Determinants of result-based control in Italian agencies

Explaining the form of public sector organizations is a major contribution that organizational theory has to offer to the study of the public sector. In this chapter, we investigate the characteristics of the relationships between agencies and their parent administration in the Italian central government. The study, rooted in organizational theory, adopts different theoretical angles, at first a ‘cultural-normative’ approach mainly embedded in a logic of appropriateness perspective (March and Olsen, 1989, 1994, 1996), then an approach focused on instrumental rationality mainly embedded in a logic of consequences perspective (for an overview of different streams in organization theory for the study of the public sector, see Christensen et al., 2007). Governments in many countries have established semi-autonomous single-purpose organizations (OECD, 2002), to which we refer as public agencies, and Italy has been no exception in this respect (Ongaro, 2006 and 2008). How to wield the control function towards agencies has become a prominent area of research in public management studies. Many studies address the topic of control as a ‘systemic’ issue, as the notion of control has been related to the way the relationships among various actors are coordinated in order to achieve a common (public) goal (Kaufmann, 1986). Other studies deal with the concept of control as an inter-organizational coordination tool (Klijn, 2005; Narad, 2006; Verhoest et al., 2007). Other authors have explored the control topic as an accountability-related theme (Bovens et al., 2008; Gregory, 2003; Mulgan, 2000) or even more generally as a central issue in public administration and public management reform (Kettl, 2000; Romzek, 2000). As a starting point, we may assume that there are two broad modalities whereby control of agencies is exercised: ex ante controls and ex post controls. Ex ante controls (Thompson, 1993; Verhoest et al., 2004; Wirth, 1986) refer to a broad set of mechanisms ranging from the appointment of the director general/CEO and the members of the board to the definition of rules and standards by the parent organization to which the agency has to comply (constraining the behaviour in order to steer the agency towards the outputs desired by the parent administration).This perspective refers both to the formal-bureaucratic aspects of the ministry-agency relationship (Egeberg, 2003) and to other forms of control that can be read (also) through the lens of the theory of clans (Ouchi, 1980) – whereby control is pursued through the selection and appointments of top executives having clannish relations (in the sense clarified by Ouchi) with the ‘political master’ in the parent administration. The second perspective is based on the assumption that the ‘parent’ administration can steer and control ‘its’ agency by setting the goals it is expected to pursue and controlling ex post their achievement. This second perspective largely refers to the management control and performance management literature (Bouckaert and Halligan, 2008; Bouckaert and van Dooren, 2003; Laegried et al., 2006a; Moynihan, 2006) and has important roots in the concept of ‘Management by Objectives and Results’: it is a performance management concept that ‘entails a redoubled attempt to operationalize goals for government and to use these goals more actively both in choosing between alternative actions and in evaluating results’ (Christensen et al., 2007, p. 89). Ex post control, or result-based control, concerns the assignment of goals to the agency and the assessment by the parent administration of the degree of achievement of the goals by the agency, as well as the possible execution of sanctions or attribution of rewards associated to the achievement of goals. The complex nature of the relationships between the parent administration and the agency based on a performance management logic has recently been studied with reference to the autonomy-control balance (Verschuere, 2007; see also Barbieri and Ongaro, 2008) or as an indicator of ‘NPMness’ of such relationships (Verhoest et al., 2004; Verschuere and Barbieri, 2009). This chapter addresses the question of what are the determinants of result-based forms of control. The analysis is focused on the ex post result control for three main reasons: the already extensive literature on ex ante control instruments on Italian agencies, which contrasts with the absence (to our knowledge) of any systematic investigation of the determinants of result-based controls; the chance to compare the findings on the impact of the selected independent variables (age, staff size, budget size, type of task and policy field) with investigations performed in other countries (see this book, and in the range of recent works on the topic the contributions by Laegreid et al., 2006b; Verschuere, 2007); the opportunity to empirically investigate the degree of implementation of a major public management reform that occurred in Italy in 1999, that had the declared goal of introducing forms of result-based control in the Italian public sector. In the study of the determinants of result-based control, we have at first considered that the forms of control between the agency and the parent administration may be determined by institutionalisation processes that unfold over time under the influence of certain factors. We assume both ministries and agencies to be institutionalized organisations and we investigate their relationship on the basis of factors that may affect the organisational form of their relationship. Factors like age, size (in both budget and staff terms), the type of task executed, the policy field in which the agency and the ministry operate have been considered (on the ground of theoretical argumentations explained in the third section, in which the hypotheses about the influence of such factors on result-based control are formulated).A supplement of investigation was required, however, since the statistical analysis did not provide conclusive evidence about the influence such factors may have on the form of control. We then adopted a different perspective, and considered the reform processes that occurred in Italy over the last two decades (Ongaro, 2009). In the range of a number of reform programmes that were launched during that period in Italy, one can be singled out as regards the area of organisational form of central government agencies and the agency-ministry relationship; in fact, in 1999 a comprehensive reform of the central government prescribed, within the framework of a general re-organisation of the central government, the establishment of a number of new agencies patterned on a well-identified model, which can be easily recognized in the UK ‘Next Steps’ model of public agency. Though it is beyond the scope of this chapter to investigate the policy process that led to the enactment of this reform programme (to what extent it was a case of policy transfer from abroad, the way different policy alternatives were considered and how this one came to be selected, etc.), for the purposes of this work we may consider that a well-specified model of public agency, centred on result-based control, was intentionally introduced in the Italian central administration. From a hierarchically-oriented instrumental perspective on reform (Christensen et al., 2007, chapters 2 and 7), we may quite straightforwardly state that the executive government at the higher level forced the introduction of a new organisational form in the Italian central public administration. What have been the effects of this reform on ex post controls in the relationship between agencies and their parent administration? And has this new organisational form, and particularly the type of result-based control which is a central component of it, institutionalised over time? As illustrated in closer detail later in this chapter, this reform has led to establishing a set of agencies that can be singled out of the overall population of the investigated central government agencies because of their high level of result-control. However, the reform has been apparently not ‘institutionalized’ and the broader population of Italian central government agencies (encompassing those agencies established subsequently to the 1999 reform) displays a level of result-based control markedly different from that of the agencies established by the 1999 reform.