Process institutionalism: toward an action-centric approach
This section goes beyond the institution-centric approach to propose a new model for understanding the relationship between actions and institutions in state extraction, namely process institutionalism. This model harmonizes theoretical and empirical research on the relationship between actions and institutions. It focuses on processes whereby actors with different interests and values initiate social actions. In contrast to the institution-centric approach, process institutionalism adopts an action-centric approach that emphasizes actions as fundamentally important to changes, not only in institutions but also in the consciousness of the actors themselves. The model suggests that the contradictions in the forms of institutions and their structural characteristics force actors to innovatively select among competing constraints and opportunities imposed by institutions and structures (see Fig. 2). In addition, contingent events may either lead to institutional and structural changes and, thus, to the emergence of institutional and structural contradictions or change actors' understandings of institutional and structural characteristics, as well as their contradictions. The latter would directly facilitate their selection among the available behavioral options (see Fig. 2). This deliberate selection process leads to subsequent changes in individual consciousness and institutions (see Fig. 2).
Theoretically, Weber's (1978Weber's ( , 2012Weber's ( [1913) interpretive understanding of social action establishes the foundation of the action-centric approach, and Barbalet's (2009a, b) action theory presents the central proposition of the action-centric approach. Hence, Weber's (1978Weber's ( , 2012Weber's ( [1913) interpretive understanding of social action and Barbalet's (2009a, b) action theory open up the path to combine theoretical and empirical research on the relationship between actions and institutions.  Weber (1978Weber ( : 20-21, 2012Weber ( [1913: 284, 288) illuminates the qualitatively contradictory constraints imposed by institutions and structures on actions, thus necessitating theoretical accounts of actions in institutional explanations. It is "unrealistic or abstract" that an actor should be assumed to be purely oriented to economic ends alone (Weber 1978: 21). Accordingly, Weber's interpretive understanding of social action, recognizing that institutions impose qualitatively contradictory motives on the action in most cases, allows the engagement of theoretical and empirical research on the relationship between actions and institutions.
Furthermore, Barbalet (2009a, b) suggests that actions are central in institutional changes and the actor's consciousness. Thus, he theoretically permits an understanding of the changes in institutions and actors' consciousness during the actual process of actions. As the institution-centric approach fails to account for the changes beyond the initial institutional and structural characteristics and the inconsistency and diversity of an actor's preferences during the process of action, Barbalet's (2009a, b) action theory, at the theoretical level, leads to an understanding of the relationship between actions and institutions that enables an empirical explanation of the changes in institutions and the consciousness of an actor during the process of action.
Regarding the empirical process of state extraction, Schumpeter (1991Schumpeter ( [1918) and Campbell (1993) emphasize that the actual process of state extraction plays a crucial role in the changes in state institutions and other related institutions in a broader context and facilitates the development of fiscal sociology. Schumpeter (1991Schumpeter ( [1918: 100-101) argues that in most historical periods, the state's fiscal needs and fiscal policy explain "economic forms, " "human types, " "industrial situations, " "modern economy, " and the "modern spirit. " Campbell (1993: 163) suggests that taxation has effects on "political revolution, " "state building, " "economic organization, " "labor force participation, " and "philanthropy. " However, neither Schumpeter (1991Schumpeter ( [1918) nor Campbell (1993) develops a consistent theoretical model for understanding the relationship between actions and institutions in state extraction.
By employing an action-centric approach to state extraction, process institutionalism engages the theoretical and empirical research on the relationship between actions and institutions. Process institutionalism, on the one hand, follows the tradition of the interpretive understanding of social action, initiated by Weber (1978Weber ( , 2012Weber ( [1913) and developed by Barbalet (2009a, b). On the other hand, process institutionalism follows Schumpeter's (1991Schumpeter's ( [1918) and Campbell's (1993) emphasis on actions in state extraction. It outlines a series of propositions on the relationships of actions and institutions in the process of state extraction.

Institutional forms and processes by which institutions operate
The process institutionalism model focuses on the forms of state institutions and other institutions in the broader social context and on the processes whereby institutions operate and interact with each other. It includes both formal and informal institutions (North 1991: 97). During the process of state extraction, formal institutions manifest in various ways, including as forms of taxation, forms of contracts between the tax-collecting authority and the tax collectors, and structures of local organizations (Azabou and Nugent 1988: 684;Kiser and Schneider 1994: 193;Choi 2009: 80-82).
Informal institutions are humanly devised and unwritten constraints that structure and moderate political, economic, and social interaction. These include taboos, customs, codes of conduct, ideologies, and other normative social entities (North 1981(North : 45-58, 1991. The forms of informal institutions are composed of symbols such as objects, numbers, and other concrete embodiments of meanings (Geertz 1973: 91).
Specifically, the forms of taxation, the contract between the tax-collecting authority and the tax collectors, the structures of local organizations, the size of the country, the extent of the development of the transportation and communication technologies, and socioeconomic characteristics (such as the financial networks, the distribution of financiers within the country, and the position of government creditors in the credit market) impose interest-specific incentives and behavioral programs on the actors, providing them with a "repertory of programs" (Williamson 1975;North 1981North , 1991Kiser and Schneider 1994;He 2013;Saylor and Wheeler 2017). The ideas (such as moral values) embodied in the forms of informal institutions are both the institutional characteristics upon which actors rely to discover the meanings of incentives and behavioral programs and the tools used to legitimize the selected behavior (Chang and Evans 2005;Chang 2007a: 9-10;Barbalet 2009a: 143-144).
The processes whereby institutions operate and interact with one another are essentially systems of interdependent behavior that link shifting coalitions of multiple actors in a broader social context (Scott 2003(Scott [1981: 29). Institutional operations are sustained by individual-level behavior, and in this context, they involve habits on the one hand and deliberate actions on the other (Barbalet 2009a: 145).
Specifically, a habit occurs "without reference to the conscious will" (James 1931: 114 cited in Barbalet 2009a: 146), and the choice of programs provided by institutions is inherent in the habit. It does not require "awareness that a choice is being made" (Nelson and Winter 1982: 73 cited in Barbalet 2009a: 146). Action, in contrast, "requires conscious selection of a course of conduct from among a number of options or choices" (Barbalet 2009a: 146) and involves modifying the incentive structures of actors, as well as searching for appropriate programs that seek to meet the desired goals (Barbalet 2009a).
The process institutionalism model focuses on the "subjective" and "meaningful" processes of actions that are understood by referring to an actor's consciousness and, thus, inevitably relate to individual experience, memory, and personal temperament (Weber 1978: 21-22) rather than the physiological and involuntary processes of habits and reactive behavior that arise from mechanical and instinctive factors and are understandable in biological terms (Weber 1978: 17). The model emphasizes the forms of formal and informal institutions that provide constraints and opportunities for the behavior of actors. It also emphasizes the processes whereby actors, by referring to individual consciousness, subjectively understand the meanings of the characteristics embodied in the forms of institutions and make a volitional choice among the contradictory constraints and opportunities available.
In contrast to the existing studies that recognize state extraction as state capacity and thus an institutional characteristic of the state, the process institutionalism model perceives state extraction as a process whereby multiple actors interact with socioeconomic institutions and structures. Process institutionalism neither centers on the state actor nor necessarily takes the state as the unit of analysis. In contrast to the studies adopting the institution-centric approach, the process institutionalism model does not recognize powerful actors as pivotal, and it may conduct region-based comparisons, as Pomeranz (2000) and Li (2000) have suggested. Therefore, the process institutionalism model enables an understanding of how state extraction leads to a multiplicity of institutions and the corresponding institutional operations at local levels within a country.

Actions, forms of institutions, and consciousness of actors
The proposition central to process institutionalism is that actions are fundamentally important for changes in the forms of state and other related institutions in a broader social context and for modifying individual actors' consciousness over time. Here, the concept of contradiction is important. When changes in the forms of institutions and structural characteristics lead to perceived contradictions (in the context of demographic, economic, technological, military, and natural events, for example), actors "have no alternative but to choose between emergent opportunities" (Barbalet 2009a: 145). When this is the case, actions are responsible for the ensuing changes in institutions and the actors' consciousness.
Referring to state extraction, actors deliberately manipulate the forms of formal institutions to enhance the efficiency of their selected behavior and add the selected goals and behavioral programs to legitimate informal institutions that legitimize the action. For instance, multiple elites in early modern Europe volitionally exploited the strategic opportunities created by religious divisions that arose from the Protestant Reformation, leading to elite conflicts that shaped the formation of the states (Lachmann 2000: 94). In another example, the rulers of Prussia initiated a specific tax collection system between 1640 and 1806, compatible with the specific structural characteristics. The specific deviations of the Prussian tax collection system from the ideal-typical bureaucracy led to the efficiency of state extraction in the corresponding period (Kiser and Schneider 1994). Siu (1989) argues that during the first half of the twentieth century, the local bosses in rural China facilitated both the persistence and the changes of traditional moral values through a series of actions. These included protecting the local community against outsider intrusion in an age of political disorder and establishing estates and other symbols in the form of traditional informal institutions. Whiting (2000: 24-25) suggests that during the reform period (1980s-1990s), the difficulty of acquiring information about the resources of private enterprises in rural China significantly raised the costs of state extraction, thereby forcing local cadres to develop "new institutional arrangements" to reduce related transaction costs (borne by themselves). Zhang (2017) suggests that the state in contemporary China under-institutionalizes the taxation system. Accordingly, it reduces effective tax rates and alleviates taxation's adverse impact on economic growth. Moreover, it facilitates tax evasion and lightens the pressure on the state to open up to taxpayers' representatives. In addition, the state develops a half-tax state with a high reliance on non-tax revenues, indirect taxes, and state-owned enterprises, which leads to a decrease in citizens' perceived tax burdens and a decline in tax collection costs and taxpayers' demand for representation in the state.
At the same time, the actions of multiple actors within institutions will lead to changes in the incentive and idea structures of the actors themselves, thus changing their consciousness. In contrast to the new institutionalism, which implies that the various incentives imposed by institutions are qualitatively homogeneous (March and Simon 1958;Simon 1972;Williamson 1975), process institutionalism, following Weberʼs (1978 argument, suggests that the multiple incentives imposed by organizations and the other institutions in a broader social context tend to be "qualitatively heterogeneous" within an actor's subjective understanding. During an action, institutionally constrained actors pursue compatibility among the incentives embodied in the institutions or select among the contradictory incentives, thus modifying their incentive structures and initiating subsequent actions. As actors do not possess a consistent structure of incentives during an action that is inevitably directed toward the future, they are forced, when initiating the action, to imagine the future incentive that the action will fulfill (Barbalet 2009a, b). Therefore, when an actor initiates an action, any entity's desire for benefits and claim to legitimacy in the future may be involved in the actor's imagination at present in view of the corresponding desire's emotional vividness (Barbalet 2009a, b). As the presumption of this consistent structure of incentives is abandoned, the process whereby an actor acts to fulfill the interest-specific incentives of the entities other than themself is subject to interpretation, as are the other actors' claims to legitimacy.
In addition, in contrast to cultural theory, which emphasizes actors' conformity to socially held ideas in informal institutions (Meyer and Rowan 1977;DiMaggio and Powell 1983;Webber and Wildavsky 1986), process institutionalism argues that actors, from a subjective understanding, deliberately select from ideas in informal institutions and use them in their actions, thereby actively constructing a relatively coherent idea structure (Wilson 2011). Thus, the consciousness of an actor, which involves the structures of incentives (related to preference frames) and of ideas (related to worldview of society), tends to be constructed "through curiosity and the acquisition of various competences" during the action (Barbalet 2009a: 152).
In state extraction, as the contradictions among the institutional and structural characteristics lead to a fiscal crisis, actors-especially authoritative actors-are impelled to search for an appropriate behavioral program to alleviate the state's fiscal risk. The subsequent process involving "experimentation and learning" leads to the "continuous accumulation of effective elements and knowledge. " It ends with creating new institutions (He 2013: 22-23). In other words, action leads to both institutional changes and the formation and modification of an actor's consciousness.

The two dimensions of action in state extraction: efficiency and legitimacy
There are two major dimensions of action in state extraction-efficiency and legitimacy dimensions. The former relates to whether an action is worth taking, while the latter relates to the validity of an action. Both are involved in rational actors' considerations, and their operations and interactions are included in the scope of rationality. 4 During the action process, an actor can face qualitative contradictions not only among the motives at different dimensions (e.g., between an interest-specific incentive in the efficiency dimension and a claim to legitimacy in the legitimacy dimension) but also among the motives within the same dimension (e.g., among the claims to different types of legitimacy). Analyzing the two dimensions is necessary for understanding the qualitative contradictions in the action process and, thus, is fundamentally important for developing the process institutionalism model.

The efficiency of an action in state extraction
The efficiency of an action is related to whether it is worth taking and is understood by perceiving and evaluating its consequences. The consequences of an action in state extraction can be classified into two types: economic and noneconomic. Economic consequences relate to both the objective revenues and the objective costs imposed on individual actors, and they play a crucial role in the fluctuations in the entire society's net production (McGuire and Olson 1996). Noneconomic consequences relate to the ethical, aesthetic, religious, or other related experiences attached to an action (Weber 1978: 24-25).
The economic consequences in state extraction tend to impose objective resource limits on the action, and a ruler tends to exclude actions resulting in lower net revenue than the minimum resource requirement from their behavioral repertoire. Only the behavior leading to adequate revenue for the operation of the state apparatus can be included in a ruler's behavioral repertoire in view of the corresponding ruler's knowledge. In extreme cases, such as war, the highly expanding short-term need of the state will force a powerful actor to search for an available behavioral program to alleviate imminent fiscal and political crises, regardless of the high inefficiency of the corresponding behavior in the long term.
For instance, to survive during a war, rulers in the early modern period were forced to select an action that led to the highest revenue in the short term despite the risk of subsequent threats to economic growth and political stability. Such an action, for example, could be the issuance of short-term borrowing bills unfunded by tax revenues. In this way, high inefficiency emerged in the long run (Levi 1988: 32-33;Campbell 1993: 166;He 2013: 183-184). The state's dependence on fiduciary credit instruments to meet its spending needs led to subsequent fiscal crises, which impelled powerful actors in England after the Civil War and in Meiji Japan to "experiment with new methods and institutional elements" to pursue an increase in the efficiency of state extraction and ultimately facilitated the centralization of tax collection (He 2013: 183-184).
However, as there are no contingent events that force state actors to exhaust the potential of the socioeconomic structures to fulfill the state's fiscal needs, state actors may select among behavioral programs that lead to adequate revenues for the state. Economic and noneconomic consequences for an individual actor and other actors may factor into the actor's consideration "on the level of meaning" (Weber 1978: 23-24). Zhang (2017), for example, argues that the contemporary Chinese state decreases effective (or perceived) tax burdens on taxpayers by under-institutionalizing the taxation system and relying heavily on non-tax revenues and indirect taxes, seeking to decrease taxpayers' demand for representation and the cost of tax collection.
In addition, in the process of state extraction, actors other than rulers do not necessarily possess the incentive to extract maximum revenue or pursue the highest efficiency, even when the state revenue is inadequate. For instance, local officials in state extraction may pursue "career success" rather than the highest revenue (Whiting 2000: 18).

The legitimacy of an action in state extraction
Legitimacy is "a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions" (Suchman 1995: 574). "Norms, values, beliefs, and definitions" are ideas possessed by actors and embodied in informal (uncodified) institutions. Following Weberʼs (1978: 31-33) argument, the legitimacy of an action is essentially the actor's subjective belief in the validity of the informal institutions (in Weberʼs words, "determinable maxims") toward which the action is oriented. In contrast to the efficiency of an action, which is related to whether an action is worth taking, an action's legitimacy is related to its righteousness.
During the action, actors rely on the ideas embodied in informal institutions, to discover meanings in the incentives and behavioral programs, and find legitimacy in appropriate behaviors. Therefore, the legitimacy of behavior is perceived through the action process whereby actors formulate and modify their subjective understandings of the meanings of the informal institutions and their subjective perceptions of the behavior's validity. In other words, the legitimacy of an action and the corresponding informal institution toward which the action is oriented depend on the actor's perceptions of the informal institution and the action rather than on the forms of the informal institution. Consequently, different actors conforming to the same informal institution may have contradictory ideas about the legitimacy of the same behavior because of their contradictory perceptions of the informal institution. Moreover, the same actors may orient their actions toward formally contradictory informal institutions (Weber 1978: 32).
The legitimacy of an action reflects an actor's subjective belief in its validity (Weber 1978: 31). Consequently, neither the habitual behavior employed unconsciously nor the action taken purely on an efficiency basis is necessarily legitimate (Weber 1978: 31). First, an action's legitimacy is the behavioral characteristic involved in an actor's consideration when they refer to their consciousness to consider whether the action is appropriate. Although a habit unconsciously employed by an actor may initially be legitimate and efficient, it may be followed afterward without referring to the actor's consciousness, even when structural changes can lead to illegitimacy. Thus, habitual behavior is not necessarily legitimate or irrelevant to legitimacy. Second, highly efficient behavior may prompt an actor to engage in it emotionally, regardless of the legitimacy of the action or lack of it. Thus, efficient behavior is not necessarily legitimate.
According to Weberʼs (1978: 36-37) argument, "the sacredness of tradition, " "belief in the legitimacy of a prophet, " "natural law, " "belief in legality, " and so on lead to belief in the legitimacy of action oriented toward informal institutions. Specifically, belief in the validity of informal institutions leads to belief in the legitimacy of their manifested formal institutions and the behavior oriented toward the institutions. Similarly, belief in the legitimacy of the source's authority leads to belief in the legitimacy of its orders (Weber 1978: 37). In addition, during the action, actors may use the ideas embodied in the legitimate informal institutions to legitimize their behavior, reinforcing others' conformity to the behavioral order.
In state extraction, state actors may pursue a decrease in the cost of enforcing constituents' compliance by investing in the legitimacy of the authority, thus improving the efficiency of state extraction. In addition, an actor in pursuit of legitimacy may select a behavior, regardless of its efficiency. For instance, the belief in the Confucian tradition allowed Chinese emperors to choose not to extract maximum revenue from their constituents to adopt "benevolent governance" (ren-zheng) (Rowe 2009: 33).
For example, Emperor Qianlong, a ruler in mid-Qing China, refused to undertake a new nationwide land survey or increase the formal land tax rates to pursue an increase in state revenue, even when there was a fiscal need for greater revenue arising from the country's enlarged population and territory (Elliott 2009: 149). In addition, the emperor decreed a nationwide amnesty of land tax four times, canceled the collection of tribute grain from the country's breadbasket provinces three times, and abolished taxes on disaster-stricken territories and newly cultivated fields several times (Elliott 2009: 151). He chose to distribute the country's wealth among the people to meet their needs rather than maximize the state's revenues (Elliott 2009: 149-151).

Action, time, and event
Action is a temporal process characterized by time (Barbalet 2009a), during which contingent event(s) facilitate(s) an actor, who is faced with qualitative contradictions, to select a specific incentive and a corresponding behavioral program (Sewell 2005). Thus, the process institutionalism model follows Sewell's (2005) "eventful temporality" and involves an understanding of time at the level of subjective meaning. Sewell (2005) argues that events can change the balance of the qualitatively contradictory causal forces imposed by institutions and structures.
Specifically, contingent events may bring about a change in actors' positions on institutions and structures and open new strategic opportunities for their volitional actions (Lachmann 2000;Sewell 2005). Actors' experiences with contingent events may also directly trigger changes in their subjective understandings of structural and institutional characteristics (Kiser and Linton 2002). Either scenario will form or change the contradictions among actors' understandings of structural and institutional characteristics, as well as their conscious behavioral selections and interactions.
Following eventful temporality, the time of action is neither totally consistent nor entirely discrete. Instead, it is nonlinear and divergent. As Sewell (2005: 110-111) noted, "social processes […] are inherently contingent, discontinuous, and open-ended. […] 'Structures' are constructed by human action, and 'societies' or 'social formations' or 'social systems' are continually shaped and reshaped by the creativity and stubbornness of their human creators. " On the one hand, events facilitate actors' selections among qualitatively contradictory incentives and ideas and modify their incentive and idea structures. Thus, this leads to causal heterogeneous temporality (Sewell 2005: 110). On the other hand, an actor's understanding and selection of a specific constraint triggered by an event are inevitably related to the earlier relevant experience and the ideas arising from it, making it partially path-dependent (Sewell 2005: 100-101). In short, the time of an action is not a teleological process manifested in a determinant behavioral choice but rather an indeterminate process involving a cluster of divergent possible behavioral choices.
Indeed, Schumpeter (1991Schumpeter ( [1918: 101) emphasizes the causal role events have played in fiscal history and the opportunities they have provided for exploring the "causal" and "symptomatic" significance of fiscal history.
"Most important of all is the insight which the events of fiscal history provide into the laws of social being and becoming and into the driving forces of the fate of nations, as well as into the manner in which concrete conditions, and in particular organizational forms, grow and pass away" (Schumpeter 1991(Schumpeter [1918: 101; emphasis in original).
Empirically, some case studies demonstrate the specific roles contingent events have played in facilitating actors' selections among available behavioral options at critical junctures toward future institutional changes. For instance, Kiser and Linton (2002) recognize that historical events, such as the Fronde in France, have shaped the relationship between taxation and revolts in the history of state building. They argue that the Fronde helped shape the subjects' considerations of the potential efficiency of behavioral options that were available and thus led to a difference in the relationships between taxation and revolts before the event than after it. Lachmann (2000: 94) similarly argues that the Protestant Reformation "opened a new cleavage in elite interests" and "transformed elite capacities, " thus shaping the courses of state formation in Europe in the sixteenth and seventeenth centuries. Wenkai He (2013) suggests that historical events imposed long-lasting effects on the direction and tempo of subsequent institutional changes, as the socioeconomic structures set crucial constraints on the repertoire of available trajectories.
Methodologically, quantitative analysis and in-depth case studies can be employed to conduct an eventful analysis of the connection between actions in state extraction and institutions. For example, Kiser and Linton (2002: 899) incorporated a dummy variable for the period before and during the Fronde that interacted with taxation into the model, thus exploring the difference in the connections between taxation and the revolt before and after the Fronde. Additionally, He (2013) developed a causal narrative to account for the divergent institutional outcomes of three historical episodes-England (1642-1752), Japan (1868-1895), and China (1851-1911)-prior to which many important points in state formation and market development were shared. The author combined a comparative study with an in-depth historical analysis of the three episodes. He (2013) suggested that the level of concentration of goods, the extent of the development of domestic financial networks, and the other characteristics of the socioeconomic structure presented the available behavioral options for the actors in historical contexts. The credit crisis arising from contingent events opened up the paths of the three countries toward divergent institutional changes, determining the success or failure of modern fiscal state formation.

Summary of preliminary propositions
The process institutionalism model proposes some preliminary propositions for state extraction as follows.
1. State extraction institutions (e.g., form of taxation, contract between rulers and agents) and other formal institutions (e.g., the structure of local organization) may impose qualitatively contradictory interest-specific incentives (e.g., incentive to revenue and incentive to promotion) on the actions of actors (e.g., tax collectors). 2. Informal institutions (e.g., state ideology and religious culture) may impose qualitatively contradictory ideas on actors (e.g., taxpayers) and facilitate contradictory understandings of the legitimacy of an action (e.g., paying taxes). 3. In the process of actions, an actor may be faced with not only contradictions among the incentives in different dimensions (e.g., a contradiction between an interest-specific incentive in the efficiency dimension and a claim to a specific type of legitimacy in the legitimacy dimension) but also contradictions among the incentives in the same dimension (e.g., contradiction among claims to different types of legitimacy). 4. In the process of actions, actors select from the available contradictory incentives and ideas, facilitating changes in state extraction institutions, formal institutions in the broader context, and informal institutions. 5. In the process of actions, contingent events impact actors' perceptions of and selections among the institutional and structural characteristics. 6. In the process of actions, actors' earlier experiences impact their perceptions of and selections among the institutional and structural characteristics. 7. Actors' (e.g., peasants as taxpayers) actions in state extraction are impacted by actions in processes other than state extraction (e.g., land reform). 8. Through actions of state extraction, actors (e.g., farmers) modify their idea and incentive structures and, thus, their consciousness.

Conclusion
This study proposes a process institutionalism model for understanding the relationship between actions and institutions in state extraction. In contrast to existing studies that recognize state extraction as an institutional characteristic of the state, this model recognizes state extraction as a process by which multiple actors choose and take different actions and interact with socioeconomic structures. The model adopts an action-centric approach to emphasize the process of action through which actors not only formulate and modify their consciousness but also facilitate institutional and structural changes in a broader context. By employing this action-centric approach, the model follows Weber's (1978Weber's ( : 4, 2012Weber's ( [1913) sociological theoretical tradition of "the interpretive understanding of social action. " It accomplishes something that previous models in the study of state extraction did not attain.
Specifically, the process institutionalism model, adopting an action-centric approach to state extraction in history, and especially the early modern period, realizes a conjuncture between theoretical and empirical research on the relationships between actions and institutions. Theoretically, Weber (1978Weber ( , 2012Weber ( [1913) and Barbalet (2009a, b) outline the action-centric approach, which, in contrast to the institution-centric approach, permits theoretical accounts of the changes in institutions and the consciousness of actors in the process of actions. Schumpeter (1991Schumpeter ( [1918) and Campbell (1993) empirically demonstrate that actions in state extraction are crucial to changes in state institutions and other related institutions. However, they do not offer a coherent theoretical model for understanding the relationship between actions and institutions in state extraction. By employing an action-centric approach to review the literature on state extraction in history, the present study develops a process institutionalism model that facilitates a combination of theoretical and empirical research on the relationship between actions and institutions.
In addition, the process institutionalism model allows for an understanding of multiple incentives for diverse actors and their actions in the extraction process, in contrast to existing studies that focus on the actions of rulers and higher-level bureaucrats and presume that authoritative actors are persistently driven by the incentive to maximize state revenue while maintaining their reign. Moreover, unlike the institution-centric approach that does not explain the divergent institutional changes arising from similar initial institutional and structural characteristics, the process institutionalism model facilitates understanding the diversity of actions and subsequent institutional changes, beginning from similar institutional and structural characteristics.
Furthermore, by emphasizing the temporal processes of actions, the model involves time and event(s) in the causal analysis of institutional changes. Following Sewell's (2005) eventful temporality, process institutionalism recognizes the time of action as neither totally consistent nor entirely discrete. Rather, it is an open-ended process involving a cluster of divergent possible behavioral choices. Along with Kiser and Linton's (2002) and He's (2013) studies of state extraction that emphasize the role of events in shaping the direction and tempo of institutional changes in state building, process institutionalism makes the indeterminate causal regularities inherent in the connection between state extraction and institutional changes accessible for analysis.
This study attempted to illustrate the theoretical foundation of the process institutionalism model and to pave the way for applying it (and, therefore, the action-centric approach) to the study of state extraction in both Eastern and Western societies. In future research, more attention should be given to how the actions of rulers, officials, heads of local organizations, common peasants, money lenders, and other actors have led to the formations and changes in the tax states across regions and over time. Additionally, the comparative study of the historical processes of state formation across Eastern and Western societies will contribute to an understanding of the mechanisms of fiscal state formation that integrate agency, structure, and contingency into one causal narrative, accounting for institutional changes (Sewell 2005;He 2013: 180).