Search

For governments within many modern welfare states it has become a key challenge to guarantee the quality and accessibility of social service provision to citizens. Under the third-party government in Belgium, this implied that far from producing the lion’s share of these services themselves, governments increasingly involved private nonprofit organizations (NPOs) to implement those policies (Salamon et al., 1999). These NPOs have become appealing partners due to their bottom-up nature, a position close to clients and their rather small scale of operation. Hence, it is expected that NPOs have greater opportunities for tailoring services to citizens’ needs and are better able to perform a radar function by signalling new or rather ‘unconventional’ needs as well (Boris and Steuerle, 1999; Salamon, 1995). As a result, the focus of government shifted from ‘rowing’ to ‘steering’ (Osborne and Gaebler, 1992). This implies a separation of the process of planning social policies from the actual implementation of those policies by NPOs (OECD, 2012). In essence, we argue that governments hereby develop a set of strategic management practices in order to steer and control the NPOs that carry out policies, while equally granting them a meaningful voice in the development of new policies as well.

In contemporary welfare states, governments increasingly rely on private actors for delivering public welfare services to citizens (Snavely & Desai, 2001; Salamon, Sokolowski et al., 2004; Anheier, 2005).1 Due to their grass-roots bottom-up nature and rather small scale of operation, non-profit organisations (NPOs) have become appealing partners for public policymakers who wish to involve the organised civil society in the policy process. The argument is that NPOs, by their nature of being exponents of organised civil society, have greater opportunities for tailoring services to clients’ needs and are better able to influence local social behaviour as well (Salamon, 1995; Boris & Steuerle, 1999). For governments that engage with the organised civil society, the key challenge then becomes one of ensuring that NPOs remain publicly accountable for the deployment of sometimes very substantial public funds. Still, this may not reap some of the indisputable benefits attributed to NPOs’ activities that to a large extent derive from their flexibility and autonomy of not just being another arm of government (Huxham, 1995; Boyle & Butler, 2003). This chapter approaches the above issues of NPOs’ accountability and autonomy from the perspective of NPOs involved in publicly funded welfare programs. A large of amount of scholarship has already focused on the impact of governmental interference on the NPOs’ functioning and autonomy, but this discussion remains far from being settled (Toepler, 2010)

Civil society is dead; hurrah for civil society! This simple statement describes the sociopolitical debate in many modern welfare states. Traditional forms of (ideologically or religiously driven) social organisation have gradually declined. Various drivers behind this development have been identified, notably individualisation and changing life patterns, a point to which we will come back later in this chapter. But states have also arguably played a role in this. Welfare-state growth propelled an era of ‘big government’ during the 1970s and 1980s. Although the ‘crowding out’ thesis of a direct trade-off between the size of government and the size of civil society has been largely disproven (see e.g. Salamon, 1995), the collaboration between civil society and the state in the provision of social services has led many organisations to lose contact with their traditional member groups and, by implication, their original legitimacy.

In this chapter we formulate hypotheses on variations in autonomy, control and internal management among agencies, based on three perspectives: (1) a structural-instrumental perspective (emphasizing formal organizational structure and rational choice of actors), (2) a cultural-institutional perspective (emphasizing organizational culture and organizational path dependency) and (3) a task-specific perspective (emphasizing the characteristics of organizational tasks). The environmental perspective (emphasizing institutional environments, isomorphism and myths) is considered mainly in Part III and functions here as a null hypothesis.

In previous chapters we discussed the features of the ideal-type agency as elements within the NPM doctrine, and asserted that there are international pressures for similarity and convergence towards this NPM model. In our model, we expect that these doctrines are being transformed when they meet state-specific polity, administrative culture and actor constellations. This transformation may lead to different responses in different states (state-level pressures for divergence and dissimilarity).

This chapter addresses the basic characteristics of our comparative research design; the subjects studied (population), the measurement instruments used (survey and document analysis) and procedures chosen for analyses (types of statistical techniques used). The first part describes the survey populations in Flanders, Norway and Ireland, based on the understanding of what constitutes a public sector agency as outlined in Chapter 2. Choices made for the purposes of enhancing the comparability of our three samples are described. We then focus on the design of the surveys in the three states, their particularities and similarities, and how we proceeded in constructing a common data set comprising similar information for all three states stemming from these surveys. This process included the conscious identification and selection of comparable data, and the exclusion of data that were not considered to be directly comparable. Lastly, the chapter addresses the various types of statistical procedures undertaken when comparing agencies in the three states, and their inherent strengths and shortcomings.

In this chapter we bring the analysis to its final stage. Having already analysed the influence of state-level and agency-level characteristics separately, we now focus on the effects of interaction between the two. More specifically, we study those agency-level characteristics that affect the autonomy, control and internal management of agencies in a direct way, independently from the influence of state-level characteristics. Integrating the findings on state-level characteristics outlined in Chapter 9, we now know which characteristics, both at state and agency level, make an impact directly and independently from one another. We discuss whether, and to what extent, the structural-instrumental perspective, the cultural-institutional perspective and the task-specific perspective, as well as the environmental perspective outlined in Chapter 3, may be of relevance in explaining the autonomy, control and internal management of agencies. Chapter 13 thus deals with the final research questions: RQ 6: How do agency-level characteristics interact with state-level characteristics in explaining agency autonomy, control and internal management? RQ 7: Which of the four theoretical perspectives has the strongest explanatory power with respect to the level and pattern of autonomy, control and internal management in the three states? Or do all four theoretical perspectives add explanations to the empirical data?

Since the 1980s, the structuring and functioning of the Organisation for Economic Co-operation and Development (OECD) public sector has undergone major shifts (cf. OECD 2002a; Pollitt and Bouckaert 2004). One of the most observed trends in public sector organization in OECD states is the shift from a centralized and consolidated public sector to a decentralized, structurally devolved and ‘autonomizing’ public sector, including the disconnection of policy design, implementation and evaluation (Christensen and Laegreid 2001a, 2006). Systems of public administration have been disaggregated into a multitude of different kinds of (semi-)autonomous organizations, denoted as ‘agencies’ or ‘quangos’ (cf. Flinders and Smith 1999; Pollitt and Talbot 2004). This disaggregation through ‘agencification’ is the result of a process of vertical and horizontal specialization, based on geography as well as on different types of purposes, tasks, customer groups or processes (Christensen et al. 2007; Roness 2007). In this process of agencification and autonomization, the responsibilities and autonomy of public organizations are redefined (structural aspect). Moreover, the way that they are controlled by government, including the mechanisms of accountability, are redesigned, mostly from ex ante to ex post, and from input to results based rationals (functional aspect).

In Chapter 3, we briefly formulated our model for comparing agencies within and across states. We described a model of conflicting pressures for similarity and dissimilarity at international, state and agency level. Figure 3.1 presents this model in a schematic way. In this chapter we ask what kind of statelevel factors/pressures for divergence may influence the actual autonomy and control of agencies in the three states. As outlined in Chapter 5, these factors may be clustered in five main sets with different levels of changeability: environmental pressures, dimensions of politico-administrative culture, structural and functional elements of polity, actor constellations and deliberate action (for example, general reform programmes).

In their comparative work on public management reforms, Pollitt and Bouckaert (2004: 40) refer to politico-administrative systems as those fundamentals of political and administrative systems that change only gradually or infrequently, and which may therefore be regarded as rather stable characteristics of the environment in a given polity. A politico-administrative system, in their view, includes structural, cultural and functional elements. Drawing from this and other sources (Christensen and Laegreid 2001b; Wollmann 2003b), we identify and distinguish here between different levels of politico-administrative analysis, which will serve to illuminate later chapters both theoretically and empirically. The emphasis in this chapter is on characterizing the politico-administrative regimes in Norway, Ireland and Flanders across a range of relevant dimensions. The implications of these regimes for agency autonomy, control and internal management are then discussed in Chapter 7.

In Chapter 8 we discussed the extent to which agencies use management techniques, and whether differences or similarities exist in their use between agencies in different states. We emphasized comparisons across states, and whether the similarities and dissimilarities were upheld for agencies with certain features. In this chapter we analyse the extent to which agency characteristics, and the autonomy and control of agencies, make a difference for their use of various management techniques. Thus, we aim to answer the fifth research question: RQ 5: What agency-level factors influence the internal management of agencies? To what extent does the level of autonomy and control affect the internal management? As shown in Chapter 9, there is a wide variation in the degree to which agencies use management techniques — our dependent variables. Taking all agencies in the three states together, we see that half of them use techniques such as allocation of resources internally on the basis of performance and cost-calculation systems. Techniques related to quality management systems are used by approximately 35 per cent of the agencies. Techniques like multiyear planning (85 per cent), public reporting (93 per cent) and customer surveys (68 per cent) are applied by a (large) majority of the agencies in Norway, Ireland and Flanders.

The previous chapter shows that the dominant pattern for managerial autonomy, result control and use of management techniques is that Norwegian agencies rank higher than the other two states, and that they are closer to the NPM ideal-type model of agencies. Flemish agencies report using more management techniques than Irish agencies, but with regard to managerial autonomy and result control, agencies in these two states are quite similar, demonstrating low to moderate levels. In this chapter we discuss the third research question: RQ 3: How can we explain similar/dissimilar levels and patterns of autonomy, control and use of management techniques between the three states by referring to aspects of their politico-administrative regimes (state-level characteristics)? In Chapter 7 we formulated hypotheses on the importance of environmental pressures, politico-administrative culture, structural and functional elements of polity, and actor constellations and deliberate actions on agency autonomy, control and internal management (see Table 7.1). In the change management approach, hypotheses focused on the influence of environmental or state-level characteristics on the extent to which a state would adopt more radical NPM-type agency reforms. Such reforms would result in agencies which report relatively higher levels of managerial autonomy, result control and use of management techniques.

This chapter maps the development of agencies in the three states, with a particular focus on ‘waves’ or agencification and related administrative reforms. We respond to the first empirical research question, setting the background of this study: RQ 1: To what extent are agencies created by the governments of the three states, in what form, with what legal-structural status, and in which policy domains? What tasks are they performing? How is the creation o f agencies evolving over time? While clearly delineating what does or does not constitute an agency differs according to politico-administrative regimes, the definitions discussed in Chapter 2 provide a compass for our inquiries. Each state is considered in turn, beginning with Norway. The last part of the chapter compares the actual agency landscapes in the three states in terms of task, size, budget, governance structures and other characteristics.

The influence of management on the performance of public sector organizations has regained academic interest with the rise of the ‘management matters’ thesis, itself inspired to a large extent by the doctrines of ‘managerialism’ and the ‘New Public Management’. These doctrines advocate the adoption of private managerial tools (e.g. Hood 1991; Osborne and Gaebler 1992) within public sector organizations in order to increase their efficiency, effectiveness, and quality of service delivery. The amount of empirical evidence on the influence of managerial styles and strategies on organizational performance has been growing rapidly, but the contextual factors remain very unclear. Besides, not much is known about the influence of specific managerial tools on performance of public sector organizations. Finally, studies in this area mainly focus on governments as a whole rather than on individual public sector organizations.

In Chapter 8 we presented the extent of perceived agency autonomy and control for all agencies and for agencies within each of the three states. We emphasized comparisons across states, and examined whether the similarities and dissimilarities were upheld for agencies with certain features. In this chapter we check whether, and to what extent, agency characteristics make a difference for their autonomy and control. Thus, we aim to answer the fourth research question: RQ 4: What agency-level factors influence the degree and pattern of autonomy and control of agencies? The analysis is based on the discussion of the importance of structural, cultural and task features for agency autonomy and control in Chapter 10. As noted, in most instances the hypotheses being formulated imply variations in opposite directions for autonomy and control. The hypotheses are summarized in Table 11.1, together with the main arguments why these agency characteristics will make an impact on agency autonomy (and control).