ONLY GOOD FENCES KEEP GOOD NEIGHBOURS! THE INSTITUTIONALIZATION OF MINISTRY–AGENCY RELATIONSHIPS AT THE SCIENCE–POLICY NEXUS IN GERMAN FOOD SAFETY POLICY
| Date | 01 March 2015 |
| Published date | 01 March 2015 |
| Author | SYLVIA VEIT,REBECCA‐LEA KORINEK |
| DOI | http://doi.org/10.1111/padm.12117 |
doi: 10.1111/padm.12117
ONLY GOOD FENCES KEEP GOOD NEIGHBOURS!
THE INSTITUTIONALIZATION OF MINISTRY–AGENCY
RELATIONSHIPS AT THE SCIENCE–POLICY NEXUS
IN GERMAN FOOD SAFETY POLICY
REBECCA-LEA KORINEK AND SYLVIA VEIT
This article makes a further contribution to opening the ‘black box’ of micropractices in
ministry–agency relationships. We argue that the mechanisms that come into play in the course
of institutionalizing agencication reforms – such as renegotiating mutual roles and rules in
ministry–agency interactions – are only poorly covered in the existing literature. To adequately
address the negotiated and contingent nature of de facto agency autonomy and political control,
we develop an interpretive approach based on the concept of ‘boundary work’. The empirical
focus is on ministry–agency interactions at the science–policy nexus in the contested policy eld
of food safety. By studying actors’ stories about the institutionalization processes following the
fundamental reorganization of the German food safety administration in the wake of the BSE crisis
from a longitudinal perspective, we show how actors manage boundary conicts via increasingly
differentiated backstage coordination.
INTRODUCTION
For almost two decades, a growing body of literature within public administration
has focused on agencication reforms – that is, the delegation of regulatory and other
state competencies to various forms of formally semi-autonomous bodies (e.g. Pollitt
et al. 2001; Gilardi 2005; Christensen and Lægreid 2006; Van Thiel et al. 2012). While this
research has provided important insights, for example with regard to the lack of a clear
relationship between formal and de facto autonomy of agencies, it is still struggling to
capture the mechanisms that come into play in the course of the institutionalization of
the agencication reforms in practice (Yesilkagit 2004). To adequately address microlevel
institutionalization dynamics, such as the renegotiation of mutual roles and rules in
ministry–agency relationships, we concur with other recent contributions to the eld
in arguing that much is to be gained from adopting an interpretive approach to the
agencication phenomenon (Moynihan 2006; Flinders 2009; Smullen 2010; Hay 2011;
Elston 2014).
This article sets out to address this microlevel practice gap by interpreting the institu-
tionalization of agencication within a specic setting, namely the science–policy nexus in
German food safety policy, which was fundamentally reorganized in the wake of the BSE
crisis in the early 2000s. To explore the question how the intra- and inter-organizational
role understandings and practices in ministry–agency relationships changed after reorga-
nization, we develop a heuristic allowing us to conceptualize de facto agency autonomy
as the result of ‘boundary work’ (Gieryn 1983). Methodically, we draw on narrative data
from semi-structured, personal expert interviews conducted with various members of the
organizations involved.
Rebecca-Lea Korinek is at the Cultural Sources of Newness Unit, at the Berlin Social Science Center Berlin (WZB), Ger-
many.Sylvia Veit is in the Faculty of Economics and Management, University of Kassel, Kassel, Germany.
Public Administration Vol.93, No. 1, 2015 (103–120)
© 2014 John Wiley & Sons Ltd.
104 REBECCA-LEA KORINEK AND SYLVIA VEIT
WHY WE NEED AN INTERPRETIVE APPROACH TO AGENCY AUTONOMY
In the 1990s scholars began to observe a growing number of semi-autonomous agen-
cies spreading across countries and sectors, resulting in the metaphorical diagnosis of
‘agency-fever’ (Pollitt et al. 2001). In addition to mapping, assessing, and explaining the
rise and renewal of agency landscapes, agencication research has focused on concep-
tualizing and empirically assessing ‘agency autonomy’. A major analytical distinction
is made between formal autonomy (‘the level of decision-making competencies’) and
de facto autonomy (‘the exemption of constraints on the actual use of decision-making
competencies’) (Verhoest et al. 2004, p. 104). According to this conception, autonomy is
inherently a quantitative phenomenon. This has considerable implications for how agency
autonomy is approached empirically, namely in terms of the varying degrees of formal or
de facto autonomy.
Most of the scholars focusing on de facto autonomy attempt to measure the autonomy
of executive agencies on the basis of large-Nsurvey research (Verhoest et al. 2012). One
remarkable nding states that in a number of countries there is still a considerable degree
of political control over the operational work of agencies. A higher degree of formal auton-
omy hence does not always imply a higher degree of de facto autonomy (e.g. Yesilkagitand
Van Thiel 2008; Maggetti 2012). At the same time, however, the patterns of autonomy and
control in the actual ministry–agency interactions are contingent on different contextual
factors (Verhoest et al. 2012). These survey results lend some support to critiques ques-
tioning one of the dominant New Public Management myths: namely, the possibility and
universal rationality of a clear separation between politics/policy (having the primacy)
and bureaucracy.
Studying Dutch agencies, Yesilkagitand Van Thiel (2008) conclude that the relationships
between autonomy and control are highly complex and that the question remains how
mechanisms of control work in practice, given the high degree of inuence that the agen-
cies ascribed to governmental actors, according to their survey. Discussing the method
used, they nally argue that the question of the practical operation of autonomy and con-
trol, that is, the way in which agencies respond to external expectations, requires ‘a more
dynamic approach’ than survey research permits (Yesilkagit and Van Thiel 2008, p. 152).
Roness et al. (2008, p. 168) argue that such ndings, which draw a rather messy picture of
how the patterns of autonomy and control function in practice, are entirely plausible con-
sidering the deeply negotiated and contingent nature of agency autonomy.Finally, Verhoest
et al. (2010), though supporting basic claims made by ‘task-specic path-dependence the-
ory’ (Pollitt et al. 2004), suggest that there should be a stronger focus in future research on
‘the mechanisms and logics in the minds of the key actors involved’ (Verhoest et al. 2010,
p. 269).
This brief literature overview shows that while previous research has produced impor-
tant insights, several questions remain unanswered: How do actors use and understand
the concept of agency autonomy and political control? What mechanisms come into play
in the course of the institutionalization of the agency model, for example in terms of ‘role
purication’ (Christensen and Lægreid 2006), following the establishment of new agencies
(Yesilkagit 2004)? If the agency model is contested, how does this contestation take place,
and how might the model reassert itself in the face of contestations? How do the different
institutional logics of politics, bureaucracy, and other social spheres shape organizational
sensemaking processes with regard to the agency model?
Public Administration Vol.93, No. 1, 2015 (103–120)
© 2014 John Wiley& Sons Ltd.
To continue reading
Request your trial