DEMSTAR

Democracy, the State, and Administrative Reforms

Conceptualizing State Capacity

Mette Kjr and Ole Hersted Hansen

with Jens Peter Frlund Thomsen

DEMSTAR Research Report No. 6

Conceptualizing State Capacity

Anne Mette Kjr and Ole Hersted Hansen with Jens Peter Frlund Thomsen

DEMSTAR Research Report No. 6

Department of Political Science, University of Aarhus April 2002 www.demstar.dk

Table of contents Conceptualizing state capacity .............................................................................. 7 What is the State....................................................................................................... 7 State formation processes ....................................................................................... 8 Empirical state capacity ........................................................................................ 10 The importance of autonomy .......................................................................... 11 Embedded autonomy........................................................................................ 13 Governed Interdependence.............................................................................. 14 State-society synergy......................................................................................... 15 Leadership and governance............................................................................. 17 Summing up: Structure and agency, autonomy and capacity.................... 19 Studying state capacity ......................................................................................... 20 Specifying the components of state capacity ..................................................... 21 Capacity enhancing mechanisms and factors enabling capacity.................... 23 Conclusions............................................................................................................. 25 Notes ........................................................................................................................ 27 References ............................................................................................................... 29

List of Figures Model 1: The dynamics of state formation........................................................... 9 Model 2: Explaining the state in functional and causal terms ........................ 20 Figure 1: Capacity as conditioning state actors. ................................................ 23 Figure 2: Capacity as a dynamic relationship.................................................... 25

List of Tables Table 1: State capacities......................................................................................... 22

Conceptualizing state capacity State capacity is generally defined as the ability of the state to formulate and implement strategies to achieve economic and social goals in society. This definition, however, does not provide guidelines as to how the phenomenon is to be studied. State capacity is not a tangible feature of the state. It is not observable in itself, but its consequences and preconditions can be observed, such as economic growth, the character of policy outcomes, or the system of revenue extraction. The present paper explores state capacity as a concept. It analyzes the various understandings described in existing literature, and the factors that influence state capacity. The aim of the paper is thus explorative; we wish to come to terms with the concept of state capacity in order to reach a better understanding of how it can be studied. The paper is divided into two parts. The first part assesses existing literature on the state. We begin by discussing the concept of state, and how to define the state. We then discuss various approaches to state formation, and we look at more recent attempts to study empirical manifestations of state capacity. The second half attempts to identify the core features of state capacity, and we tentatively set up a hierarchy of state capacities, of which the ability to extract resources and secure public support are essential. Finally, we explore how the concept of state capacity may be further developed, allowing us to ask new questions when studying the phenomenon. What is the State During the heyday of behavioralism and systems theory it was commonplace to prefer the concepts of, say, political system or government to that of state.1 However, in the 1980s the concept of state enjoyed a revival in political science. American scholars came to advance so-called statecentered perspectives, arguing that the state is an important institution when analyzing the course of societal development. The purpose of this introductory section is to discuss definitions of the state. The concept is slippery because the state cannot be defined along a single level of analysis. The state is theoretically as well as empirically a multidimensional phenomenon. Defined in its most abstract sense, the state clearly has to be defined in terms of both structure and agency inasmuch as it embodies particular organizations, rules and human officials. Parliament, constitutional rules and elected politicians are all central to the state apparatus. In methodological terms it must therefore be acknowledged that the doctrines of methodological individualism and collectivism must supplement each other when approaching the state. State officials transform state

organizations into agents by and through their individual or collective actions. State structures, in turn, either facilitate or constrain certain lines of state action. Some statist scholars emphasize the structural characteristics of the state, while other stress agency. Theda Skocpol, for example, defines the state primarily in collectivist and structuralist terms:
a set of administrative, policing, and military organizations headed, and more or less well coordinated by, an executive authority. Any state first and fundamentally extracts resources from society and deploys these to create and support coercive and administrative organizations (Skocpol, 1985: 29).

Another leading statist, Eric A. Nordlinger defines the state in individualistic terms:
The state refers to all those individuals who occupy offices that authorize them, and them alone, to make and apply decisions that are binding upon any and all segments of society. Quite simply, the state is made up of and limited to those individuals who are endowed with society-wide decisionmaking authority (Nordlinger, 1981: 11).

Skocpols definition is essentially structural, whereas Nordlinger defines the state in individualistic terms. There is no major contradiction between their views because the state has to be defined in both holistic and individualistic terms. Skocpol indicates that state organizations cannot be reduced to the individuals that operate within them. Officials come and go, but organizations persist over time. Nordlinger, on the other hand, points to the fact that organizations cannot and do not act: only individuals possess the ability to act and to pursue particular preferences. State officials constantly draw upon resources that are intrinsic to state organizations. Such resources are derived from society at large, or they may derive from outside the system, from territory gained in warfare or from foreign aid. This suggests that in order to discuss the existence of state there must be some sort of internalization of the concept of state in the populace. The state must further be recognized by the international society of states. Sovereignty, autonomy, and legitimacy are therefore inextricably linked in the understanding of what constitutes a state. And, as we shall see, state leaders play an important role in generating legitimacy and they may therefore affect the availability of organizational resources. State formation processes A number of scholars have addressed the issues of state building in order to understand why some state-builders have succeeded and others failed. They

have all tried to come to terms with the fact that states consist of both actors and structures.
Model 1: The dynamics of state formation.

Structural variables underpinning institutions (socio-economic and the international context)

Institutional constraints on rulers

Policy choices of rulers

The model illustrates how the dynamics of state formation are often approached in order to include agency as well as structure in the analysis. The rulers range of policy options is circumscribed by institutional constraints consisting of both formal and informal rules, norms, or patterns of behavior. For instance, if the bargaining power of societal groups vis--vis the state is very weak, the ruler may chose to use a coercive rather than a collaborative strategy in order to collect tax revenue (Levi, 1988; Crone, 1988). And finally, these institutional constraints have been developed within the constraints of endemic structural variables. The modern nation state evolved in the interface of war and capitalist development. The dynamics of inter-state warfare made it necessary to generate domestic revenue and to expand state administration. This advanced the development of weapons technology and production, leading to the expansion of industrialism and capitalism. The nation-state, characterized by authority and administrative capacity, emerged from these mutually reinforcing dynamics between increased surveillance, capitalist and industrial development, and military inter-state violence (Giddens, 1985: 311). The states in this process gradually lost despotic power, i.e. the range of actions the state elite is empowered to undertake without routine, institutionalized negotiations with civil society groups (Mann, 1988: 5). At the same time, the state increasingly gained infrastructural power the capacity to actually penetrate civil society and to implement political decisions throughout the realm (ibid: 5). The impetus to build infrastructural power is thus seen to derive from war: War drives state formation and transformation (Tilly, 1990: 20). Conquest means that state actors expand their administration over new territories. And preparations for war lead to efforts to increase resource extraction. As a by-product of preparations for war, rulers initiated activities and organizations that eventually resulted in the materialization of modern

bureaucracy: courts, treasuries, tax systems, regional administrations, and public assemblies (Tilly, 1998: p. 75). First, a central message of these historical sociological contributions is that they emphasize how state organizations evolve in interaction with their surroundings. State capacity cannot be generated if the state apparatus remains insulated. State capacity is a product of continuous interaction between rulers and citizens, as well as interaction between sovereign states. As we shall see below, the concept of governance may be useful in order to focus on the interface between rulers and citizens and between autonomy and capacity. Second, the contributions strive to illustrate the importance of seeing the dynamics of state capacity as involving both actors and structures. State actors are restrained by path dependencies, but they are also capable of initiating changes that alter the direction of existing paths. Whether they succeed, however, is another matter. It may be argued that when studying changes over long periods of time, the historical sociological approach is appropriate. Studying changes in state capacity over a more modest time span, say one or two decades, requires an approach focusing more on individual agency. The concept of governance may be useful in this regard. Third, if we are to learn from historical observations, the character of economic production matters because it affects the composition of and the way in which fiscal revenue is generated. Rich Arab states, for example, have the wherewithal to provide an array of services, but being sufficiently funded by oil revenues they lack incentives to construct an efficient organizational apparatus to collect domestic revenue. If the oil wells dry up, the state apparatus may collapse. Finally, the international context matters. War, the nature of warfare and the types of relationships with international donors and lending institutions all affect how states relate to their own citizens.2 If, for instance, poor states depend on aid for a major part of their income, they may end up being more accountable to the donors than to their own citizens (Brautigam, 1992; Moore, 1997). The macro-sociological approach brings to attention the international and structural processes in the evolution of the state. Recent studies, to which we shall turn below, have focused more on the specific institutional manifestations of state capacity. Empirical state capacity During the 1980s scholars showed renewed interest in the problems of state capacity in the third world in an effort to distance themselves from the dominant theoretical conceptions of a dependent third world as well as the neo-liberalist interpretation of state failure. Dissatisfied with the tendency in

10

both dependency and modernization theories to reduce politics to socioeconomic variables, and dissatisfied with the neglect of the central roles played by state actors in developing societies, scholars attempted to reassert the importance of the state as an explanatory variable. The general claim is that state actors are able to generate policy preferences that emerge from within the state apparatus, and that they on occasion use state power and resources to pursue their own interests rather than those of domestic or foreign actors.3 The statist perspective is an aggregation of attempts by different scholars to bring the state back in to the analysis of development, and this movement is first and foremost associated with the renowned work of the same title by Skocpol, Evans & Rueschemeyer.4 Skocpol criticizes the tendency in previous approaches to neglect the role of the state as an independent actor. She argues that in the late dependency approaches, many possible forms of autonomous state action are thus ruled out by definitional fiat.5 Functional theories of state tend to portray states as products of society and the concept political system is used more often than the concept of state. Skocpol conceptualizes the state as an actor or an institution that shapes societal outcomes. In this vein the concept of autonomy therefore becomes crucial for understanding why some states, better than others, are capable of transcending societal boundaries and enact their own policies. State capacity is derived from state autonomy:
States conceived as organizations claiming control over territories and people may formulate and pursue goals that are not simply reflective of the demands or interests of social groups, classes, or society. This is what is usually meant by state autonomy. Unless such independent goal formulation occurs, there is little need to talk about states as important actors.6

The central issue that emerges is how the insulation of states and their bureaucracies is linked to developmental success (Leftwich, 1993; 2000). While most of the initial work in the statist perspective tended to equate autonomy with capacity, describing states as either strong or weak due to varying degrees of autonomy, newer versions engage in a more elaborate analysis. The importance of autonomy Empirical scholars have proposed a more elaborate outline of the role of the state by stressing that studies have to be set in a framework describing each states particular relation to society.7 These scholars recognize that strong and autonomous states develop quite differently. The contrast between the successful authoritarian regimes in East Asia and the breakdown of bureaucratic authoritarian regimes of Latin America was a puzzle that did not 11

readily lend itself to be explained by the narrow concept of state autonomy. Confronting the tendency to reduce capacity to the concept of autonomy, Rueschemeyer and Evans emphasize that some states with a high degree of autonomy were seemingly unable to implement and carry out their intended goals. The concept of relative autonomy8 was therefore introduced, both to compensate for the somewhat simplistic characterization of institutions as either autonomous or penetrated, and to demonstrate how some states are able to act effectively as corporate actors involving societal interests.
the states very success in building its role as a corporate actor may undercut its ability to remain autonomous and effective intervention may increase the extent to which the state becomes an arena of social conflict.9

The central issue is the depiction of the state-society relationship as involving a trade-off between an insulated but not necessarily effective state and a less insulated but perhaps more effective state that may risk capture and corruption. There is a clearer recognition that the main issue concerns how such trade-offs take place in diverse societal settings and that the strength of a state is a two-edged concept. The distinction between despotic and infra-structural powers introduced by Michael Mann is in this respect a relevant contribution because it induces scholars to focus on the relation between autonomy and state capacity, and how this relation is conditioned by societal context. The implication of the distinction is that low capacity states rely primarily on despotic powers rather than infra-structural powers because of their inability to build effective state administrations and to impose centralized administration over a fixed territory. This relativist conception of the relation between strength and autonomy has been approached in at least two different ways. One focuses on the central role of the state in creating autonomy from a hostile environment, while the second focuses on states that succeed in creating strong infrastructural powers without losing autonomy. In Strong Societies and Weak States, Migdal (1988) observes that most third world states are weak in the sense that they confront powerful groups in society that resist every attempt at political reform. He concludes that the major impediment to development is the apparent lack of state autonomy and control vis--vis society. His analysis is in that sense a reminder to statist theory of the futility of talking about deliberate state action under adverse societal conditions where the degree of state-ness is severely contested. Another type of approach, represented by first generation developmental state theories,10 focuses on how the East Asian states used autonomy to set coherent developmental goals and discipline a weak industrial sector 12

through active intervention in the market. The developmental success of the East Asian states is predominantly viewed as a result of deliberate state action, where domestic industries are used as instruments to achieve state policy goals. This perspective conforms to the statist perspective by emphasizing that states are strong only when initiating policies independent of major societal actors. The second generation theories of developmental states are reactions to this rather unilateral focus on state actions. Examples are found in the works of Peter Evans and Linda Weiss in which perceptions of the state and its capacities are both moderated and differentiated by providing a version of the relation between the state and industry in the East Asian experience which focuses more on cooperation than coercion. Embedded autonomy Stating that autonomy is one of the central tenets of a strong state, Evans (1995) argues that the relation between state and society goes beyond the claims of statist approaches. He proposes an alternative view stressing the possibility of mutuality between society and the state. Drawing on Granovetters11 concept of embeddedness, Evans hypothesizes that in order to be developmental, states must combine autonomy with close ties with certain parts of society. Ideally such ties provide states with consistent information on the problems of development and how to adapt policies to address these problems, whereas the autonomy of bureaucratic structures from politicians and societal groupings improves coherence, and hence effectiveness, in the implementation of policies. The state must be embedded in some sort of social relation in order to be effective and accepted:
A state that was only autonomous would lack both sources of intelligence and the ability to rely on decentralized private implementation. Dense connecting networks without a robust internal structure would leave the state incapable of resolving collective action problems of transcending the individual interest of its private counterparts.12

Bureaucratic autonomy is according to Evans not derived from being insulated from societal actors. Instead autonomy is a result of the status accorded the central administration because of its high level of coherence and quality. The way in which Evans uses the concept of state embeddedness does not clearly specify the emergence and specific traits of the special relationships between industry and the state. A possible reason for this is applicability. As the concept of embeddedness is constructed on the basis of the experiences of East Asian countries, it requires a higher degree of specification if it is to be used in broader contexts. In a later article,13 Evans offers a somewhat more elaborate explanation of the construction of embeddedness in the 13

East Asian states. Embeddedness emerged because a political leadership was capable of constructing a collective project of national development,14 but he then reveals that the specific development of the close ties was even more convoluted and counter intuitive than the evolution of the bureaucracy itself. The process involved elements of deliberate bargaining between the state and private enterprises that could never have been implemented without a foundation of concrete interaction between economic bureaucrats and corporate managers.15 The argument is ambiguous. The role that states employ is predisposed by the degree of embeddedness, but embeddedness is itself contingent upon politicians and bureaucrats alike that embody these specific roles. Is embeddedness then a result of a progressive leadership, or is it the result of already existing networks of elites sharing interests across the publicprivate divide? Embeddedness is difficult to describe empirically, and it is only in a very few passages that Evans is able to show how it works in practice. It nevertheless seems that the concept of embedded autonomy is an advance over traditional dichotomies like state versus market or state versus society, because it softens the assumption of a direct causal link between autonomy and state capacity. Governed Interdependence In The Myth of the Powerless State, Linda Weiss (1998) attempts to elaborate on Evans concept of embeddedness in relation to domestic industry. Transformative capacity is defined as the ability to device and implement policies that augment a societys investible surplus.16 She endeavors to show that coordination and cooperation between state and society is vital to state capacity. Building on Evans notion of embedded autonomy, she introduces the concept of governed independence (GI), referring to a distinctive kind of institutionalized linkage between state and society that does not focus solely on the strength or embeddedness of a state, but emphasizes how states delegate authority to societal actors. According to Weiss it is the successful integration of industry into the transformative project of the state that allows the state to share more of its powers with industry while retaining its coordinating role,
in a system of GI, as described for Japan and the NICs, the question of who initiates loses much of its meaning and importance. Both the state and industry can and do take policy initiatives, but this takes place within a negotiated relationship in which the state retains a guiding role, exercising leadership either directly or by delegation to industry.17

In a situation of governed independence, a state in effect gains power because of the synergistic links with industry, which makes the implementation of 14

transformative measures easier to accomplish. Whereas in statist theories capacity tends to be rather fixed, transformative capacity is far more volatile. It belongs neither to the state nor society and needs to be reaffirmed on a frequent basis. Using GI, Weiss stresses the possibility of institutionalizing the features of embeddedness, thereby allowing for a development in which societal actors implement developmental goals formulated in conjunction with the representatives of the state. In a developed context of GI, state actors provide guidelines and draw upon the states administrative and technical institutions to provide information and advice in exchange for the commitment of industry to shared goals of development. This institutionalization of contact between state and societal actors is not conditioned by state strength in the statist sense, but is a strategy that places heavy demands on the quality and innovativeness of bureaucracy and state leaders. Thus, Weiss shares with Evans the emphasis on well-developed bureaucracies as a necessary though insufficient condition for cooperation between state and society. State-society synergy It would thus appear that the construction of state capacity is very volatile in the empirical world, and this is paralleled theoretically. Weiss account of the enabling sources is comprehensive and renders the concept of GI contingent upon each case study. In a comparative perspective it is therefore difficult to define and extract the relevant learning experiences of successful transformers. Still, one of the main insights to be drawn from the description of the approaches taken by Peter Evans and Linda Weiss is their emphasis on synergy between states and society. Synergy is a rather new concept in the discussion of state capacity and is found in those cases where state and social forces cooperate to pursue a joint project of economic transformation. The question is whether a strategy of synergy is possible in the process of political development, and hence in other spheres than the purely economic? After all, the role models of East Asia are authoritarian states that to some extent were able to ignore demands from non-industrial societal groupings and in fact repressed parts of civil society. Does this limit the argument as stated in the accounts of Evans and Weiss, or are the structures of embeddedness or governed independence applicable to other relationships between state and society? Evans addresses this question by showing that embeddedness is not always an exclusive feature of relations between industrial elites and the state. In fact he hypothesizes that by extending statesociety links to a broader

15

base of societal groups, the embedded autonomy of the state tends to become more robust and adaptive.18 Robert Putnam draws similar conclusions. He argues that the success of democracy rests upon norms of reciprocity and networks of civic engagement across the public-private divide (Putnam, 1994). The synergy described by Putnam represents a relationship in which a vibrant and organized civil society facilitates the resolution of collective action problems. Putnams approach thus counters the strong society vs. weak state notion embraced in the statist theories. Whereas Putnam primarily observes a causal effect from a vibrant civil society to the effectiveness of the state, others, for instance Elinor Ostrom (1996), have emphasized the importance of state institutions devoted to cooperative endeavors with a civil society that may be characterized by different levels of organization. In her contrasting account of the construction of urban infrastructure in Nigeria and Brazil she shows how different strategies chosen by public officials toward civil society shape the results of such projects. Brazilian public officials embarked on a strategy of co-production by which they encouraged public participation both in the planning and implementation stages of safe-water projects, and as a side effect they created local organizations to coordinate actions between neighborhoods, whereas the process of implementation in the case of Nigeria devolved solely upon bureaucracy. It is a useful example of mutual empowerment in which the state, on the one hand, delivers services more effectively, while on the other enhancing its infrastructural powers by stimulating the creation of civil organizations that are tied to the state. Likewise, in her analysis of successful government programs in the Brazilian state of Ceara, Judith Tendler shows how public officials enhanced municipal performance by creating and strengthening local civic organizations that worked in programs designated by local governments (Tendler, 1997). Her main assertion is that frequent interaction between all levels of government rather than historical circumstances empowered the community to play an important role. These examples all subscribe to the view that the establishment of reciprocity between state and society is important,19 and they thus transcend the traditional zero-sum vision inherent in older theories. In most of these accounts of synergy it seems that the state is able to initiate cooperation even in cases where the basic social conditions and previous historical records did not facilitate such cooperative efforts. In contrast to Migdals somewhat unilateral view of social control mechanisms a variety of methods for achieving infrastructural powers are given here. The underlying problem, of course, is how to scale up these findings to provide general guidelines on how to 16

construct such mutuality on an aggregate level. Staying on the level of small islands of synergy, as in the case of Ceara in Brazil, is useful to illustrate the logic of the argument, but can these lessons be transferred to large-scale problems such as legitimating tax regimes or ensuring broad compliance on government projects? The problem of how to scale up local social capital assets relates closely to the final issue of discussion in this section, that of leadership and governance. Leadership and governance A problem inherent in these stories of embedded autonomy, governed interdependence, and state-society synergy is how these cases of good government arose in the first place. Robert Putnam argues, for instance, that state capacity appears whenever society is strong, thus excluding state agency as an explanatory factor (Tarrow, 1996: 395). Likewise, Peter Evans embedded autonomy appears to have evolved out of a preexisting culture of elite career patterns and networks. However, Evans (1995: 52) does note the importance of the stark difference between Parks and Rhees leadership in South Korea, although he does not acknowledge this importance theoretically. The term governance has recently emerged as an important concept in the social sciences. Governance appears to be useful when discussing state capacity because it is situated between structures and actors, state and society. The focus on governance acknowledges individual agency without disregarding preexisting patterns of behavior. And the concept of governance transcends the state-society division because it focuses on the formulation of the political rules of the game and the actors involved in this continuous process of formulating and changing the rules (Feeny, 1988). However, no consensus on the exact definition of governance exists, as the examples below illustrate:
*Governance signifies a change in the meaning of government, referring to a new process of governing; or a changed condition of ordered rule, or the new method by which society is governed (Rhodes, 1996: 652). *Governance is the stewardship of formal and informal political rules of the game. Governance refers to those measures that involve setting the rules for the exercise of power and settling conflicts over such rules (Hyden, 1999: 185). *Governance is the exercise of authority with or without the formal institutions of government (Rosamund, 2000: 109). *Governance is the capacity of government to make and implement policy, in other words, to steer society. (Pierre and Peters, 2000: 1).

17

*Governance is the manner in which power is exercised in the management of a countrys economic and social resources for development (World Bank, 1989: 60). *Governance is the institutional capability of public organizations to provide the public and other goods demanded by a countrys citizens or their representatives in an effective, transparent, impartial, and accountable manner, subject to resource constraints (World Bank, 2000: 48).

Some of these definitions are largely identical to the definition of state capacity applied in this paper: the ability to formulate and implement policies. But most of the definitions are qualitatively different in that they seek to transcend the public-private or state-society divide and instead focus on networks of groups and individuals affecting public policy formulation and public outcomes.20 Such an understanding of governance allows us to understand existing patterns of behavior while acknowledging that such patterns may be affected by individuals in times of change or when windows of opportunity arise, as they did when colonies achieved independence in the 1950s and 1960s, or when the Berlin wall crumbled.21 Adopting a governance lens also confers the advantage of being able to identify the specific rules and actors that affect a particular outcome. For instance, in weak states formal rules rarely guide individual behavior. When studying such states, individual leaders and informal rules should be in focus (Jackson and Rosberg, 1982). A governance approach may thus pave the way for an analysis of the kind of politics that affects institutional outcomes, providing an understanding of coalitions of interests and ideologies affecting and changing the capacities of the state. Merilee Grindles (1996) analysis of state capacity in Latin America and Africa deserves to be highlighted in this context, because it has the quality of a governance approach: She operates with four dimensions of state capacity: Institutional capacity describes states having authoritative and effective rules of the game to regulate economic and political behavior. Technical capacity is the ability to formulate and manage macro-economic policies. Administrative capacity describes effective administration of basic physical and social infrastructure the ability to perform basic administrative functions essential for economic development and social welfare. Political capacity consists of effective and legitimate channels for societal demand making, representation and conflict resolution and having responsive political leaders and administrators (ibid. p. 8). Grindle hypothesizes that Latin American and African states have responded to the economic crisis of the 1980s by improving technical capacity, while in general the other capacities have deteriorated. Improved technical capacity reflects the IFIs demands that more technocrats 18

be hired to manage the macro-economy. With regard to the other capacities, Grindle sees the economic crisis as having increased conflicts over the rules of the game, weakened administrative ability to deliver basic services and carry out normal functions of government, and limited the responsive capacities of state leaders and managers (1996: 8). Grindle thus analyses how state leaders have responded to the economic crisis and she manages to incorporate autonomy (political and institutional capacity) and capacity (administrative and technical capacity) as well as structure and agency in a telling analysis of state responses to economic crisis. Summing up: Structure and agency, autonomy and capacity Most of the contributions referred to in this paper deal with the issues of autonomy and capacity, although in different ways and with varying views as to how the two are interrelated. Some approaches emphasize the importance of establishing autonomy before capacity can be built. The stories of statesociety synergy however illustrate that capacity can develop in situations where public authorities are far from being insulated from their surroundings. Do states need to consolidate their capacities prior to developing the ability to act autonomously toward their societies? Or is the opposite logic, that autonomy is a precondition for building state capacity, understood as keeping societal actors at an arms length, a more accurate description? What do we mean when discussing autonomy: are we prioritizing the autonomy of bureaucracy either internally or externally, or the autonomy of the leaders that allows them to formulate their own policies? Evans and other authors work on East Asian states do not distinguish between political and bureaucratic autonomy. An autonomous and coherent civil service would not make much difference if a government were unable to assert its autonomy towards non-state actors. However, bureaucratic autonomy on its own does not guarantee state autonomy. Weiss argues that autonomy is a poor predictor of capacity. History is strewn with examples of autonomous states that failed to develop capacity (Weiss, 1998: 29). Czarist Russia is an example of such an overly autonomous state. In Michael Mann's words, Russia had much despotic power; i.e. the ability to implement decisions without routinized negotiations with groups in society. According to Mann, as states develop they tend to lose despotic power but gain infrastructural power; in other words, states lose autonomy but gain capacity. It thus seems that there may be a threshold of autonomy: below this threshold capacity will deteriorate due to vulnerability to particularistic interests. Above it, capacity will deteriorate because the state will become

19

overly autonomous, coercive or outright repressive and thereby fail to secure the compliance of societal groups. Therefore, on a general level, there is no trade-off between autonomy and capacity. Some autonomy from societal interests is necessary for state actors to carry out important functions. But an overly autonomous state risks cutting itself off from important information needed to make enlightened policy decisions, and it risks failing to secure compliance with those policies. Thus, when autonomy is low, a higher degree of insulation will tend to increase state capacity. But if autonomy is already high, additional autonomy will not increase state capacity. The challenge becomes one of identifying the specific type of autonomy characterizing the state in question.22 State capacity, then, is the ability of states to formulate and implement policies. Studying the presence or absence of this ability in a particular state, as well as how this ability evolves, requires that the most relevant actors, institutions and structures in a given context be identified. It also requires recognizing that the state relates to external as well as domestic pressure, and that the state may possess autonomy in relation to both societal and external actors. How these aspects may be identified is the focus of the next section. Studying state capacity At a general level it can be argued that the state is related to society in two rather different ways. This is illustrated in model 2. The model should not be interpreted as a defense of functionalist arguments in the sense that societal needs automatically translate into public policies. The point is a rather different one. Societal needs exist at any given point in time, but the extent to which these needs will be met depends on the amount of capacities available to state actors. Capacities constitute the crucial feedback mechanism between state and society. If state actors command few resources they will either respond passively to particularistic societal demands or be unable to enact policies that promote economic and social development (Migdal, 1988).
Model 2: Explaining the state in functional and causal terms State Functional Logic: State responding to Causal logic: State initiating change Society

change

20

State actors need more than a strong will in order to promote societal development. State actors in very poor countries, for instance, cannot provide adequate health care and public education when fiscal resources are virtually nonexistent. Weak states often find themselves caught up in a functional logic that prevents them from actively initiating change. They are able to stay afloat, but unable to navigate (Jackson and Rosberg, 1982). In highly industrialized countries states also face various functional needs, but a general feature of these states is that they are also capable of changing and modifying the course of societal development. Given this difference, however, there is room for alternative ways of utilizing and improving the capacities at hand, as was argued in the literature referred to in the above. Moreover, state actors may succeed in improving some capacities while others deteriorate (Grindle, 1996, Weiss, 1998). Specifying the components of state capacity The statist and neostatist literature does not explicitly discuss the relative importance of different capacities. However, the foregoing analyses suggest that a set of capacities that lie at the core of any state apparatus may be identified. Table 1 distinguishes between central capacities and capacity generating arrangements. The first column lists the features that have historically proven beneficial to states as pointed out in the previous section. The evolution of the state and its central capacities is a dynamic and relational process and can best be described as an outcome of the interaction of existing capacities, different capacity enabling mechanisms and capacity enabling factors. An example of the characteristics that the three elements may consist of is found in Table 1. The direction of causality is not clear, as already described in the analysis of the existing literature in the previous sections. The main point to stress in this context, however, is that the purpose is not to identify specific state actors but rather to explain the presence of specific capacities in terms of the relational developments over time. The ability to mobilize fiscal revenue is the bottom-line of state capacity (Levi, 1988; Therkildsen, 2000). In order to attract elected and appointed officials the state must have revenue-generating capacity to cover the public payroll. Likewise, there can be no public sector and production of services without adequate and permanently in-flowing revenues. The amount of fiscal revenues extracted from the economy and the methods of extraction evidently vary between nations: states with limited fiscal resources face many unthinkable options; in fact, most of the essential welfare services within education, health, pensions, infrastructure, and social security cannot

21

be provided by the state if public revenues are small (Therkildsen, 1995). However, the ability to generate revenue is itself contingent upon other factors, such as the level of societal organization and technological and organizational possibilities, the absence of external threats to the security of the state, and the character of governance.
Table 1: State capacities Central capacities Fiscal revenue creation Capacity enhancing mechanisms Capacity enabling factors Governance generating legitimacy, trust and reciprocity Growing private sector, internal organization High levels of education Level of societal organization External threats to the state Position in the international division of labor

Close ties or corporatist Principles of bureaucratic arrangements (conflict solving capacity) i.e. organization institutionalized Specialized and bargaining, types of differentiated governance of the market administrative system system Strong Treasury control Empowered local of public finances authorities (delivery of (internal state control services) Outward oriented Flexible public sector military defense (national Leadership, role of autonomy politicians (governance) Types of ties with external actors and institutions

States that are capable of formulating and implementing important policy decisions are characterized by a number of administrative and organizational features. Elected officials can do little without a permanent staff of highly skilled appointed officials. Appointed officials recruited on merit criteria are essential at most policy stages as they prepare and design policy and also often are involved in the implementation process. Administrative capacity in turn requires specialization and a clear division of labor within the central administration (Weber, 1978; Evans, 1996). Judicial matters cannot be addressed properly without a justice department, and likewise, taxation issues require units of experts. The specific manifestation of these organizational features can vary. Evans and Weiss both stress the Weberian traits of bureaucratic organization as indispensable conditions for capacity, but this is by no means an exhaustive criterion for success. Later works on the future of the developmental state tend to stress flexibility and decentralization of the state apparatus as a central to capacity in the course of acting in ever-changing surroundings (ORiain: 2000). 22

Military force is the means by which the state eliminates internal and external threats to its authority. As sketched above, inter-state war and preparation for war have been linked to domestic revenue generation and institutionalized bargaining (Tilly, 1990). War has served to promote not only domestic capacity building processes but also to secure internal pacification. A functional dimension of capacity might thus be an outward oriented military force (whose primary aim is defense) rather than an inward oriented military force (whose primary aim is to solve civil disorder) (Giddens, 1985). From a strictly legal point of view courts are not part of the state. In a functional sense, however, it is clear that the state is highly dependent upon an effective judicial system. The judiciary enhances state autonomy when solving conflicts and effectuating sanctions in relation to unlawful behavior. An independent judiciary also serves to hold state actors accountable for their doings. In sum, states that cannot rely upon an effective judiciary will most likely find themselves in a context of evasive and unpredictable behavior. Capacity enhancing mechanisms and factors enabling capacity Table 1 lists the elements of state capacity in a rather static manner. A more dynamic model of the causal relationships between state, capacities, and societal development is developed in Figure 1. The central logic of the model is that assuming state actors have preferences concerning the development of their societies; they will be constrained by their ability to put policies into effect. This ability is contingent upon the functional dimensions described above. It is pivotal for understanding capacity that it refers to ability and not to fixed entities or structural features of the organization.
Figure 1: Capacity as conditioning state actors.

STATE ACTORS

Preferences Abilities Capacity

Societal Development

If a dynamic perspective is applied to account for the evolution of capacities the picture evidently becomes more detailed. As depicted in the second column of Table 1, state capacity develops in a dynamic process involving 23

the interaction of capacity enhancing mechanisms that to a certain degree depend on country specific path dependencies. For example, most developed states have created and relied upon a plethora of corporatist bargaining structures. There may be many reasons for this, but state actors see corporatist arrangements as attractive solutions to distributional conflicts. State actors may be forced to do something they would not otherwise do, but by making compromises with vested interests they transform a political constraint into a facilitating capacity: Thus, a policy of non-consultation may eventually lead to less state control as vested interests will use their unique power to affect implementation (Haggard and Kaufman, 1992, 1995). The content of the second column in Table 1 should is not an attempt to develop an exhaustive list of these mechanisms but should be interpreted as an illustration of the general logic of the argument. Some of these mechanisms are more generally applicable than others, and their impact on capacity may vary, but the central idea is that state actors have the possibility of selecting mechanisms that enhance capacity, and these mechanisms can become more or less institutionalized as bases of capacity for the state to draw upon. Figure 2 demonstrates this logic by linking state actors and the capacity enhancing mechanisms. The model it also indicates (see Table 1, right hand column) that the notions of state capacity and political legitimacy are inextricably linked. A state is powerful only if it enjoys the support of the majority of citizens within its territory. The existing level of support may again be contingent upon the character of governance: whether the political rules of the game are designed so as to legitimize the regime. Consequently, a distinction can be made between the state, the regime and the government. The state as an organization is more permanent than the regime. Regimes may collapse (as was the case with some of the post-communist regimes) while the state persists (Hyden, 1999). Regimes are more than governments because a government can lose legitimacy, while regime rules are still accepted. In many weak states, however, the distinction between state, regime and government is blurred. If the incumbent leadership fails to design rules of succession that reflect a compromise between various powerful interests, the legitimacy of the regime and the state may become weakened and trigger a downward spiral that severely hampers capacity building. Column three also indicates that state capacity is contingent upon societal conditions such as education, size of private economy and organizational density, and that these rather fixed entities are formed by general societal development. However, the policy choices of the incumbent leadership may set in motion new directions of social and economic development. The 24

crucial aspect is therefore the extent to which state actors are able actively to shape development, which places the left side of the model at the very heart of the analysis.
Figure 2: Capacity as a dynamic relationship.

Political Regime

Preferences

Society Abilities

STATE

Capacity

Degree of autonomy

Capacity enhancing mechanisms

Capacity enabling societal conditions

Conclusions Table 1 provides a rather static snapshot of state capacity while the two models reflect more dynamic pictures of state capacity, but all can be used for empirical analysis. Thus, if the aim is to compare the degree of state capacity across a range of countries, the snapshot approach would be useful, and indicators such as level of economic development, resource mobilization, and bureaucratic quality would be relevant. If however the aim is to identify processes of change within a defined time span, an approach focusing on the dynamics of institutional change should be adopted. Such an approach should identify the institutional setup represented by examples of capacity enabling mechanisms. Relevant actors and interest groups that affect the outcome in question should also be identified. The concept of state capacity is too broad to function as an adequate guide to country case studies. Students of state capacity would be well advised to look to scholars such as Weiss or Evans, who focus on sector specific measures of capacity, or to Merilee Grindle, who discerns four specific dimensions of state capacity.

25

It is nonetheless possible to identify capacities that are central if a state is to function as such. Model 1 attempts to identify just such core capacities. In any study there will be a trade-off between detail and generalizability. If the ambition were to reach conclusions about the overall capacity of the state in question, an approach similar to that of Merilee Grindle would be recommendable. However, such an approach cannot be very detailed. Another viable solution would be to choose a policy area that could represent a crucial case for analyzing state capacity. This can be done in two ways (Nrgaard and Mller, 2001). A policy area could be chosen in which the state has a long tradition for good performance. If the state proves unable to handle policy in this area, it may not sufficient the capacity to manage policy in other more demanding areas. Alternatively, a policy area could be chosen in which the state is exposed to many hazards, international pressures, strong particularistic interests, technical complexity, and rapid development. If the state handles these challenges well, it may be capable in other sectors as well. Country case studies would require that relevant structures, institutions and actors be identified. For instance, if the ability to collect revenue were to be scrutinized, the appropriate structures to study would be the mode of production, the size of taxable surplus and the tax base in general. Relevant institutions would be the history of taxation and existing rules and regulations within the field. Actors to be studied include tax collectors, ministers of finance and heads of revenue authorities, as well as politicians, the political elite and societal organizations. If, on the other hand, the object of study is a states ability to deliver social services, the relevant structures would be the physical infrastructure for service provision, institutions would include the history of service provision, and the actors would include service providers, health workers, teachers, and top officials in the relevant ministries, the elite and the leadership. In third world countries, donors should also be included as they play a large role in service provision. The main lesson when undertaking study of state capacity is, however, to focus on output as well as outcomes, that is, to devote attention, on the one hand, to the capacity to shape policy on the political and administrative levels, while on the other focusing on how these policies are brought to realization in the society.

26

Notes
See, for instance, Binder et al. (1971), Easton (1979), or Almond and Powell (1978). 2. For an authoritative and recent account of how international relations and domestic politics interact, see Srensen (2001). 3. See for example Bates (1981). 4. Evans, Rueschemeyer, and Skocpol (1985). 5. Ibid., p. 5. 6. Ibid. 7. Katzenstein (1985), Crone (1988). 8. The concept of relative autonomy is a Marxist notion first introduced by Friedrich Engels and Nicos Poulantzas. It was used in late dependency theories to describe the relation between center and semi-peripheral states in the world economic system. 9. Ibid. 10. Amsden (1989), Wade (1990). 11. Granovetter (1985) uses the concept of embeddedness to show how the behavior in institutions on the market are affected more by the nature of social relations than by the universalistic tenets of human behavior described in neoclassical economic concept. 12. Evans (1995, p. 12). 13. Evans (1998). 14. Ibid., p. 74. 15. Ibid., p. 76. 16. Ibid., p. 5. 17. Ibid., p. 71. 18. Evans (1995, p. 228). 19. For further readings confirming these trends, see Evans (1996), Fox (1996), Lam (1996), and Heller (1996). 20. Rhodes (1996) distinguises between five types of governance: governance as the minimal state (IMF and World Bank in the 1980s for example); governance as corporate governance; governance as new public management; governance as a socio-cybernetic system (Weiss GI); and governance as autonomous, organizational networks (the examples of synergy). These perceptions have been criticized, for example by Leftwhich (1993), for not acknowledging politics and interests. The concept of governance adopted in this paper is distinct from the concept of state capacity precisely because it includes politics and thereby agency and the affect of individual choices on state capacities. 21. Deborah Brautigam uses a governance lens when in her account of the Mauritius miracle she argues that because institutions tend to change slowly, the institutional legacies of previous regimes as well as the institutional choices made at the time of transition to democracy tend to structure the political options and
1.

27

development strategies of subsequent governments. However, at certain junctures individuals may be able to affect the direction of future institutional paths. She argues that three initial factors may explain why efficient and responsive states developed in some post-colonial states but not in others. These factors are (i) an institutional design that promotes coalition building and compromise, (ii) an initial set of economic and social policy choices that creates constituencies for broadbased growth, and (iii) limits on the power of landed elites and the military (Brautigam, 1996). 22. Theda Skocpol, for instance, defines state capacity as the ability of states to implement goals, especially over the actual or potential opposition of powerful social groups or in the face of recalcitrant socioeconomic circumstances (Skocpol, 1985:9 quoted in Weiss, 1998: 27, Weiss emphasis). In accordance with the argument above, we can add that Skocpols definition is only valid below the autonomy threshold, not above it.

28

References Almond, G. and Powell, G. (1978). Comparative Politics , Boston, Little, Brown. Amsden, Alice H. (1989). Asias Next Giant. New York: Oxford University Press. Bates, Robert H. (1981). Markets and States in Tropical Africa: The Political Basis of Agricultural Policies, Berkeley, University of California Press. Binder et. al (1971). Crises and Sequences in Political Development, Princeton, NJ: Princeton University Press. Brautigam, Deborah (1992). Governance, Economy, and Foreign Aid, Studies in Comparative International Development, Vol. 27, No. 3, pp. 3-25. Brautigam, Deborah (1996). State Capacity and Effective Governance, in Benno Ndulu and Nicolas van de Walle (eds.), Agenda for Africas Economic Renewal. Transaction Publishers. Crone, Donald K. (1988) State, Social Elites and Government Capacity in Southeast Asia, World Politics, Vol. 40, No. 2, pp. 252-268. Easton, David (1979). A Systems Analysis of Political Life, Chicago: The University of Chicago Press. Evans, Peter, Dietrich Reuschemeyer, and Theda Skocpol (eds) (1985). Bringing the State Back In. New York: Cambridge University Press. Evans, Peter (1995). Embedded Autonomy. States and Industrial Transformation. Princeton: Princeton University Press. Evans, Peter (1996). Government Action, Social Capital and Development: Reviewing the Evidence on Synergy in World Development, Vol. 24, No 6, pp. 1119-1132. Evans, Peter (1998) Transferable Lessons? Re-examining the Institutiotnal Prerequisites of East Asian Economic Policies in Journal of Developmental Studies, Vol. 34, No. 6, pp. 66-86. Feeny, David (1988). The demand for and supply of institutional arrangements in V. Ostrom et al. (eds.) Rethinking Institutional Analysis and Development: Issues, Alternatives and Choices, San Fransisco. Fox, Jonathan (1996). How Does Civil Society Thicken? The Political Construction of Social Capital in Rural Mexico in World Development, Vol. 24, No. 6, pp. 1089-1103. Giddens, Anthony (1984). The Constitution of Society. Cambridge: Polity Press. Giddens, Anthony (1985). The Nation-State and Violence. Berkeley: University of California Press. Granovetter, Mark (1985). Economic Action and Social Structure: The Problem of Embeddedness in American Journal of Sociology, Vol. 91, No. 3, pp. 481-510.

29

Grindle, Merilee S. (1996). Challenging the State: Crisis and Innovation in Latin America and Africa, Cambridge: Cambridge University Press. Haggard, Stephan and Robert R. Kaufman (eds) (1992). The Politics of Economic Adjustment. Princeton: Princeton University Press. Haggard, Stephan and Robert R. Kaufman (1995). The Political Economy of Democratic Transitions. Princeton: Princeton University Press. Heller, Patrick (1996). Social Capital as a Product of Class Mobilization and State Intervention: Industrial Workers in Kerala, India, World Development, Vol. 24, No. 6, pp. 1055-1071. Hyden, Goran (1999). Governance and the Reconstitution of Political Order, in Richard Joseph (ed). State, Conflict, and Democracy in Africa, Boulder: Lynne Rienner. Hyden, Goran (2000). The Governance Challenge in Africa, in Hyden, Olowu and Ogendo (eds.), African Perspectives on Governance, Africa World Press, Inc. Jackson, Robert H. and Carl G. Rosberg (1982). "Personal Rule in Black Africa: Prince, Autocrat, Prophet, Tyrant", Berkeley: University of California Press. Katzenstein, Peter J. (1985). Small States in World Markets. Ithaca NY: Cornell University Press. Lam, Wai Fung (1996). Institutional Design of Public Agencies and Coproduction: A Study of Irrigation Associations in Taiwan in World Development, Vol. 24, No 6., pp. 1039-1054. Leftwhich, Adrian (1993). Governance, Democracy and Development in the Third World, Third World Quarterly, Vol. 14, No. 3. Leftwich, Adrian (2000). States of Development, On the Primacy of Politics in Development. Polity Press. Levi, Margaret (1988). Of rule and revenue. Berkeley, CA: University of California Press. Migdal, Joel S. (1988). Strong Societies and Weak States. State-Society Relations and State Capabilities in the Third World. Princeton, NJ: Princeton University Press. Mann, Michael (1988). The Autonomous power of the State: Its origins, mechanisms and results, Michael Mann (ed): States, War and History, Oxford: Blackwell Publishers. Moore, Mick (1997). Death without taxes: Democracy, state capacity and aid dependence in the fourth world in G. White and M. Robinson (eds), Towards a Democratic Developmental State, Oxford University Press. Nordlinger, Eric A. (1981). On the Autonomy of the Democractic State. Cambridge: Harvard University Press.

30

Nrgaard, Ole and Luise Pape Mller (2001), Telecom development and state capacity in transition countries. A framework for analyses. DEMSTAR Research Report No. 5. Aarhus: University of Aarhus, Department of Political Science. Ostrom, Elinor (1996). Crossing the Great divide: Coproduction, Synergy and Development, World Development, Vol. 24, No. 6., pp. 1073-1087. Pierre, Jon and B. Guy Peters: Governance, politics and the State, London: St. Martins Press, 2000. Putnam, Robert D. (1994). Making Democracy Work; Civic Traditions in Modern Italy, Princeton: Princeton University Press. Rhodes, R.A.W (1996). The New Governance: Governing without government, Political Studies, Vol. XLIV, 652-667. ORiain, Sean (2000). The Flexible Developmental State: Globalization, Information Technology, and the Celtic Tiger, Politics and Society, Vol. 28, No. 2, pp. 157. Rosamund, Ben (2000). Theories of European Integration. London: Macmillan Press. Singh, J.P. (1999). Leapfrogging Development? The Political Economy of Telecommunications Restructuring. Rochester, NY: State University of New York Press. Skocpol, Theda (1985). Bringing the State Back In: Strategies of Analysis In Current Research in Evans et al. (eds) Bringing the State Back In. New York: Cambridge University Press. Srensen, Georg (2001). Changes in Statehood. The Transformation of International Relations. London: Palgrave. Tarrow, Sidney (1996). Making Social Science Work Across Space and Time: A critical Reflection on Robert Putnams Making Democracy Work, American Political Science Review. Vol. 90, No. 2, June: 389-397. Tendler, Judith (1997). Good Government in the Tropics, Baltimore and London: The Johns Hopkins University Press. Therkildsen, Ole (1995). Civil Service Reform from Below: Extreme Resource Scarcity and Unrealistic Performance Requirements in Msambichaka et al. (eds) Structural Adjustment Programmes in Tanzania. Successes, Failures and New Perspectives, Dar es Salam: Economic Research Bureau, University of Dar es Salaam. Therkildsen, Ole (2000). Comparative studies of relevance to resource mobilization and democratization in African countries: a critical review. Paper presented at the conference The State under Pressure, The Norwegian Association for Development Research, 5-6 October 2000, Bergen.

31

Tilly, Charles (1990). Coercion, Capital and European States, AD 990-1992. Oxford: Blackwell Publishers. Wade, Robert, 1990. Governing the Market. Princeton NJ: Princeton University Press. Weiss, Linda (1998). The Myth of the Powerless State. Governing the Economy in a Global Era. Polity Press. Weber, Max (1978). Economy and Society. An outline of interpretive sociology. Guenther Roth and Claus Wittich. Berkely, CA: University of California Press. World Bank (1989). Sub-Saharan Africa: From Crisis to Sustainable Growth, Washington D.C.: The World Bank. World Bank (2000). Can Africa Claim the 21st Century? Washington D.C.: The World Bank.

32