Passing the buck? Analyzing the delegation of discretion after transposition of European Union law

Abstract

This article seeks to map and explain the extent to which national legislators constrain discretion contained in European Union directives during transposition. To this end, we use standard hypotheses from the domestic delegation literature regarding the necessity of policy conflict and transaction costs. Our empirical approach is based on a focused comparison of the transposition of several provisions of the Asylum Reception Conditions Directive in France, Germany, and the Netherlands. In order to capture content-specific aspects of discretion we employ an innovative measurement tool, the so-called Institutional Grammar Tool. The study shows that while all three states formally comply with the directive, the level of European Union discretion delegated to practical implementers varies considerably across the cases. Standard delegation theory cannot fully explain the patterns. Instead, existing delegation theories have to be adjusted to the transposition context, by accounting for domestic preferences regarding the status quo. © 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.

Full text

Passing the buck? Analyzing the delegation of discretion
after transposition of European Union law
Nora Dörrenbächer *, Ellen Mastenbroek
Institute for Management Research, Radboud University, Nijmegen, the Netherlands
Abstract
This article seeks to map and explain the extent to which national legislators constrain discretion contained in European Union
directives during transposition. To this end, we use standard hypotheses from the domestic delegation literature regarding the
necessity of policy conflict and transaction costs. Our empirical approach is based on a focused comparison of the transposition
of several provisions of the Asylum Reception Conditions Directive in France, Germany, and the Netherlands. In order to capture
content‐specific aspects of discretion we employ an innovative measurement tool, the so‐called Institutional Grammar Tool. The
study shows that while all three states formally comply with the directive, the level of European Union discretion delegated to
practical implementers varies considerably across the cases. Standard delegation theory cannot fully explain the patterns. Instead,
existing delegation theories have to be adjusted to the transposition context, by accounting for domestic preferences regarding the
status quo.
Keywords: asylum law, delegation of discretion, EU transposition, reception conditions directive, transaction costs.
1. Introduction
A unique characteristic of European Union (EU) directives is that they are only binding as to the results to be achieved.
Yet directives leave member states variant discretion. Several studies have focused on the conditions under which EU
legislation delegates discretion to member states (Pollack 1997; Franchino 2007; Thomson & Torenvlied 2011). The
exact level of discretion is considered a deliberate choice of the European legislator aimed at finding compromises at
the EU level and at allowing member states to adjust their laws to local circumstances (Franchino 2007; Hartmann
2016).
However, discretion is not only relevant for the stage of EU lawmaking. It also plays an important role in the phase
of transposition, or the conversion of directives into national law (Zhelyazkova & Torenvlied 2009; Hartmann 2016). In
principle, discretion in EU directives provides transposition actors with control over practical implementation and
compliance. The implicit assumption in the literature on EU compliance is that transposition actors settle for a partic-
ular point within, or outside, the margins of EU discretion (Dimitrova & Steunenberg 2013). So far, the possibility that
transposition actors further delegate EU discretion to practical implementers has received no explicit scholarly atten-
tion (for research touching upon this possibility, see Steunenberg 2006). However, this possibility is highly relevant
for compliance. Transposition actors that delegate discretion, “pass the buck”by shifting responsibility for actual
compliance to practical implementers who thus become, in essence, EU lawmakers.
Studies on delegation in a national context have extensively investigated how political context affects the political
decision to limit agency discretion (McCubbins & Schwartz 1984; Calvert et al. 1989; Epstein & O’Halloran 1999;
Huber & Shipan 2002). Two key assumptions derive from this literature. First, a high level of risk that the agent will
not comply with decision‐makers’preferences is necessary for legislators to want to constrain discretion (Epstein &
O’Halloran 1999). The risk is high when there is a policy conflict between decisionmaker and implementer (Epstein
&O’Halloran 1999; Huber & Shipan 2002). Second, legislators have to be capable of constraining discretion. Crucially,
Correspondence: Nora Dörrenbächer, Radboud University, Institute for Management Research, P.O. Box 9108, 6500 HK Nijmegen,
the Netherlands. Email: n.dorrenbacher@fm.ru.nl
Accepted for publication 07 February 2017.
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
Regulation & Governance (2017) ••,••–•• doi:10.1111/rego.12153
This is an openaccess article under the terms of the CreativeCommons Attribution‐NonCommercial‐NoDerivs License, which permits use and distribution in any medium, pro-
vided the original work is properly cited, the use is non‐commercial and no modifications or adaptations are made.
discretion will only be constrained if the transaction costs of writing detailed legislation are low (Huber et al. 2001;
Oosterwaal et al. 2012, p. 803).
This study asks the question to what extent transposition actors indeed “pass the buck,”that is, delegate discretion
contained in EU directives down to practical implementers and to what extent domestic delegation theories account for
constraining transposition measures. By answering this question, this study seeks to add to the literature in three ways.
First, we apply insights from the domestic delegation literature to the EU implementation context. Thereby, the study
tackles Huber and Shipan’s claim that analyzing the delegation of discretion in parliamentary systems has its limits,
because of the absence of a “natural experiment that requires a wide range of political systems to address the same issue
at roughly the same time”(2002, p. 65). Following Huber and Shipan (2002 p 181, fn 8), the transposition of EU
directives might present such a natural experiment.
1
Moreover, by testing to what extent established domestic delega-
tion theories apply to the transposition context, the paper accounts for the fact that across EU member states, a
considerable amount of legislation has its origin in EU obligations.
Second, by exploring transposition outcomes beyond compliance, this study goes beyond existing studies on
transposition (Treib 2014). With the focus on EU discretion left to practical implementers, this study contributes to
the recent discussion on customization of EU law (Thomann 2015) and enhances our understanding of the regulatory
patchwork of the EU (Héritier 1996). In this way, the paper highlights the fact that EU implementation does not end
after transposition and draws attention to the role of practical implementers (Versluis 2007).
Finally, we complement previous statistical findings on the delegation of discretion with a nuanced and in‐depth
account of the multidimensional concept of discretion (Epstein & O’Halloran 1999; Huber & Shipan 2002; Oosterwaal
et al. 2012). We do so by employing a new systematic measure of discretion that draws inspiration from the
Institutional Grammar developed by Crawford and Ostrom (1995). Toshkov (2013) suggested this approach as a tool
to add content‐specific elements to existing quantitative measures of discretion.
We empirically investigate the level of discretion for the transposition of several provisions of a directive in the
highly topical field of asylum: Council Directive 2003/9/EC, the so‐called Reception Condition Directive (RCD). So
far, the implementation of EU asylum law has only received marginal scholarly attention. However, in light of the high
recent refugee influx, understanding implementation outcomes in this sensitive policy field is more crucial than ever.
The transposition processes and delegation outcomes are studied in a structured focused comparison (George &
Bennett 2005, pp. 67–72) of transposition in France, Germany, and the Netherlands. In order to investigate to what
extent policy conflict and transaction costs are necessary conditions for constraining transposition, we select countries
with highly diverse political contexts. Extensive document study and interviews with stakeholders in all three states
serve to investigate the theoretical expectations.
The analysis reveals that in all three cases, transposition acts delegated EU discretion to practical implementers.
However, the level of discretion varied considerably between member states and provisions. Strikingly, the analysis
shows that domestic delegation theories face some limitations in explaining this pattern because policy conflict and
low transaction costs do not seem to be necessary conditions for constrained discretion. To account for the findings,
the transposition context needs to be taken into account. During transposition, the implicit assumption of domestic
delegation theory that the agenda‐setter advocates change, is often not fulfilled. EU law may force transposition actors
to reopen national lawmaking on issues they would not otherwise have initiated themselves. Thus, this study
inductively suggests that preferences to maintain the status quo, which have received more explicit attention in the
Europeanization and EU implementation literature, matter for the delegation of discretion (see Héritier 1996; Duina
1997; Knill & Lenschow 1998).
2
Preferences to maintain the status quo can trigger transposition actors to maintain
constraining national legislation, despite the absence of necessary conditions assumed in the standard delegation
literature.
2. Theorizing the delegation of European Union discretion
The transposition of EU directives tends to be regarded as a political process (Mastenbroek & Kaeding 2006;
Steunenberg 2006; Toshkov 2011; Dimitrova & Steunenberg 2013). Depending on the bargaining environment, differ-
ent interests at the practical implementation and decisionmaking stages will confront the ministry that coordinates
transposition and drafts the transposition laws (Dimitrova & Steunenberg 2013, p. 250). Following these assumptions,
transposition is influenced by a range of domestic players, similar to domestic lawmaking (Steunenberg 2006).
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
2© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
Regulation & Governance (2019) 13, 70–85 doi:10.1111/rego.12153

Passing the buck? Analyzing the delegation of discretion
after transposition of European Union law
Nora Dörrenbächer *, Ellen Mastenbroek
Institute for Management Research, Radboud University, Nijmegen, the Netherlands
Abstract
This article seeks to map and explain the extent to which national legislators constrain discretion contained in European Union
directives during transposition. To this end, we use standard hypotheses from the domestic delegation literature regarding the
necessity of policy conflict and transaction costs. Our empirical approach is based on a focused comparison of the transposition
of several provisions of the Asylum Reception Conditions Directive in France, Germany, and the Netherlands. In order to capture
content‐specific aspects of discretion we employ an innovative measurement tool, the so‐called Institutional Grammar Tool. The
study shows that while all three states formally comply with the directive, the level of European Union discretion delegated to
practical implementers varies considerably across the cases. Standard delegation theory cannot fully explain the patterns. Instead,
existing delegation theories have to be adjusted to the transposition context, by accounting for domestic preferences regarding the
status quo.
Keywords: asylum law, delegation of discretion, EU transposition, reception conditions directive, transaction costs.
1. Introduction
A unique characteristic of European Union (EU) directives is that they are only binding as to the results to be achieved.
Yet directives leave member states variant discretion. Several studies have focused on the conditions under which EU
legislation delegates discretion to member states (Pollack 1997; Franchino 2007; Thomson & Torenvlied 2011). The
exact level of discretion is considered a deliberate choice of the European legislator aimed at finding compromises at
the EU level and at allowing member states to adjust their laws to local circumstances (Franchino 2007; Hartmann
2016).
However, discretion is not only relevant for the stage of EU lawmaking. It also plays an important role in the phase
of transposition, or the conversion of directives into national law (Zhelyazkova & Torenvlied 2009; Hartmann 2016). In
principle, discretion in EU directives provides transposition actors with control over practical implementation and
compliance. The implicit assumption in the literature on EU compliance is that transposition actors settle for a partic-
ular point within, or outside, the margins of EU discretion (Dimitrova & Steunenberg 2013). So far, the possibility that
transposition actors further delegate EU discretion to practical implementers has received no explicit scholarly atten-
tion (for research touching upon this possibility, see Steunenberg 2006). However, this possibility is highly relevant
for compliance. Transposition actors that delegate discretion, “pass the buck”by shifting responsibility for actual
compliance to practical implementers who thus become, in essence, EU lawmakers.
Studies on delegation in a national context have extensively investigated how political context affects the political
decision to limit agency discretion (McCubbins & Schwartz 1984; Calvert et al. 1989; Epstein & O’Halloran 1999;
Huber & Shipan 2002). Two key assumptions derive from this literature. First, a high level of risk that the agent will
not comply with decision‐makers’preferences is necessary for legislators to want to constrain discretion (Epstein &
O’Halloran 1999). The risk is high when there is a policy conflict between decisionmaker and implementer (Epstein
&O’Halloran 1999; Huber & Shipan 2002). Second, legislators have to be capable of constraining discretion. Crucially,
Correspondence: Nora Dörrenbächer, Radboud University,Institute for Management Research, P.O. Box 9108, 6500 HKNijmegen,
the Netherlands. Email: n.dorrenbacher@fm.ru.nl
Accepted for publication 07 February 2017.
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
Regulation & Governance (2017) ••,••–•• doi:10.1111/rego.12153
This is an openaccess article under the terms of the Creative Commons Attribution‐NonCommercial‐NoDerivs License, which permits use anddistribution in any medium, pro-
vided the original work is properly cited, the use is non‐commercial and no modifications or adaptations are made.
discretion will only be constrained if the transaction costs of writing detailed legislation are low (Huber et al. 2001;
Oosterwaal et al. 2012, p. 803).
This study asks the question to what extent transposition actors indeed “pass the buck,”that is, delegate discretion
contained in EU directives down to practical implementers and to what extent domestic delegation theories account for
constraining transposition measures. By answering this question, this study seeks to add to the literature in three ways.
First, we apply insights from the domestic delegation literature to the EU implementation context. Thereby, the study
tackles Huber and Shipan’s claim that analyzing the delegation of discretion in parliamentary systems has its limits,
because of the absence of a “natural experiment that requires a wide range of political systems to address the same issue
at roughly the same time”(2002, p. 65). Following Huber and Shipan (2002 p 181, fn 8), the transposition of EU
directives might present such a natural experiment.
1
Moreover, by testing to what extent established domestic delega-
tion theories apply to the transposition context, the paper accounts for the fact that across EU member states, a
considerable amount of legislation has its origin in EU obligations.
Second, by exploring transposition outcomes beyond compliance, this study goes beyond existing studies on
transposition (Treib 2014). With the focus on EU discretion left to practical implementers, this study contributes to
the recent discussion on customization of EU law (Thomann 2015) and enhances our understanding of the regulatory
patchwork of the EU (Héritier 1996). In this way, the paper highlights the fact that EU implementation does not end
after transposition and draws attention to the role of practical implementers (Versluis 2007).
Finally, we complement previous statistical findings on the delegation of discretion with a nuanced and in‐depth
account of the multidimensional concept of discretion (Epstein & O’Halloran 1999; Huber & Shipan 2002; Oosterwaal
et al. 2012). We do so by employing a new systematic measure of discretion that draws inspiration from the
Institutional Grammar developed by Crawford and Ostrom (1995). Toshkov (2013) suggested this approach as a tool
to add content‐specific elements to existing quantitative measures of discretion.
We empirically investigate the level of discretion for the transposition of several provisions of a directive in the
highly topical field of asylum: Council Directive 2003/9/EC, the so‐called Reception Condition Directive (RCD). So
far, the implementation of EU asylum law has only received marginal scholarly attention. However, in light of the high
recent refugee influx, understanding implementation outcomes in this sensitive policy field is more crucial than ever.
The transposition processes and delegation outcomes are studied in a structured focused comparison (George &
Bennett 2005, pp. 67–72) of transposition in France, Germany, and the Netherlands. In order to investigate to what
extent policy conflict and transaction costs are necessary conditions for constraining transposition, we select countries
with highly diverse political contexts. Extensive document study and interviews with stakeholders in all three states
serve to investigate the theoretical expectations.
The analysis reveals that in all three cases, transposition acts delegated EU discretion to practical implementers.
However, the level of discretion varied considerably between member states and provisions. Strikingly, the analysis
shows that domestic delegation theories face some limitations in explaining this pattern because policy conflict and
low transaction costs do not seem to be necessary conditions for constrained discretion. To account for the findings,
the transposition context needs to be taken into account. During transposition, the implicit assumption of domestic
delegation theory that the agenda‐setter advocates change, is often not fulfilled. EU law may force transposition actors
to reopen national lawmaking on issues they would not otherwise have initiated themselves. Thus, this study
inductively suggests that preferences to maintain the status quo, which have received more explicit attention in the
Europeanization and EU implementation literature, matter for the delegation of discretion (see Héritier 1996; Duina
1997; Knill & Lenschow 1998).
2
Preferences to maintain the status quo can trigger transposition actors to maintain
constraining national legislation, despite the absence of necessary conditions assumed in the standard delegation
literature.
2. Theorizing the delegation of European Union discretion
The transposition of EU directives tends to be regarded as a political process (Mastenbroek & Kaeding 2006;
Steunenberg 2006; Toshkov 2011; Dimitrova & Steunenberg 2013). Depending on the bargaining environment, differ-
ent interests at the practical implementation and decisionmaking stages will confront the ministry that coordinates
transposition and drafts the transposition laws (Dimitrova & Steunenberg 2013, p. 250). Following these assumptions,
transposition is influenced by a range of domestic players, similar to domestic lawmaking (Steunenberg 2006).
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
2© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 71

Domestic lawmaking has been theorized extensively by the delegation literature, which has developed policy‐specific
explanations for discretion in national statutes (Epstein & O’Halloran 1994; Huber & Shipan 2002).
The starting point of the delegation literature is that delegation serves to transfer power from the legislative princi-
pal to an implementing agent (Pollack 1997). Delegation of discretion is considered a strategic bargaining game in
which legislators are understood as rational actors (Moe 1982; McCubbins et al. 1987, p. 256). Political factors are
the main explanations for differences in the level of discretion in national legal statutes (Epstein & O’Halloran 1994;
Huber & Shipan 2002; Bendor & Meirowitz 2004).
We take Huber and Shipan’s (2002, 2011) approach toward delegation as a point of departure because it combines
recent theories on domestic delegation of discretion in the most parsimonious way. The authors distinguish four
hypotheses on the political dynamics of delegation, emphasizing the influence of policy conflict between principal
and agent, transaction costs, policy uncertainty, and non‐statutory factors. We focus on the first two factors, controlling
for policy uncertainty and non‐statutory factors by keeping the policy context constant.
The central theoretical argument in the delegation literature relates to policy conflict between principal and agent.
The assumption here is that both legislators and implementers seek to realize implementation close to their preferences.
The chances that an agent implements the law not in the interest of the principal are higher under discretionary
legislation than under constraining legislation (Bendor & Meirowitz 2004). As a consequence, delegation theory
assumes that if the policy preferences of legislators and practical implementers diverge, the former have strong
incentives to constrain the latters’room for maneuver (Epstein & O’Halloran 1999; Huber & Shipan 2002).
Respectively, without policy conflict the legislator will have no incentive to constrain discretion (Huber et al. 2001).
According to this view, policy conflict between principal and agents is a necessary condition for the adoption of
constraining legislation (Huber et al. 2001, p. 343). Applied to the EU transposition context, this leads to the following
expectation:
Expectation 1: Transposition actors will only constrain discretion granted by EU directives in the case of policy
conflict between transposition actors and practical implementers concerning the directive.
Yet following Huber et al. (2001, p. 343) the risk of divergence is not a sufficient condition for adopting constraining
legislation. Crucially, transposition actors must also be able to cover the costs of writing and adopting constraining
legislation (Moe 1984; Huber & Shipan 2002, p. 79). These transaction costs involve resources, such as time and exper-
tise required when legislators need to convince critical veto players of their legal proposals. These costs are resources
that legislators could otherwise channel into other tasks. The bargaining environment determines the height of these
transaction costs (Huber et al. 2001). As argued by Oosterwaal et al. (2012, p. 803), transaction costs increase when
many veto players holding divergent preferences are involved in lawmaking. In such situations, adopting constraining
legislation is highly costly (Epstein & O’Halloran 1994, 1999; Huber & Shipan 2002) and typically leads to the adoption
of compromises with low detail and few specific descriptions (Oosterwaal et al. 2012).
In the EU transposition context, the involvement and preferences of veto players may vary across national
systems. In line with the transposition literature, we define veto players as actors who are either formally or
informally involved in the transposition process and who are required to find a compromise (Dimitrova &
Steunenberg 2013 p. 250). Sometimes, transposition involves only the relevant ministry, for example, when
delegated acts are adopted (Dimitrova & Steunenberg 2013, p. 249). Under such circumstances, no compromise
needs to be negotiated and transaction costs are low. In other transposition processes, powerful policy‐specific
players, such as several ministries and parliament, are actively involved (Dimitrova & Steunenberg 2000; Haverland
2000). These actors are often linked through networks with powerful non‐state actors that become informal veto
players (Dimitrova 2010, p 145). The inclusion of informal veto players takes into account that some member
states and policy areas involve strong interest groups in political decisionmaking (Lijphart & Crepaz 1991). Formal
and informal veto players may have preferences that diverge from those of the transposing ministry, which raises
the costs of agreeing on constraining legislation and makes the transposing minister incapable of writing detailed
legislation (Steunenberg 2006). This leads to the second expectation:
Expectation 2: Transpostion actors will only constrain discretion granted by EU directives in the case of low trans-
action costs resulting from low conflict between the agenda‐setter and veto players involved in the transposition
process.
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 3
Huber and Shipan (2002) suggest two additional variables that influence the level of discretion, which we control
for in this study as they relate to the policy context. The first condition refers to policy uncertainty. This condition entails
that principals will delegate discretion to agents if the latter possess highly technical expertise necessary for
implementing complex policies (Huber & Shipan 2002, 2011). For the purpose of this study, we control for this variable
by selecting a relatively non‐technical policy area.
Additionally, Huber et al. (2001) argue that the legislator might have cheaper non‐statutory mechanisms for moni-
toring and controlling the actions of agents that make it unnecessary to invest in detailed legislation. For example,
legislators could veto rules adopted by agencies during hearings. Alternatively, decisionmakers could rely on external
actors, such as judges, to hold agents accountable (Epstein & O’Halloran 1994; Huber et al. 2001, pp. 334–335;
Blom‐Hansen 2005). In the context of EU transposition and asylum law, such mechanisms are largely absent. While
the EU Commission can be seen as an external actor to control member state under‐compliance, the Commission
tends to focus on legal transposition and less on the actions of national agencies on the ground. In case the Commission
does target practical implementation, the possibility of enforcement applies to all member states, which reduces its
suitability to explain cross‐country differences in discretion. Finally, litigation by asylumseekers might be an alternative
to correct under‐compliance. However, asylum seekers are unlikely to use litigation in such a way to represent the
interest of the national principal against the agency; instead they have their own interests.
Moreover, there are no non‐statutory mechanisms to monitor over‐compliance. Over‐compliance exists when the
agency gold‐plates the requirements of the directive against the preferences of the transposition actors. Such activities
are unlikely to be followed up by courts because the law’s target group, as possible litigants, would benefit from such
over‐compliance. Finally, in fields such as migration and asylum law, many policies are applied in street‐level
situations (e.g. Ellermann 2006). This makes non‐statutory monitoring difficult, as legislative veto or alternative
mechanisms do not exist to monitor street‐level bureaucratic behavior.
3. Methods and data
The expectations developed above will be investigated through a set of in‐depth case studies, using a structured focused
comparison (George & Bennett 2005, p. 69). As will be elaborated, the qualitative design allows the employment of a
more nuanced and context‐specific measure of the dependent variable of the relative share of discretionary and
constraining provisions than commonly used in quantitative measures of discretion. The downside of this approach
is that inferences based on a small sample might have lower external validity than large‐N correlation studies. However,
our qualitative approach allows us to look into the underlying causal mechanisms in detail. Additionally, because this
study focuses on necessary conditions, a low‐N study is suitable (Mahoney & Goertz 2006, p. 232). For such set‐
theoretical assumptions, a single contradicting case can disqualify the necessity of a certain condition (Goertz & Starr
2003). Moreover, the qualitative design makes it possible to inductively explore alternative mechanisms that specify
existing theories, in case the established necessary conditions do not fully account for the expected outcomes.
We investigate our expectations through an analysis of the transposition of several provisions of the Asylum
Reception Conditions Directive 2003/9/EC in three EU member states. By focusing on the provision level, we add
leverage to the low‐N study because dependent variable and explanatory conditions can vary on the provision level.
The RCD was chosen because, while constituting a prescriptive framework, many of its provisions provide ample
discretion to the member states, which is a scope condition for exploring our expectations (Huber & Shipan 2002,
p. 181). In addition, it forms part of the highly politicized field of asylum, which provides promising conditions for
the expected mechanisms to operate.
The directive’s key aim is to limit “asylum shopping”which was thought to stem from varying reception standards
between EU member states (Zaun 2016). The directive prescribes minimum standards for the reception of asylum
seekers that will normally suffice to ensure them a dignified standard of living. Additionally, it regulates various types
of material support for asylum seekers, including housing, food, clothing, and non‐material conditions such as
healthcare and employment. Member states may adopt or maintain provisions that are more favorable.
The Council of Ministers of the EU negotiated the directive from 1997 until 2003. During this time, the EU faced a
moderate influx of asylum seekers, compared with the years that were to follow. Nevertheless, member states were
divided on the issue. Council documents stressed repeatedly that the final act should leave “sufficient room for
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
4© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
72

Domestic lawmaking has been theorized extensively by the delegation literature, which has developed policy‐specific
explanations for discretion in national statutes (Epstein & O’Halloran 1994; Huber & Shipan 2002).
The starting point of the delegation literature is that delegation serves to transfer power from the legislative princi-
pal to an implementing agent (Pollack 1997). Delegation of discretion is considered a strategic bargaining game in
which legislators are understood as rational actors (Moe 1982; McCubbins et al. 1987, p. 256). Political factors are
the main explanations for differences in the level of discretion in national legal statutes (Epstein & O’Halloran 1994;
Huber & Shipan 2002; Bendor & Meirowitz 2004).
We take Huber and Shipan’s (2002, 2011) approach toward delegation as a point of departure because it combines
recent theories on domestic delegation of discretion in the most parsimonious way. The authors distinguish four
hypotheses on the political dynamics of delegation, emphasizing the influence of policy conflict between principal
and agent, transaction costs, policy uncertainty, and non‐statutory factors. We focus on the first two factors, controlling
for policy uncertainty and non‐statutory factors by keeping the policy context constant.
The central theoretical argument in the delegation literature relates to policy conflict between principal and agent.
The assumption here is that both legislators and implementers seek to realize implementation close to their preferences.
The chances that an agent implements the law not in the interest of the principal are higher under discretionary
legislation than under constraining legislation (Bendor & Meirowitz 2004). As a consequence, delegation theory
assumes that if the policy preferences of legislators and practical implementers diverge, the former have strong
incentives to constrain the latters’room for maneuver (Epstein & O’Halloran 1999; Huber & Shipan 2002).
Respectively, without policy conflict the legislator will have no incentive to constrain discretion (Huber et al. 2001).
According to this view, policy conflict between principal and agents is a necessary condition for the adoption of
constraining legislation (Huber et al. 2001, p. 343). Applied to the EU transposition context, this leads to the following
expectation:
Expectation 1: Transposition actors will only constrain discretion granted by EU directives in the case of policy
conflict between transposition actors and practical implementers concerning the directive.
Yet following Huber et al. (2001, p. 343) the risk of divergence is not a sufficient condition for adopting constraining
legislation. Crucially, transposition actors must also be able to cover the costs of writing and adopting constraining
legislation (Moe 1984; Huber & Shipan 2002, p. 79). These transaction costs involve resources, such as time and exper-
tise required when legislators need to convince critical veto players of their legal proposals. These costs are resources
that legislators could otherwise channel into other tasks. The bargaining environment determines the height of these
transaction costs (Huber et al. 2001). As argued by Oosterwaal et al. (2012, p. 803), transaction costs increase when
many veto players holding divergent preferences are involved in lawmaking. In such situations, adopting constraining
legislation is highly costly (Epstein & O’Halloran 1994, 1999; Huber & Shipan 2002) and typically leads to the adoption
of compromises with low detail and few specific descriptions (Oosterwaal et al. 2012).
In the EU transposition context, the involvement and preferences of veto players may vary across national
systems. In line with the transposition literature, we define veto players as actors who are either formally or
informally involved in the transposition process and who are required to find a compromise (Dimitrova &
Steunenberg 2013 p. 250). Sometimes, transposition involves only the relevant ministry, for example, when
delegated acts are adopted (Dimitrova & Steunenberg 2013, p. 249). Under such circumstances, no compromise
needs to be negotiated and transaction costs are low. In other transposition processes, powerful policy‐specific
players, such as several ministries and parliament, are actively involved (Dimitrova & Steunenberg 2000; Haverland
2000). These actors are often linked through networks with powerful non‐state actors that become informal veto
players (Dimitrova 2010, p 145). The inclusion of informal veto players takes into account that some member
states and policy areas involve strong interest groups in political decisionmaking (Lijphart & Crepaz 1991). Formal
and informal veto players may have preferences that diverge from those of the transposing ministry, which raises
the costs of agreeing on constraining legislation and makes the transposing minister incapable of writing detailed
legislation (Steunenberg 2006). This leads to the second expectation:
Expectation 2: Transpostion actors will only constrain discretion granted by EU directives in the case of low trans-
action costs resulting from low conflict between the agenda‐setter and veto players involved in the transposition
process.
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 3
Huber and Shipan (2002) suggest two additional variables that influence the level of discretion, which we control
for in this study as they relate to the policy context. The first condition refers to policy uncertainty. This condition entails
that principals will delegate discretion to agents if the latter possess highly technical expertise necessary for
implementing complex policies (Huber & Shipan 2002, 2011). For the purpose of this study, we control for this variable
by selecting a relatively non‐technical policy area.
Additionally, Huber et al. (2001) argue that the legislator might have cheaper non‐statutory mechanisms for moni-
toring and controlling the actions of agents that make it unnecessary to invest in detailed legislation. For example,
legislators could veto rules adopted by agencies during hearings. Alternatively, decisionmakers could rely on external
actors, such as judges, to hold agents accountable (Epstein & O’Halloran 1994; Huber et al. 2001, pp. 334–335;
Blom‐Hansen 2005). In the context of EU transposition and asylum law, such mechanisms are largely absent. While
the EU Commission can be seen as an external actor to control member state under‐compliance, the Commission
tends to focus on legal transposition and less on the actions of national agencies on the ground. In case the Commission
does target practical implementation, the possibility of enforcement applies to all member states, which reduces its
suitability to explain cross‐country differences in discretion. Finally, litigation by asylumseekers might be an alternative
to correct under‐compliance. However, asylum seekers are unlikely to use litigation in such a way to represent the
interest of the national principal against the agency; instead they have their own interests.
Moreover, there are no non‐statutory mechanisms to monitor over‐compliance. Over‐compliance exists when the
agency gold‐plates the requirements of the directive against the preferences of the transposition actors. Such activities
are unlikely to be followed up by courts because the law’s target group, as possible litigants, would benefit from such
over‐compliance. Finally, in fields such as migration and asylum law, many policies are applied in street‐level
situations (e.g. Ellermann 2006). This makes non‐statutory monitoring difficult, as legislative veto or alternative
mechanisms do not exist to monitor street‐level bureaucratic behavior.
3. Methods and data
The expectations developed above will be investigated through a set of in‐depth case studies, using a structured focused
comparison (George & Bennett 2005, p. 69). As will be elaborated, the qualitative design allows the employment of a
more nuanced and context‐specific measure of the dependent variable of the relative share of discretionary and
constraining provisions than commonly used in quantitative measures of discretion. The downside of this approach
is that inferences based on a small sample might have lower external validity than large‐N correlation studies. However,
our qualitative approach allows us to look into the underlying causal mechanisms in detail. Additionally, because this
study focuses on necessary conditions, a low‐N study is suitable (Mahoney & Goertz 2006, p. 232). For such set‐
theoretical assumptions, a single contradicting case can disqualify the necessity of a certain condition (Goertz & Starr
2003). Moreover, the qualitative design makes it possible to inductively explore alternative mechanisms that specify
existing theories, in case the established necessary conditions do not fully account for the expected outcomes.
We investigate our expectations through an analysis of the transposition of several provisions of the Asylum
Reception Conditions Directive 2003/9/EC in three EU member states. By focusing on the provision level, we add
leverage to the low‐N study because dependent variable and explanatory conditions can vary on the provision level.
The RCD was chosen because, while constituting a prescriptive framework, many of its provisions provide ample
discretion to the member states, which is a scope condition for exploring our expectations (Huber & Shipan 2002,
p. 181). In addition, it forms part of the highly politicized field of asylum, which provides promising conditions for
the expected mechanisms to operate.
The directive’s key aim is to limit “asylum shopping”which was thought to stem from varying reception standards
between EU member states (Zaun 2016). The directive prescribes minimum standards for the reception of asylum
seekers that will normally suffice to ensure them a dignified standard of living. Additionally, it regulates various types
of material support for asylum seekers, including housing, food, clothing, and non‐material conditions such as
healthcare and employment. Member states may adopt or maintain provisions that are more favorable.
The Council of Ministers of the EU negotiated the directive from 1997 until 2003. During this time, the EU faced a
moderate influx of asylum seekers, compared with the years that were to follow. Nevertheless, member states were
divided on the issue. Council documents stressed repeatedly that the final act should leave “sufficient room for
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
4© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 73

maneuver”as this was seen as the only way to find a compromise (see list of Council of Ministers documents in the
online appendix).
Four main provisions of the directive and their 19 sub‐provisions were chosen for in‐depth study (for wording of
the provisions, see Table S1). These provisions go to the core of the directive and constitute discretionary obligations:
•Art 11 (Access to Labor market): Access shall be given within one year but states shall establish the conditions for
doing so using national labor market policies.
•Art 13 (Material Conditions): States should guarantee a decent way of living. Benefits may be in monetary terms,
through voucher, or in kind. No provision specifies the amount of financial allowances. Benefits may be limited if
applicants possess sufficient own resources.
•Art 14 (Accommodation): Applicants shall be housed in centers, hotels, or apartments. There is no precision on
minimum standards beyond a guarantee of human dignity. Family members should be hosted together but can
be separated if appropriate. Access to non‐governmental organizations (NGOs) shall be given. Emergency
reception that limits the standards to emergency health care is possible.
•Art 15 & Art 17 (Health Care and Vulnerability): States must guarantee necessary health care and at least
emergency care and essential treatment of illness. There is no definition of “necessary healthcare.”
In measuring the dependent variable of our study, which is the relative share of discretionary and constraining
provisions in national transposition measures, we go beyond previous empirical studies. Most studies on discretion
used an index dividing the number of norms, which provide explicit leeway by the overall number of norms in a
statute (Epstein & O’Halloran 1994; Franchino 2007; Zhelyazkova & Torenvlied 2009). Huber and Shipan (2002)
used statutes’word counts to infer the level of discretion left to implementers. In our view, both measures
disregard characteristics of discretion connected to vague wordings at the provision level and, therefore, cannot
reflect different levels of discretion within provisions. For this reason, we follow the suggestion by Toshkov
(2013) to measure discretion systematically using the Institutional Grammar Tool (IGT; Crawford & Ostrom
1995; Basurto et al. 2010; Siddiki et al. 2012).
The IGT is a coding scheme that can be used to analyze legal documents. It provides a grammatical syntax to
classify legal provisions into several components, each of which may serve to either constrain or delegate discre-
tion. First, provisions specify an attribute, that is, the actor to whom the statement applies. In all provisions of the
RCD, the attribute is the member state. After transposition, attributes are the practical implementers to whom
discretion can be delegated, such as the operators of asylum reception centers. Mostly, this component leaves
no discretion.
Second, each provision has a deontic, indicating whether actions are obliged, permitted, or forbidden. In EU
directives, obligations take the shape of ‘shall’or ‘must clauses’. Permission, for its part, is signaled through
‘may clauses’. For example, Art 13.4 RCD states that “Member States may require asylum applicants (...) to con-
tribute to the cost (...)”of reception. National law can either preserve the may clause, or specify during transpo-
sition that the reception agency “should require (…)”applicants to contribute to the costs of reception. So far, the
deontic has been the main focus of existing delegation studies, because may clauses most explicitly signal discre-
tion (Epstein & O’Halloran 1994; Franchino 2007; Zhelyazkova & Torenvlied 2009). Of the selected RCD
sub‐provisions, five are may clauses and 14 are shall clauses. Focusing only on the deontic, the RCD appears
to be relatively discretion‐constrained.
However, thirdly, each legal provision also has an aim, meaning the task that should or should not be performed.
Aims can also add or decrease discretion. For example, Art 15.1 RCD holds that states shall “provide necessary
healthcare.”Despite the shall clause, the aim leaves discretion by leaving open the definition of “necessary healthcare.”
Finally, most legal provisions include conditions specifying the temporal, spatial, or procedural boundaries within
which a task is or is not to be performed. For example, Art 14.3 RCD obliges reception authorities to lodge children
together with their parents, “if appropriate,”which leaves member states discretion regarding the definition of
appropriateness.
We measured our dependent variable (DV) by first determining the aim, deontic, and condition components for
the 19 RCD sub‐provisions (Table S1). Secondly, for each component, we identified the corresponding national trans-
position provisions. Transposition reports, such as the Odysseus and AIDA country reports were used to identify the
relevant national legal texts (see Table S2 and a full list of reports in the online appendix).
3
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 5
Consequently, each national transposition component was scored with a minus (−) if it provided less discretion
than the corresponding legal provision of the directive, zero (0) if comparable discretion was granted, and a plus (+)
if transposition laws allowed for more discretion.
4
Table A1 provides the coding results.
While identifying the IGT components in the individual sub‐provisions of the RCD was relatively straightforward,
the coding of the transposition measures was more complex: it required legal interpretation based on transposition
reports, as well as making coding decisions (as explained in the online appendix).
5
We took the RCD as benchmark
for the coding. This means that if the aim of the directive provision was to establish conditions (e.g. Art 11.2 RCD),
the conditions that are eventually established in the transposition were coded as aim component and not as condition
component. Moreover, we focused on discretion given to practical implementers and did not consider it as discrete
when asylum seekers are allowed to choose from different options. Such provisions were interpreted in light of the
discretion practical implementers are left with.
While it is beyond the scope of this article to elaborate on the coding of each sub‐provision, we provide the original
wording of the transposition laws and identify the different components in the online appendix. Additionally, we
provide examples of difficult coding decisions to exemplify the coding. In the analysis, we add content analysis of
the provisions to illustrate the delegation of discretion across the components and to deepen the insights provided
by the IGT coding.
To measure the explanatory conditions, we conducted extensive document studies of Council documents, parlia-
mentary debates, committee reports, hearings, NGO reports, and the Odysseus country reports on transposition (see
online appendix). The data were complemented with interviews and email enquiries with respondents of relevant
ministries, implementing agencies, NGOs, and experts who followed the transposition processes. We investigated
the level of policy conflict between the implementing organizations and the agenda‐setting ministry regarding the
directive by outlining preferences concerning reception conditions of the ministry that was in charge of transposition.
The data sources were triangulated to identify preferences. First, in case the same minister was already in charge
of negotiating the directive, the preferences of agenda‐setters regarding specific articles of the directive could be
derived from Council documents. This provided information on which articles agenda‐setters were critical of
and in which direction they aimed to change the directive. In case a different minister was in charge of transpo-
sition, policy documents regarding transposition were investigated to trace their policy preferences regarding the
directive. Additionally, the preferences of the ministry were discussed in the interviews and email inquiries with
respondents who either worked for the agenda‐setting ministry at the time of transposition and/or who were
otherwise involved in the transposition process. Respondents were asked which goals the ministry aimed to realize
during transposition.
Next, the practical implementers of asylum reception, such as local authorities and organizers of receptions centers,
and their preferences were identified per country. The policy preferences of the practical implementers were traced by
asking representatives of implementing organizations which aims they had and which challenges they faced during the
transposition process and to what extent they had different goals as the ministry. This information was cross‐checked
with the reports of the Odysseus network, which include information on national policy debates.
If policy preferences of the relevant ministries and practical implementers converged and if respondents did not
recall any conflict, the risk of agency drift was considered to be low. If respondents indicated differences in opinion
but no explicit policy conflict between ministry and practical implementers emerged, this was scored as a moderate risk
of agency divergence. We assume that moderate risk still creates a desire on behalf of the agenda‐setter to constrain
discretion. When the agenda‐setting ministry and practical implementers recalled explicit policy conflict, the risk of
agency drift was considered to be high.
As for transaction costs, we first identified the players involved in the transposition process. To this end, first the
transposition instruments were identified. This provided a first indication of players formally involved in transposition,
such as parliament and coalition partners, and which informal actors had to be consulted, such as interest groups. In
order to find out which players had potential veto capacity and to identify substantive preferences, the individual
transposition processes were investigated using parliamentary documents of plenary debates, position papers, and
hearing reports, complemented with interviews with relevant actors. If no veto players were involved in transposition,
or if veto players shared the leading ministry’policy preference, transaction costs for constraining transposition
measures were considered to be low. If veto players with conflicting preferences were involved in the establishment
of the transposition law, transaction costs were classified as high.
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
6© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
74

maneuver”as this was seen as the only way to find a compromise (see list of Council of Ministers documents in the
online appendix).
Four main provisions of the directive and their 19 sub‐provisions were chosen for in‐depth study (for wording of
the provisions, see Table S1). These provisions go to the core of the directive and constitute discretionary obligations:
•Art 11 (Access to Labor market): Access shall be given within one year but states shall establish the conditions for
doing so using national labor market policies.
•Art 13 (Material Conditions): States should guarantee a decent way of living. Benefits may be in monetary terms,
through voucher, or in kind. No provision specifies the amount of financial allowances. Benefits may be limited if
applicants possess sufficient own resources.
•Art 14 (Accommodation): Applicants shall be housed in centers, hotels, or apartments. There is no precision on
minimum standards beyond a guarantee of human dignity. Family members should be hosted together but can
be separated if appropriate. Access to non‐governmental organizations (NGOs) shall be given. Emergency
reception that limits the standards to emergency health care is possible.
•Art 15 & Art 17 (Health Care and Vulnerability): States must guarantee necessary health care and at least
emergency care and essential treatment of illness. There is no definition of “necessary healthcare.”
In measuring the dependent variable of our study, which is the relative share of discretionary and constraining
provisions in national transposition measures, we go beyond previous empirical studies. Most studies on discretion
used an index dividing the number of norms, which provide explicit leeway by the overall number of norms in a
statute (Epstein & O’Halloran 1994; Franchino 2007; Zhelyazkova & Torenvlied 2009). Huber and Shipan (2002)
used statutes’word counts to infer the level of discretion left to implementers. In our view, both measures
disregard characteristics of discretion connected to vague wordings at the provision level and, therefore, cannot
reflect different levels of discretion within provisions. For this reason, we follow the suggestion by Toshkov
(2013) to measure discretion systematically using the Institutional Grammar Tool (IGT; Crawford & Ostrom
1995; Basurto et al. 2010; Siddiki et al. 2012).
The IGT is a coding scheme that can be used to analyze legal documents. It provides a grammatical syntax to
classify legal provisions into several components, each of which may serve to either constrain or delegate discre-
tion. First, provisions specify an attribute, that is, the actor to whom the statement applies. In all provisions of the
RCD, the attribute is the member state. After transposition, attributes are the practical implementers to whom
discretion can be delegated, such as the operators of asylum reception centers. Mostly, this component leaves
no discretion.
Second, each provision has a deontic, indicating whether actions are obliged, permitted, or forbidden. In EU
directives, obligations take the shape of ‘shall’or ‘must clauses’. Permission, for its part, is signaled through
‘may clauses’. For example, Art 13.4 RCD states that “Member States may require asylum applicants (...) to con-
tribute to the cost (...)”of reception. National law can either preserve the may clause, or specify during transpo-
sition that the reception agency “should require (…)”applicants to contribute to the costs of reception. So far, the
deontic has been the main focus of existing delegation studies, because may clauses most explicitly signal discre-
tion (Epstein & O’Halloran 1994; Franchino 2007; Zhelyazkova & Torenvlied 2009). Of the selected RCD
sub‐provisions, five are may clauses and 14 are shall clauses. Focusing only on the deontic, the RCD appears
to be relatively discretion‐constrained.
However, thirdly, each legal provision also has an aim, meaning the task that should or should not be performed.
Aims can also add or decrease discretion. For example, Art 15.1 RCD holds that states shall “provide necessary
healthcare.”Despite the shall clause, the aim leaves discretion by leaving open the definition of “necessary healthcare.”
Finally, most legal provisions include conditions specifying the temporal, spatial, or procedural boundaries within
which a task is or is not to be performed. For example, Art 14.3 RCD obliges reception authorities to lodge children
together with their parents, “if appropriate,”which leaves member states discretion regarding the definition of
appropriateness.
We measured our dependent variable (DV) by first determining the aim, deontic, and condition components for
the 19 RCD sub‐provisions (Table S1). Secondly, for each component, we identified the corresponding national trans-
position provisions. Transposition reports, such as the Odysseus and AIDA country reports were used to identify the
relevant national legal texts (see Table S2 and a full list of reports in the online appendix).
3
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 5
Consequently, each national transposition component was scored with a minus (−) if it provided less discretion
than the corresponding legal provision of the directive, zero (0) if comparable discretion was granted, and a plus (+)
if transposition laws allowed for more discretion.
4
Table A1 provides the coding results.
While identifying the IGT components in the individual sub‐provisions of the RCD was relatively straightforward,
the coding of the transposition measures was more complex: it required legal interpretation based on transposition
reports, as well as making coding decisions (as explained in the online appendix).
5
We took the RCD as benchmark
for the coding. This means that if the aim of the directive provision was to establish conditions (e.g. Art 11.2 RCD),
the conditions that are eventually established in the transposition were coded as aim component and not as condition
component. Moreover, we focused on discretion given to practical implementers and did not consider it as discrete
when asylum seekers are allowed to choose from different options. Such provisions were interpreted in light of the
discretion practical implementers are left with.
While it is beyond the scope of this article to elaborate on the coding of each sub‐provision, we provide the original
wording of the transposition laws and identify the different components in the online appendix. Additionally, we
provide examples of difficult coding decisions to exemplify the coding. In the analysis, we add content analysis of
the provisions to illustrate the delegation of discretion across the components and to deepen the insights provided
by the IGT coding.
To measure the explanatory conditions, we conducted extensive document studies of Council documents, parlia-
mentary debates, committee reports, hearings, NGO reports, and the Odysseus country reports on transposition (see
online appendix). The data were complemented with interviews and email enquiries with respondents of relevant
ministries, implementing agencies, NGOs, and experts who followed the transposition processes. We investigated
the level of policy conflict between the implementing organizations and the agenda‐setting ministry regarding the
directive by outlining preferences concerning reception conditions of the ministry that was in charge of transposition.
The data sources were triangulated to identify preferences. First, in case the same minister was already in charge
of negotiating the directive, the preferences of agenda‐setters regarding specific articles of the directive could be
derived from Council documents. This provided information on which articles agenda‐setters were critical of
and in which direction they aimed to change the directive. In case a different minister was in charge of transpo-
sition, policy documents regarding transposition were investigated to trace their policy preferences regarding the
directive. Additionally, the preferences of the ministry were discussed in the interviews and email inquiries with
respondents who either worked for the agenda‐setting ministry at the time of transposition and/or who were
otherwise involved in the transposition process. Respondents were asked which goals the ministry aimed to realize
during transposition.
Next, the practical implementers of asylum reception, such as local authorities and organizers of receptions centers,
and their preferences were identified per country. The policy preferences of the practical implementers were traced by
asking representatives of implementing organizations which aims they had and which challenges they faced during the
transposition process and to what extent they had different goals as the ministry. This information was cross‐checked
with the reports of the Odysseus network, which include information on national policy debates.
If policy preferences of the relevant ministries and practical implementers converged and if respondents did not
recall any conflict, the risk of agency drift was considered to be low. If respondents indicated differences in opinion
but no explicit policy conflict between ministry and practical implementers emerged, this was scored as a moderate risk
of agency divergence. We assume that moderate risk still creates a desire on behalf of the agenda‐setter to constrain
discretion. When the agenda‐setting ministry and practical implementers recalled explicit policy conflict, the risk of
agency drift was considered to be high.
As for transaction costs, we first identified the players involved in the transposition process. To this end, first the
transposition instruments were identified. This provided a first indication of players formally involved in transposition,
such as parliament and coalition partners, and which informal actors had to be consulted, such as interest groups. In
order to find out which players had potential veto capacity and to identify substantive preferences, the individual
transposition processes were investigated using parliamentary documents of plenary debates, position papers, and
hearing reports, complemented with interviews with relevant actors. If no veto players were involved in transposition,
or if veto players shared the leading ministry’policy preference, transaction costs for constraining transposition
measures were considered to be low. If veto players with conflicting preferences were involved in the establishment
of the transposition law, transaction costs were classified as high.
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
6© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 75

Transposition of the provisions of the directive was investigated in three member states: France, Germany, and the
Netherlands. The three countries share some general features. Crucially, none of these states had to build a reception
system from scratch and they were among the strongest negotiators of the directive in the Council negotiations (Zaun
2016, p. 138). Thus, the challenge of transposing the directive was similarly low for all three member states.
Additionally, the three countries have similar administrative and financial capacity to host asylum seekers. Third, all
three countries have a civil law tradition, which is, following Huber and Shipan (2002), an important scope condition
for similar levels of legal discretion. Additionally, in all three countries, migration was relatively high on the political
agenda, which is a favorable condition for observing policy conflict and transaction costs. Furthermore, during
transposition the states received comparable numbers of asylum applicants (Eurostat 2007).
Previous studies found that some member states literally transpose EU directives into national law that could influ-
ence levels of discretion (Steunenberg & Voermans 2006). While the Netherlands has the reputation to sometimes use
this copy‐out method, this was not the case for the RCD (Odysseus_Report_NL, 2006 p. 4, Odyssey_Report_general,
2006 p. 17).
6
The three countries were selected in such a way as to maximize the chances of observing variance on the two
explanatory conditions. France is a centralized state but practical implementation of asylum reception is delegated
to a range of practical implementers at the central and local level. Thus, there are good conditions to observe
policy conflict between transposition actors and the practical implementers. Additionally, France, with its major-
itarian system, is not known for its inclusion of veto players in the lawmaking process (Lijphart & Crepaz 1991).
Thus, the country provides good conditions to observe low transaction costs, making the expected mechanism of
constrained discretion likely.
Germany, as a federal state, delegates practical implementation of asylum reception to a range of state and local
actors. Following Huber and Shipan (2002), federal arrangements provide good conditions for policy conflict between
federal legislators and local practical implementers. The federal structure also suggests a strong role of formal veto
players with divergent preferences, which increases transaction costs for constraining transposition, providing good
conditions to explore the mechanisms of expectation two.
Finally, the Netherlands, as a centralized state, delegates practical implementation of asylum reception to a single
state agency, providing favorable conditions for low policy conflict between decisionmaker and practical implementers.
Additionally, Dutch lawmaking is known for its consensus‐oriented decisionmaking, which tends to include a range of
formal and informal veto players (Lijphart & Crepaz 1991). This may increase transaction costs for constraining trans-
position and makes the Dutch case an unlikely one for constrained discretion. Table 1 summarizes the case selection
criteria.
In the following, the case descriptions are presented. These have a theory‐testing character. We start with a most
likely case for constraining mechanisms (France), followed by a critical case to investigate expectation 2 (Germany),
and conclude with a least‐likely case for constraining mechanisms (Netherlands). For each case, we first assess the level
of policy conflict between practical implementers and the leading ministry as agenda‐setter. Second, we explore the
transaction costs by identifying the preferences of players involved in the transposition and their capacity to veto
decisionmaking. Third, we compare our expectations to the characteristics of the case, resulting in a case‐specific
expectation for the degree of discretion. We then analyze this expectation by assessing the extent of discretion across
the transposition provisions.
Table 1 Case Selection
Likelihood of France Germany Netherlands
Risk of deviation: Policy conflict between practical
implementers and leading ministry
High High Low
Transaction costs: Policy conflict between
transposition actors
Low High High
Discretion Low Low High
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 7
4. France
First, we investigate the application of the delegation theories in the French context. To assess the desire of the
transposition actors to constrain discretion in the French transposition law, first possible policy conflict between
legislator and implementing organization(s) vis‐à‐vis the directive is investigated. During negotiation and transposition
of the RCD, a conservative government governed France with a unified majority for the Union for a Popular Movement
that held restrictive views on asylum. The French delegation of the Ministry of Immigration aimed at limiting the scope
of the directive during Council negotiations, albeit without success (Resp. 1, Council doc. 5791/02). It upheld this
approach during transposition (Resp. 2, 21, Odysseus_Report_ Fr 2006).
Concerning the preferences of practical implementers, it is important to note that in France, several actors implement
reception policy. Besides the central French Immigration and Integration Service, which was founded in its current form
during the transposition phase (Aida 2013; Resp. 6), NGOs and local authorities at the département level are involved. As
a result, this agency’s preferences were uncertain. NGOs, which operate reception centers on the ground, favored a lib-
eral reception system. Nevertheless, they were not collectively organized and did not speak with one voice (Resp. 2, 21).
Some implementing NGOs articulated their opposition to the agenda‐setting ministry, but most NGOs did not want to
risk budget cuts for being too critical (Resp. 2). Overall, there was no open policy conflict between practical implemen-
ters and transposition actors (Odysseus_Report_Fr 2006). However, policy preferences of practical implementers were
divergent or uncertain. In light of our first expectation,this implied a moderate risk of divergence for the ministry.
Turning to the transaction costs, it is important to note that France did not adopt a specific statute to transpose the
directive, but only modified existing decrees. The decreesdid not require approval by parliament or other potential veto
players (Resp. 2, 5). Instead, transposition was the province of the Secretary‐General of the Ministry of Immigration,
Stefanini. He pulled the strings in the Ministry and did not consult any other actors during transposition (Resp. 2).
Consequently, neither the governing parties nor the opposition explicitly discussed transposition (Odysseus_Report_Fr
2006; Resp. 2, 4). Additionally, there was no hearing of informal veto players, such as NGOs. Thus, because of the ab-
sence of veto players, the transaction costs of negotiating constraining provisions were low.Given the presence of mod-
erate policy conflict combined with low transaction costs, we expect French transposition to be relatively constraining.
Our analysis of French reception laws shows, first of all, that while France notified the Commission of three decrees,
French transposition laws were more complex. France had no unified national asylum law, but a range of very different
general laws that transposed the directive, such as general welfare provisions, labor rights, laws on social institutions,
and general health care, as well as temporary waiting and emergency regulations.
7
In line with the expectations, French
legislation on reception entailed mainly constraining provisions, as compared to the directive. Table A1 shows that 18
components of French legislation constrained discretion. Two optional clauses were not transposed. Fifteen compo-
nents were comparable to the directive, and in five components, discretion extended beyond the directive. Discretion
also existed where two obligatory provisions were not transposed. The deontic components remained mostly
unchanged compared to the directive, particularly regarding “shall”provisions. The aim and condition components
were mostly constraining or comparable to the directive.
Examination of the substance of discretion reveals that compared to the RCD, French regulations constrained discre-
tion, particularly in the conditions for access to the labor market and application procedures. However, French law explic-
itly allowed prefects to establish additional conditions to access the labor market (aim of Art 11.2 RCD). Also concerning
material reception, French law was constraining (aim of Art 13 RCD). For example, it obliged implementers to provide
benefits in monetary terms. While French law obliged reception centers to demand a contribution from applicants in case
they had sufficient own resources (deontic of Art 13 RCD) and specified the condition under which contributions could
be asked, the amount and enforcement of the contribution was left to prefectures and local implementers.
For those applicants who received a place in a reception center, accommodation was regulated with detail by
specifying the living space per applicant and hygiene criteria (aim of Art 13.1 RCD, 14.1 RCD).
8
Access to NGOs
was regulated only implicitly because NGOs operate the reception centers. Health care was regulated in detail under
the universal healthcare scheme. Some discretion existed regarding securing family unity (constraining deontic, but
vague conditions), training of reception center staff, and transfer between centers (both unregulated).
For applicants without a place in a reception center, emergency legislation was established in response to the direc-
tive. The legislation was more elaborate than Art 14.8 RCD. However, in cases of budget shortage or overfill, French law
allowed for the circumventing of several constraining standards set in the other transposition provisions.
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
8© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
76

Transposition of the provisions of the directive was investigated in three member states: France, Germany, and the
Netherlands. The three countries share some general features. Crucially, none of these states had to build a reception
system from scratch and they were among the strongest negotiators of the directive in the Council negotiations (Zaun
2016, p. 138). Thus, the challenge of transposing the directive was similarly low for all three member states.
Additionally, the three countries have similar administrative and financial capacity to host asylum seekers. Third, all
three countries have a civil law tradition, which is, following Huber and Shipan (2002), an important scope condition
for similar levels of legal discretion. Additionally, in all three countries, migration was relatively high on the political
agenda, which is a favorable condition for observing policy conflict and transaction costs. Furthermore, during
transposition the states received comparable numbers of asylum applicants (Eurostat 2007).
Previous studies found that some member states literally transpose EU directives into national law that could influ-
ence levels of discretion (Steunenberg & Voermans 2006). While the Netherlands has the reputation to sometimes use
this copy‐out method, this was not the case for the RCD (Odysseus_Report_NL, 2006 p. 4, Odyssey_Report_general,
2006 p. 17).
6
The three countries were selected in such a way as to maximize the chances of observing variance on the two
explanatory conditions. France is a centralized state but practical implementation of asylum reception is delegated
to a range of practical implementers at the central and local level. Thus, there are good conditions to observe
policy conflict between transposition actors and the practical implementers. Additionally, France, with its major-
itarian system, is not known for its inclusion of veto players in the lawmaking process (Lijphart & Crepaz 1991).
Thus, the country provides good conditions to observe low transaction costs, making the expected mechanism of
constrained discretion likely.
Germany, as a federal state, delegates practical implementation of asylum reception to a range of state and local
actors. Following Huber and Shipan (2002), federal arrangements provide good conditions for policy conflict between
federal legislators and local practical implementers. The federal structure also suggests a strong role of formal veto
players with divergent preferences, which increases transaction costs for constraining transposition, providing good
conditions to explore the mechanisms of expectation two.
Finally, the Netherlands, as a centralized state, delegates practical implementation of asylum reception to a single
state agency, providing favorable conditions for low policy conflict between decisionmaker and practical implementers.
Additionally, Dutch lawmaking is known for its consensus‐oriented decisionmaking, which tends to include a range of
formal and informal veto players (Lijphart & Crepaz 1991). This may increase transaction costs for constraining trans-
position and makes the Dutch case an unlikely one for constrained discretion. Table 1 summarizes the case selection
criteria.
In the following, the case descriptions are presented. These have a theory‐testing character. We start with a most
likely case for constraining mechanisms (France), followed by a critical case to investigate expectation 2 (Germany),
and conclude with a least‐likely case for constraining mechanisms (Netherlands). For each case, we first assess the level
of policy conflict between practical implementers and the leading ministry as agenda‐setter. Second, we explore the
transaction costs by identifying the preferences of players involved in the transposition and their capacity to veto
decisionmaking. Third, we compare our expectations to the characteristics of the case, resulting in a case‐specific
expectation for the degree of discretion. We then analyze this expectation by assessing the extent of discretion across
the transposition provisions.
Table 1 Case Selection
Likelihood of France Germany Netherlands
Risk of deviation: Policy conflict between practical
implementers and leading ministry
High High Low
Transaction costs: Policy conflict between
transposition actors
Low High High
Discretion Low Low High
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 7
4. France
First, we investigate the application of the delegation theories in the French context. To assess the desire of the
transposition actors to constrain discretion in the French transposition law, first possible policy conflict between
legislator and implementing organization(s) vis‐à‐vis the directive is investigated. During negotiation and transposition
of the RCD, a conservative government governed France with a unified majority for the Union for a Popular Movement
that held restrictive views on asylum. The French delegation of the Ministry of Immigration aimed at limiting the scope
of the directive during Council negotiations, albeit without success (Resp. 1, Council doc. 5791/02). It upheld this
approach during transposition (Resp. 2, 21, Odysseus_Report_ Fr 2006).
Concerning the preferences of practical implementers, it is important to note that in France, several actors implement
reception policy. Besides the central French Immigration and Integration Service, which was founded in its current form
during the transposition phase (Aida 2013; Resp. 6), NGOs and local authorities at the département level are involved. As
a result, this agency’s preferences were uncertain. NGOs, which operate reception centers on the ground, favored a lib-
eral reception system. Nevertheless, they were not collectively organized and did not speak with one voice (Resp. 2, 21).
Some implementing NGOs articulated their opposition to the agenda‐setting ministry, but most NGOs did not want to
risk budget cuts for being too critical (Resp. 2). Overall, there was no open policy conflict between practical implemen-
ters and transposition actors (Odysseus_Report_Fr 2006). However, policy preferences of practical implementers were
divergent or uncertain. In light of our first expectation,this implied a moderate risk of divergence for the ministry.
Turning to the transaction costs, it is important to note that France did not adopt a specific statute to transpose the
directive, but only modified existing decrees. The decreesdid not require approval by parliament or other potential veto
players (Resp. 2, 5). Instead, transposition was the province of the Secretary‐General of the Ministry of Immigration,
Stefanini. He pulled the strings in the Ministry and did not consult any other actors during transposition (Resp. 2).
Consequently, neither the governing parties nor the opposition explicitly discussed transposition (Odysseus_Report_Fr
2006; Resp. 2, 4). Additionally, there was no hearing of informal veto players, such as NGOs. Thus, because of the ab-
sence of veto players, the transaction costs of negotiating constraining provisions were low.Given the presence of mod-
erate policy conflict combined with low transaction costs, we expect French transposition to be relatively constraining.
Our analysis of French reception laws shows, first of all, that while France notified the Commission of three decrees,
French transposition laws were more complex. France had no unified national asylum law, but a range of very different
general laws that transposed the directive, such as general welfare provisions, labor rights, laws on social institutions,
and general health care, as well as temporary waiting and emergency regulations.
7
In line with the expectations, French
legislation on reception entailed mainly constraining provisions, as compared to the directive. Table A1 shows that 18
components of French legislation constrained discretion. Two optional clauses were not transposed. Fifteen compo-
nents were comparable to the directive, and in five components, discretion extended beyond the directive. Discretion
also existed where two obligatory provisions were not transposed. The deontic components remained mostly
unchanged compared to the directive, particularly regarding “shall”provisions. The aim and condition components
were mostly constraining or comparable to the directive.
Examination of the substance of discretion reveals that compared to the RCD, French regulations constrained discre-
tion, particularly in the conditions for access to the labor market and application procedures. However, French law explic-
itly allowed prefects to establish additional conditions to access the labor market (aim of Art 11.2 RCD). Also concerning
material reception, French law was constraining (aim of Art 13 RCD). For example, it obliged implementers to provide
benefits in monetary terms. While French law obliged reception centers to demand a contribution from applicants in case
they had sufficient own resources (deontic of Art 13 RCD) and specified the condition under which contributions could
be asked, the amount and enforcement of the contribution was left to prefectures and local implementers.
For those applicants who received a place in a reception center, accommodation was regulated with detail by
specifying the living space per applicant and hygiene criteria (aim of Art 13.1 RCD, 14.1 RCD).
8
Access to NGOs
was regulated only implicitly because NGOs operate the reception centers. Health care was regulated in detail under
the universal healthcare scheme. Some discretion existed regarding securing family unity (constraining deontic, but
vague conditions), training of reception center staff, and transfer between centers (both unregulated).
For applicants without a place in a reception center, emergency legislation was established in response to the direc-
tive. The legislation was more elaborate than Art 14.8 RCD. However, in cases of budget shortage or overfill, French law
allowed for the circumventing of several constraining standards set in the other transposition provisions.
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
8© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 77

Overall, the French case fits our expectations relatively well. In line with the expectations, moderate policy con-
flict combined with low transaction costs fit the overall constraining reception laws. However, the analysis also shows
that while French legislation formally complied with the directive, it “passed the buck”to practical implementers in
some instances, such as emergency situations or regarding staff training. Thus, contrary to the expectations, French
transposition measures also contain discretionary provisions, particularly in the aim and condition components or
through missing regulations that allowed practical implementers to interpret EU obligations. Hence, the French case
shows that transposition actors do not always settle for point‐specific transposition laws, even if necessary conditions
are present.
5. Germany
Next, we investigate the delegation of discretion in a federal context. We start by investigating the policy conflict between
transposition actors and implementers. The German transposition law was drafted in 2007, two years after the trans-
position deadline, by the Ministry of Internal Affairs led by the Christian‐Democratic (CDU) Minister, Schäuble, in
cooperation with the Ministry of Social Affairs led by the Social Democratic (SPD) Minister, Müntefering.
During negotiation and transposition of the directive, preferences of the Ministries were rather restrictive (Resp. 10).
9
The German delegation stressed continuously during Council negotiations that it would only agree to changes in German
lawifthe16Bundesländer agreed (Resp. 1, 10). With the exception of access to the labor market, the Länder were fully
responsible for the implementation of asylum reception (Resp. 10, 8). In line with the principle of concurrent legislation,
the Länder could even pass subsequent regulations as long as the federal level did not do so. Bavaria, Baden‐Württemberg,
and Hessen were most determined not to agree on any modifications of the federal law (Resp. 10, 11).
The Länder were not the only practical implementers as they had delegated much reception implementation down
to municipalities, which in turn had delegated to public actors, hired welfare organizations, or private companies to run
reception centers at the local level. The municipalities coordinated their preferences within the Städtetag, which is an
Association of German cities (Resp. 8). Interests between municipalities, NGOs, and private actors who implement
reception on the ground were highly diverse within and across the Länder. Thus, there was some debate among prac-
titioners on the directive (Odysseus_Report_De 2006). While this debate did not lead to an open policy conflict with
the government, the diversity of practical implementers and their varying interests at the local level can be understood
as creating at least a moderate risk of divergence, similar to the French case.
Turning to the second condition, transaction costs, the existence and preferences of veto players are relevant. As two
ministries from different coalition parties acted as agenda‐setters for the transposition measure, they could have vetoed
each other and thus raised transaction costs. However, there was no divergence in preferences regarding the directive
between the two Ministries (Resp. 10). The Bundestag was another formal veto player because Germany adopted a
transposition instrument that required parliamentary approval. However, although several Social‐Democratic parlia-
mentarians criticized the general package law (Drucksache_16/5527, 30. 05. 2007; Drucksache_16/5654, 13. 06.
2007; Drucksache 16/5621, 13. 06. 2007), no actual parliamentary debate concerning the RCD ensued, indicating that
there was no explicit conflict regarding the directive amongparliamentarianscompared to the agenda‐setting ministries
(Odysseus_Report_De 2006, Resp. 11).
Another formal veto player with strong policy preference was the Bundesrat, whose consent was required (Resp. 1,
7, 8, 10, 11). The Bundesrat represents the interests of the Länder. As elaborated above, the majority of the Länder did
not want to change the existing law. They particularly opposed the more liberal reception laws because they financed
asylum reception and feared rising costs (Resp. 8, 10). Following the respondents, the Bundesrat took these preferences
as basis for negotiation with the Ministries.
Following Dimitrova & Steunenberg (2013), NGOs can also be informal veto players because of state capture. In
Germany, a range of NGOs was formally consulted in parliamentary hearings and nine NGOs established a common
position, trying to pressure the government to ensure more liberal reception conditions (Resp. 7, 8, 13). However,
investigating their capacity to effectively veto the transposition law shows that their evaluations were in fact not required
to find a compromise. Thus, even though the NGOs held divergent preferences compared to the agenda‐setting
ministries, they cannot be considered as veto players with the capacity to raise transaction costs.
Overall, despite the large number of actors involved in the transposition process, players either held convergent
preferences regarding the directive, or had no veto capacities. Thus, the transaction costs in the German case were
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 9
low. Following the established theories, this leads to the case‐specific expectation that constraining transposition would
be adopted.
Surprisingly, however, German reception laws delegated extensive discretion to the practical implementers. In
several components, the law remained even more discretionary than allowed by the directive. As shown in Table A1
German transposition law constrained discretion in only 12 components. Ten components delegated similar levels
of discretion as the RCD, and two optional clauses were not transposed. In 14 components, German law was even more
discretionary than the directive. Similar to France, two obligatory provisions were not transposed. Again, most deontics
remained similar to those of the RDC’s provisions. The aim and conditions components differed in the levels of
discretion compared with the RCD.
Transposition measures were partly constraining with regard to access to the labor market, describing a relatively
detailed priority right for EU citizens (aim of Art 11 RCD), while leaving some discretion regarding the condition by
only indirectly connecting the restrictions to the date of making the asylum application.
German law delegated more explicit discretion regarding material conditions. For example, regarding Art 13.5
RCD, German law gives priority to benefits in kind but “if necessary due to the circumstances”(discretionary condi-
tion) benefits can (keeps “may”deontic) be given in cash or through vouchers (keeps discretion in aim). Regarding
Art 13.2 RCD, most discretion is constrained as German law set a specific amount for pocket money. However, the
transposition law did not convert the amounts of pocket money into Euro. Instead, German law kept the pre‐existing
amounts written in the former German currency. This led to several years of legal uncertainty regardingthe appropriate
amounts and eventually required a Constitutional Court ruling.
Turning to accommodation, no regulation determined who should operate reception centers at the Länder level.
Thus, for Art 14 RCD, the attribute remained unspecified. There are also no federal quality standards regarding recep-
tion centers and their staff. Only a minority of Bundesländer had binding standards. Access to NGOs was only vaguely
regulated, similar to provisions on family unity.
Health care was also transposed in a particularly discretionary way. German law stated that necessary medical or
dental treatment only had to be provided under acute illnesses and conditions of pain. Thus, the transposition of
Art 15.1 RCD shows that the deontic can be constraining but the overall provision remains discretionary as it is not
specified when and which treatment is covered. Other benefits may be granted “if they are indispensable in an individ-
ual case to secure health.”Thus, regarding Art 15.2 RCD, it remained under local discretion if chronic sickness, dental
care, and treatment of vulnerable people were covered (Art 15.2 RCD deontic and aim).
The high level of discretion in the German transposition law highlights some particularities of the transposition
context. While domestic delegation theories assume implicitly that at least the legislative actor who initiates new legis-
lation is an advocate of change, this was not the case during the transposition of the RCD in Germany. Although mem-
bers of different parties led the two leading federal Ministries, they agreed to modify the existing reception conditions as
little as possible because renegotiating the status quo would have been politically and financially expensive (Resp. 11).
10
This feature highlights the importance of preferences for the status quo as a special aspect of the transposition context
where the EU level imposes legal change. In the German case, the unwillingness to invest in constraining legislation
was not a result of high transaction costs because of policy conflict between veto players, as assumed by domestic del-
egation theories. Instead, there was a unified preference among decisionmakers to maintain the legal status quo. Based
on the inductive findings for the German case, preferences for the status quo deserve more explicit attention in
explaining discretion after transposition than is the case in the domestic delegation literature.
6. The Netherlands
In order to investigate the desire to constrain discretion in the Dutch context, we first explore the level of policy conflict
between transposition and implementing actors. Here, one should note that the Dutch decisionmakers held less critical
views toward the RCD than in the other two states (Resp. 1). Transposition was the responsibility of a coalition of the
Christian Democratic Appeal, the People’s Party for Freedom and Democracy, and Democrats 66. While the respon-
sible Minister, Verdonk (People’s Party for Freedom and Democracy), held restrictive views on migration, centrist
parties held a majority. The government took a rather liberal stance during transposition (Resp. 14).
The agency responsible for implementing reception measures in the Netherlands is the Central Agency for the
Reception of Asylum Seekers (COA). This central state agency is an independent administrative body operating under
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
10 © 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
78

Overall, the French case fits our expectations relatively well. In line with the expectations, moderate policy con-
flict combined with low transaction costs fit the overall constraining reception laws. However, the analysis also shows
that while French legislation formally complied with the directive, it “passed the buck”to practical implementers in
some instances, such as emergency situations or regarding staff training. Thus, contrary to the expectations, French
transposition measures also contain discretionary provisions, particularly in the aim and condition components or
through missing regulations that allowed practical implementers to interpret EU obligations. Hence, the French case
shows that transposition actors do not always settle for point‐specific transposition laws, even if necessary conditions
are present.
5. Germany
Next, we investigate the delegation of discretion in a federal context. We start by investigating the policy conflict between
transposition actors and implementers. The German transposition law was drafted in 2007, two years after the trans-
position deadline, by the Ministry of Internal Affairs led by the Christian‐Democratic (CDU) Minister, Schäuble, in
cooperation with the Ministry of Social Affairs led by the Social Democratic (SPD) Minister, Müntefering.
During negotiation and transposition of the directive, preferences of the Ministries were rather restrictive (Resp. 10).
9
The German delegation stressed continuously during Council negotiations that it would only agree to changes in German
lawifthe16Bundesländer agreed (Resp. 1, 10). With the exception of access to the labor market, the Länder were fully
responsible for the implementation of asylum reception (Resp. 10, 8). In line with the principle of concurrent legislation,
the Länder could even pass subsequent regulations as long as the federal level did not do so. Bavaria, Baden‐Württemberg,
and Hessen were most determined not to agree on any modifications of the federal law (Resp. 10, 11).
The Länder were not the only practical implementers as they had delegated much reception implementation down
to municipalities, which in turn had delegated to public actors, hired welfare organizations, or private companies to run
reception centers at the local level. The municipalities coordinated their preferences within the Städtetag, which is an
Association of German cities (Resp. 8). Interests between municipalities, NGOs, and private actors who implement
reception on the ground were highly diverse within and across the Länder. Thus, there was some debate among prac-
titioners on the directive (Odysseus_Report_De 2006). While this debate did not lead to an open policy conflict with
the government, the diversity of practical implementers and their varying interests at the local level can be understood
as creating at least a moderate risk of divergence, similar to the French case.
Turning to the second condition, transaction costs, the existence and preferences of veto players are relevant. As two
ministries from different coalition parties acted as agenda‐setters for the transposition measure, they could have vetoed
each other and thus raised transaction costs. However, there was no divergence in preferences regarding the directive
between the two Ministries (Resp. 10). The Bundestag was another formal veto player because Germany adopted a
transposition instrument that required parliamentary approval. However, although several Social‐Democratic parlia-
mentarians criticized the general package law (Drucksache_16/5527, 30. 05. 2007; Drucksache_16/5654, 13. 06.
2007; Drucksache 16/5621, 13. 06. 2007), no actual parliamentary debate concerning the RCD ensued, indicating that
there was no explicit conflict regarding the directive amongparliamentarianscompared to the agenda‐setting ministries
(Odysseus_Report_De 2006, Resp. 11).
Another formal veto player with strong policy preference was the Bundesrat, whose consent was required (Resp. 1,
7, 8, 10, 11). The Bundesrat represents the interests of the Länder. As elaborated above, the majority of the Länder did
not want to change the existing law. They particularly opposed the more liberal reception laws because they financed
asylum reception and feared rising costs (Resp. 8, 10). Following the respondents, the Bundesrat took these preferences
as basis for negotiation with the Ministries.
Following Dimitrova & Steunenberg (2013), NGOs can also be informal veto players because of state capture. In
Germany, a range of NGOs was formally consulted in parliamentary hearings and nine NGOs established a common
position, trying to pressure the government to ensure more liberal reception conditions (Resp. 7, 8, 13). However,
investigating their capacity to effectively veto the transposition law shows that their evaluations were in fact not required
to find a compromise. Thus, even though the NGOs held divergent preferences compared to the agenda‐setting
ministries, they cannot be considered as veto players with the capacity to raise transaction costs.
Overall, despite the large number of actors involved in the transposition process, players either held convergent
preferences regarding the directive, or had no veto capacities. Thus, the transaction costs in the German case were
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 9
low. Following the established theories, this leads to the case‐specific expectation that constraining transposition would
be adopted.
Surprisingly, however, German reception laws delegated extensive discretion to the practical implementers. In
several components, the law remained even more discretionary than allowed by the directive. As shown in Table A1
German transposition law constrained discretion in only 12 components. Ten components delegated similar levels
of discretion as the RCD, and two optional clauses were not transposed. In 14 components, German law was even more
discretionary than the directive. Similar to France, two obligatory provisions were not transposed. Again, most deontics
remained similar to those of the RDC’s provisions. The aim and conditions components differed in the levels of
discretion compared with the RCD.
Transposition measures were partly constraining with regard to access to the labor market, describing a relatively
detailed priority right for EU citizens (aim of Art 11 RCD), while leaving some discretion regarding the condition by
only indirectly connecting the restrictions to the date of making the asylum application.
German law delegated more explicit discretion regarding material conditions. For example, regarding Art 13.5
RCD, German law gives priority to benefits in kind but “if necessary due to the circumstances”(discretionary condi-
tion) benefits can (keeps “may”deontic) be given in cash or through vouchers (keeps discretion in aim). Regarding
Art 13.2 RCD, most discretion is constrained as German law set a specific amount for pocket money. However, the
transposition law did not convert the amounts of pocket money into Euro. Instead, German law kept the pre‐existing
amounts written in the former German currency. This led to several years of legal uncertainty regardingthe appropriate
amounts and eventually required a Constitutional Court ruling.
Turning to accommodation, no regulation determined who should operate reception centers at the Länder level.
Thus, for Art 14 RCD, the attribute remained unspecified. There are also no federal quality standards regarding recep-
tion centers and their staff. Only a minority of Bundesländer had binding standards. Access to NGOs was only vaguely
regulated, similar to provisions on family unity.
Health care was also transposed in a particularly discretionary way. German law stated that necessary medical or
dental treatment only had to be provided under acute illnesses and conditions of pain. Thus, the transposition of
Art 15.1 RCD shows that the deontic can be constraining but the overall provision remains discretionary as it is not
specified when and which treatment is covered. Other benefits may be granted “if they are indispensable in an individ-
ual case to secure health.”Thus, regarding Art 15.2 RCD, it remained under local discretion if chronic sickness, dental
care, and treatment of vulnerable people were covered (Art 15.2 RCD deontic and aim).
The high level of discretion in the German transposition law highlights some particularities of the transposition
context. While domestic delegation theories assume implicitly that at least the legislative actor who initiates new legis-
lation is an advocate of change, this was not the case during the transposition of the RCD in Germany. Although mem-
bers of different parties led the two leading federal Ministries, they agreed to modify the existing reception conditions as
little as possible because renegotiating the status quo would have been politically and financially expensive (Resp. 11).
10
This feature highlights the importance of preferences for the status quo as a special aspect of the transposition context
where the EU level imposes legal change. In the German case, the unwillingness to invest in constraining legislation
was not a result of high transaction costs because of policy conflict between veto players, as assumed by domestic del-
egation theories. Instead, there was a unified preference among decisionmakers to maintain the legal status quo. Based
on the inductive findings for the German case, preferences for the status quo deserve more explicit attention in
explaining discretion after transposition than is the case in the domestic delegation literature.
6. The Netherlands
In order to investigate the desire to constrain discretion in the Dutch context, we first explore the level of policy conflict
between transposition and implementing actors. Here, one should note that the Dutch decisionmakers held less critical
views toward the RCD than in the other two states (Resp. 1). Transposition was the responsibility of a coalition of the
Christian Democratic Appeal, the People’s Party for Freedom and Democracy, and Democrats 66. While the respon-
sible Minister, Verdonk (People’s Party for Freedom and Democracy), held restrictive views on migration, centrist
parties held a majority. The government took a rather liberal stance during transposition (Resp. 14).
The agency responsible for implementing reception measures in the Netherlands is the Central Agency for the
Reception of Asylum Seekers (COA). This central state agency is an independent administrative body operating under
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
10 © 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 79

the political responsibility of the Ministry of Security and Justice. During both negotiation and transposition of the
RCD, the Ministry consulted COA extensively, in order to find out if the COA already complied with the directive
and what had to be changed (Resp. 14, 19). Respondents from the Ministry and the COA recalled no differences in
opinion regarding the directive. The only problematic issue was that more categories of asylum seekers had to be guar-
anteed reception under the RCD than under existing Dutch law, which created some capacity problems for the COA
(Resp. 14). Because of the limited policy conflict, the risk of agency drift was low for the Ministry, suggesting that the
desire to constrain discretion was also low.
Secondly, transaction costs were moderate. Although the Netherlands transposed the directive with a ministerial de-
cree (Rva 2005) that did notrequire parliamentary approval, parliament was consulted, as wellas the Dutch Council for
Refugees. This central NGO became an active player during transposition (Odysseus_Report_Nl 2006). This informal
player successfully started an initiative that triggered a debate among formal veto players in parliament. The debate con-
cerned the financial allowances for asylum seekers (Parliamentary proceeding 47 2005; Resp. 17). In response to this
debate, the opposition successfully carried a motion to modify the transposition decree in such a way that financial
allowances gradually increase. The Minister agreed to adopt this motion but stressed that it was not necessary to
comply with the RCD (DVB 2005). Additionally, a parliamentary debate on reception conditions for Dublin cases
emerged because of the directive, eventually leading to the inclusion of these cases under the reception scheme
(Odysseus_Report_Nl 2006). Thus, although the Dutch transposition measure was a decree, some veto players
with opposing views were involved in the transposition process. It can be argued that the pressure to find a
compromise with these actors increased the transaction costs of constraining legislation.
Following theories on domestic delegation, we expect that lowpolicy conflict and moderate transaction costs would
trigger the Ministry to adopt discretionary transposition laws. However, analysis of Dutch transposition reveals that,
contrary to the expectations, most discretion in the directive was constrained. As illustrated in Table A1 in 25 compo-
nents, discretion was limited, and 18 components were framed similarly to the directive. Still, in six components,
discretion extended beyond the directive. Contrary to the other two cases, all obligatory provisions were transposed.
Similarly to the other two cases, the deontic components remained mostly similar to those in the directive. The
conditions and aims mostly constrained discretion.
The conditions regarding access to the labor market were more specific in aims and conditions than in the RCD.
Dutch transposition measures detailed the period for which asylum seekers may work per year, and the documents
and administrative steps to be taken by applicants and employers (condition of Art 13.2 RCD).
Coming to material conditions, Art 13.1 RCD contained no condition component, but Dutch law introduced a
constraining condition specifying that benefits need to be given “in any case.”Weekly allowances for food, clothes,
and other expenses need to be in cash and are listed in detail, including public transport allowances. Although the
COA could decide whether the asylum seekers’own resources are checked, the conditions under which this was
possible constrained the agency’s discretion. In contrast to the other two cases, access to NGOs was regulated in more
detail than in the directive (Art 14.5 RCD). The main issue not regulated by fixed guidelines was the distribution of
asylum seekers across the centers. Contrary to the directive, which holds that transfer from one center to another
should only occur “if necessary,”there was no such condition in Dutch law (Art 14.4 RCD). However, the other
two countries did not include this provision at all in their legislation.
Similarly to France, an emergency system existed but it only delegated discretion in exceptional cases to the Secre-
tary of Justice, not to practical implementers (Art. 15.2 RCD). Concerning health care, the transposition decree referred
to the health insurance scheme, which specifies health coverage in detail (Art. 15.1 RCD).
The Dutch case is puzzling in light of the case‐specific theoretical expectations. Despite the absence of marked pol-
icy conflict and moderate transaction costs, transposition measures were rather constraining. Features rather character-
istic of the transposition context can, again, explain this surprising finding. Similar to the German case, pre‐existing
preferences of the agenda‐setter to change as little as possible in the domestic laws dominated the Dutch transposition
process (Resp. 14, 19, Odysseus_Report_Nl 2006). Thus, the directive did not modify preferences of the Dutch trans-
position actors to stay in close control of the central reception agency. Introducing more discretionary provisions into
Dutch legislation would have raised new transaction costs by opening a national debate on reception beyond the direc-
tive. The legislator had no incentives to introduce new discretion into already detailed reception laws. Similarly to the
German case, the Dutch case shows that the implicit assumption of domestic delegation theories that leading ministries
prefer a change in the status quo does not always hold in the transposition context.
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 11
So far, the effect of the status quo on transposition has mainly received attention from Europeanization scholars,
who have posited the goodness of fit argument (see Duina 1997; Knill & Lenschow 1998). The goodness of fit argument
holds that successful compliance depends on the fit between European policy requirements and existing institutions at
the national level. The argument received mixed support in explaining compliance, as it fails to account for domestic
preferences (Haverland 2000; Mastenbroek & Kaeding 2006; Falkner et al. 2007). Yet even though domestic actors will
not always want to maintain the status quo, they may want to do so in specific cases –a situation that breaks with
delegation theories. Thus, the German and Dutch cases show that domestic delegation theories need to account for
the possibility that domestic actors involved in transposition want to maintain the status quo. Table 2 summarizes
the findings of the three case studies.
7. Conclusion
This paper set out to investigate to what extent transposition actors delegate discretion contained in EU directives to
practical implementers and to what extent domestic delegation theories account for the variation. The article formu-
lated two theoretical expectations based on the well‐established domestic delegation literature (Epstein & O’Halloran
1999; Huber & Shipan 2002). First, policy conflict between principal and agent is necessary to make transposition actors
willing to constrain discretion. Second, discretion is only constrained under low transaction costs.
The expectations were investigated using the transposition laws of key provisions of the 2003 Asylum Reception
Condition Directive in France, Germany, and the Netherlands. We analyzed discretion after transposition systemati-
cally by using an innovative approach tomeasure discretion in legal documents, the so‐called IGT (Crawford & Ostrom
1995; Siddiki et al. 2012). Three main conclusions derive from this study.
First, the analysis showed that all three countries delegated some of the directive’s discretion to practical
implementers, or even extended discretion beyond the margins set by the directive. Thus, even if states formally comply
with EU law, they do not always settle for point‐specific positions within the margins of discretion, as implicitly
assumed by the transposition literature. Additionally, contrary to assumptions of some EU delegation studies, member
states do not always use EU discretion to tailor their national law to stay in control of practical implementation
(Franchino 2007). Instead, states differed in the level of customization of EU law (Thomann 2015). Such passing of
discretion to practical implementers forces implementers on the ground to determine the eventual outcomes of EU
law. This highlights the importance of also studying the activities of national administrators beyond legal compliance
when applying EU law in practice. Connected to this aspect, the considerable level of discretion left after transposition
to practical implementers raises normative questions of accountability and legal certainty.
The second main finding of this analysis is that in the transposition context, theories developed to explain the
delegation of discretion in national statutes could only partly explain the variance of discretion left to practical
implementers. The French case fits the theories best. Under moderate policy conflict between decisionmakers
and practical implementers, coupled with the absence of veto players with divergent preferences and thus, low
transaction costs, French transposition laws were relatively constrained. However, the German and Dutch cases
show that the particularities of the transposition context need to be taken into account as well. While there
was moderate policy conflict between transposition actors and practical implementers in the German case, trans-
position actors had unified preferences to maintain the legal status quo of asylum reception. The Dutch case
showed this mechanism even more explicitly by challenging the necessity of established conditions. Dutch
Table 2 Findings of the three transposition cases
Country Risk of deviation: Policy
conflict between practical
implementers and leading
ministry
Transaction costs:
Conflict between
transposition actors
Discretion Particularity of
transposition
process
France Moderate Low Low (with some exceptions)
Germany Moderate Low High Unified preferences
for status quoNetherlands Low Moderate Low
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
12 © 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
80

the political responsibility of the Ministry of Security and Justice. During both negotiation and transposition of the
RCD, the Ministry consulted COA extensively, in order to find out if the COA already complied with the directive
and what had to be changed (Resp. 14, 19). Respondents from the Ministry and the COA recalled no differences in
opinion regarding the directive. The only problematic issue was that more categories of asylum seekers had to be guar-
anteed reception under the RCD than under existing Dutch law, which created some capacity problems for the COA
(Resp. 14). Because of the limited policy conflict, the risk of agency drift was low for the Ministry, suggesting that the
desire to constrain discretion was also low.
Secondly, transaction costs were moderate. Although the Netherlands transposed the directive with a ministerial de-
cree (Rva 2005) that did notrequire parliamentary approval, parliament was consulted, as wellas the Dutch Council for
Refugees. This central NGO became an active player during transposition (Odysseus_Report_Nl 2006). This informal
player successfully started an initiative that triggered a debate among formal veto players in parliament. The debate con-
cerned the financial allowances for asylum seekers (Parliamentary proceeding 47 2005; Resp. 17). In response to this
debate, the opposition successfully carried a motion to modify the transposition decree in such a way that financial
allowances gradually increase. The Minister agreed to adopt this motion but stressed that it was not necessary to
comply with the RCD (DVB 2005). Additionally, a parliamentary debate on reception conditions for Dublin cases
emerged because of the directive, eventually leading to the inclusion of these cases under the reception scheme
(Odysseus_Report_Nl 2006). Thus, although the Dutch transposition measure was a decree, some veto players
with opposing views were involved in the transposition process. It can be argued that the pressure to find a
compromise with these actors increased the transaction costs of constraining legislation.
Following theories on domestic delegation, we expect that lowpolicy conflict and moderate transaction costs would
trigger the Ministry to adopt discretionary transposition laws. However, analysis of Dutch transposition reveals that,
contrary to the expectations, most discretion in the directive was constrained. As illustrated in Table A1 in 25 compo-
nents, discretion was limited, and 18 components were framed similarly to the directive. Still, in six components,
discretion extended beyond the directive. Contrary to the other two cases, all obligatory provisions were transposed.
Similarly to the other two cases, the deontic components remained mostly similar to those in the directive. The
conditions and aims mostly constrained discretion.
The conditions regarding access to the labor market were more specific in aims and conditions than in the RCD.
Dutch transposition measures detailed the period for which asylum seekers may work per year, and the documents
and administrative steps to be taken by applicants and employers (condition of Art 13.2 RCD).
Coming to material conditions, Art 13.1 RCD contained no condition component, but Dutch law introduced a
constraining condition specifying that benefits need to be given “in any case.”Weekly allowances for food, clothes,
and other expenses need to be in cash and are listed in detail, including public transport allowances. Although the
COA could decide whether the asylum seekers’own resources are checked, the conditions under which this was
possible constrained the agency’s discretion. In contrast to the other two cases, access to NGOs was regulated in more
detail than in the directive (Art 14.5 RCD). The main issue not regulated by fixed guidelines was the distribution of
asylum seekers across the centers. Contrary to the directive, which holds that transfer from one center to another
should only occur “if necessary,”there was no such condition in Dutch law (Art 14.4 RCD). However, the other
two countries did not include this provision at all in their legislation.
Similarly to France, an emergency system existed but it only delegated discretion in exceptional cases to the Secre-
tary of Justice, not to practical implementers (Art. 15.2 RCD). Concerning health care, the transposition decree referred
to the health insurance scheme, which specifies health coverage in detail (Art. 15.1 RCD).
The Dutch case is puzzling in light of the case‐specific theoretical expectations. Despite the absence of marked pol-
icy conflict and moderate transaction costs, transposition measures were rather constraining. Features rather character-
istic of the transposition context can, again, explain this surprising finding. Similar to the German case, pre‐existing
preferences of the agenda‐setter to change as little as possible in the domestic laws dominated the Dutch transposition
process (Resp. 14, 19, Odysseus_Report_Nl 2006). Thus, the directive did not modify preferences of the Dutch trans-
position actors to stay in close control of the central reception agency. Introducing more discretionary provisions into
Dutch legislation would have raised new transaction costs by opening a national debate on reception beyond the direc-
tive. The legislator had no incentives to introduce new discretion into already detailed reception laws. Similarly to the
German case, the Dutch case shows that the implicit assumption of domestic delegation theories that leading ministries
prefer a change in the status quo does not always hold in the transposition context.
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 11
So far, the effect of the status quo on transposition has mainly received attention from Europeanization scholars,
who have posited the goodness of fit argument (see Duina 1997; Knill & Lenschow 1998). The goodness of fit argument
holds that successful compliance depends on the fit between European policy requirements and existing institutions at
the national level. The argument received mixed support in explaining compliance, as it fails to account for domestic
preferences (Haverland 2000; Mastenbroek & Kaeding 2006; Falkner et al. 2007). Yet even though domestic actors will
not always want to maintain the status quo, they may want to do so in specific cases –a situation that breaks with
delegation theories. Thus, the German and Dutch cases show that domestic delegation theories need to account for
the possibility that domestic actors involved in transposition want to maintain the status quo. Table 2 summarizes
the findings of the three case studies.
7. Conclusion
This paper set out to investigate to what extent transposition actors delegate discretion contained in EU directives to
practical implementers and to what extent domestic delegation theories account for the variation. The article formu-
lated two theoretical expectations based on the well‐established domestic delegation literature (Epstein & O’Halloran
1999; Huber & Shipan 2002). First, policy conflict between principal and agent is necessary to make transposition actors
willing to constrain discretion. Second, discretion is only constrained under low transaction costs.
The expectations were investigated using the transposition laws of key provisions of the 2003 Asylum Reception
Condition Directive in France, Germany, and the Netherlands. We analyzed discretion after transposition systemati-
cally by using an innovative approach tomeasure discretion in legal documents, the so‐called IGT (Crawford & Ostrom
1995; Siddiki et al. 2012). Three main conclusions derive from this study.
First, the analysis showed that all three countries delegated some of the directive’s discretion to practical
implementers, or even extended discretion beyond the margins set by the directive. Thus, even if states formally comply
with EU law, they do not always settle for point‐specific positions within the margins of discretion, as implicitly
assumed by the transposition literature. Additionally, contrary to assumptions of some EU delegation studies, member
states do not always use EU discretion to tailor their national law to stay in control of practical implementation
(Franchino 2007). Instead, states differed in the level of customization of EU law (Thomann 2015). Such passing of
discretion to practical implementers forces implementers on the ground to determine the eventual outcomes of EU
law. This highlights the importance of also studying the activities of national administrators beyond legal compliance
when applying EU law in practice. Connected to this aspect, the considerable level of discretion left after transposition
to practical implementers raises normative questions of accountability and legal certainty.
The second main finding of this analysis is that in the transposition context, theories developed to explain the
delegation of discretion in national statutes could only partly explain the variance of discretion left to practical
implementers. The French case fits the theories best. Under moderate policy conflict between decisionmakers
and practical implementers, coupled with the absence of veto players with divergent preferences and thus, low
transaction costs, French transposition laws were relatively constrained. However, the German and Dutch cases
show that the particularities of the transposition context need to be taken into account as well. While there
was moderate policy conflict between transposition actors and practical implementers in the German case, trans-
position actors had unified preferences to maintain the legal status quo of asylum reception. The Dutch case
showed this mechanism even more explicitly by challenging the necessity of established conditions. Dutch
Table 2 Findings of the three transposition cases
Country Risk of deviation: Policy
conflict between practical
implementers and leading
ministry
Transaction costs:
Conflict between
transposition actors
Discretion Particularity of
transposition
process
France Moderate Low Low (with some exceptions)
Germany Moderate Low High Unified preferences
for status quoNetherlands Low Moderate Low
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
12 © 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 81

transposition actors maintained constraining laws, even though policy conflict between implementers and
decisionmakers was absent and transaction costs were moderate.
The phenomenon that in contrast to domestic lawmaking, transposition does not necessarily include actors who
advocate changes in the legal status quo, can explain this surprising finding. With this inductive finding, we add to
the domestic delegation and the EU implementation literature. With regard to the domestic delegation literature, we
add that this literature’s implicit assumption that domestic actors want to change the status quo may not hold in the
transposition context. Domestic actors confronted with EU legislation may want to maintain the domestic legal status
quo. As a result, in the transposition context, policy conflict and transaction costs are only necessary conditions for
constraining transposition if the costs of modifying pre‐existing levels of discretion and transposition actors’prefer-
ences for the legal status quoare incorporated. This aspect may apply more broadly to domestic lawmaking. So far, del-
egation theories have focused on major legal revisions where agenda‐setters can be assumed to be proponents of change
(Epstein & O’Halloran 1999; Huber & Shipan 2002). However, federal systems may force lower‐level agenda‐setters to
introduce legislation, just like in the EU setting. In addition, the eagerness of agenda‐setters to invest in minor legal
revisions, which constitute a large part of domestic lawmaking, may not always be pronounced.
Moreover, our study contributes to the EU implementation and compliance literature. Thisliterature devoted some
attention to the status quo when relying on the goodness‐of‐fit to explain timely and correct transposition (Héritier
1996; Duina 1997). However, the assumption that member states always aim to keep the status quo received consider-
able criticism (Haverland 2000; Mastenbroek & Kaeding 2006; Falkner et al. 2007). Our theoretical model goes beyond
the original goodness‐of‐fit literature by allowing for both conservative and reformist preferences.
Finally, from a methodological point of view, this study showed that the coding of discretion with the help of the
IGT provided much richer information on the substance of discretion than existing quantitative measures. The tool
revealed that the deontic, which is the most widely used aspect to evaluate level of discretion in quantitative studies,
is not the only relevant component that determines the level of discretion in legislation. This study showed that trans-
position actors might also pass the buck to implementers by maintaining or extending discretion along the aim and
condition components. This can have considerable implications for the outcomes of implementation and needs further
attention when studying discretion.
Having said this, reflections about external validity are in place: we tested the delegation theories in the specific pol-
icy area of asylum. Accordingly, the findings of this study might be limited to the presence of several scope conditions.
First of all, through our case selection we controlled for policy uncertainty and non‐statutory mechanisms. Because of
the high costs of asylum reception and the political sensitivity of the topic, the costs of modifying pre‐existing levels of
discretion were relatively high in all three cases, which strengthened preferences regarding the status quo. While this
mechanism might be typical for many transposition processes, there is in principle also the possibility that national
legislators are in favor of changing national laws because of EU directives (Mastenbroek & Kaeding 2006). In these
instances, transposition will resemble classical delegation theories more closely.
Consequently, a next step would be to apply the IGT in a quantitative way to contribute to our understanding of
how states decide to delegate EU discretion in the transposition phase for a wider population of directives, member
states, and over time. Such follow‐up studies would allow for more fine‐grained understanding of the origins of the
EU’s regulatory patchwork (Héritier 1996), by testing if the peculiarities of the transposition context found inductively
in this study apply more widely.
Acknowledgements
The authors would like to thank Dr. Asya Zhelyazkova and Professor Jale Tosun for comments on a previous version of
this paper. A special thanks extends to all respondents for their time.
Notes
1 They considered data available to be too limited for quantitative analysis.
2 It must be noted that we do not go as far as the original goodness of fit argument, which holds that domestic actors involved in
EU‐induced legislative processes are necessarily inclined to maintain the policy status quo. Instead, we take into account the
possibility that they want to do so, a situation that is not recognized in the domestic delegation literature (Mastenbroek &
Kaeding 2006).
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 13
3 We took 2007 as the cut‐off point for the end of transposition, which was one year before the EU Commission initiated a recast of
the RCD based on the transposition results.
4 Two sub‐articles overlapped substantively and were combined.
5 For example, national law often has no explicit deontic and gives rights directly to asylum seekers. We deal with this complication
by assuming that phrases imply “shall”deontics for practical implementers. If the law explicitly mentioned that authorities have a
choice or that these rights “can”be granted, only then did we treat this as a “may”deontic.
6 Article 14.2b RCD constitutes an exception. This literal transposition provides for a higher likelihood of the Netherlands to keep
discretion in the national regulation.
7 This made the application of the IGT to the French caseparticularly challenging as often a range of diverse national provisions in
combination transposed individual directive provisions and had to be coded in combination.
8 Conditions were different for asylum seekers in the accelerated producer and Dublin cases without access to the centers.
9 During negotiations, a coalition of SPD and Green party governed Germany.
10 The more recent renegotiation of German reception law in response to a changed asylum situation shows the high transaction
costs of reopening lawmaking in this field.
References
Aida Asylum Information Database (2013) National Country Report France. Available from URL: http://www.asylumineurope.org/
reports (accessed April 2016).
Basurto X, Kingsley G, McQueen K, Smith M, Weible CM (2010) A Systematic Approach to Institutional Analysis: Applying
Crawford and Ostom’s Grammar. Political Research Quarterly 63, 523–537.
Bendor J, Meirowitz A (2004) Spatial Models of Delegation. American Political Science Review 98, 293–310.
Blom‐Hansen J (2005) Principals, Agents, and the Implementation of EU Cohesion Policy. Journal of European Public Policy 12,
624–648.
Calvert RL, McCubbins MD, Weingast BR (1989) A Theory of Political Control and Agency Discretion. American Journal of Political
Science 33, 588–611.
Council doc 5791/02 (ASILE 5) Proposal for a Council Directive laying down minimum standards for the reception of applicantsfor
asylum in Member States‐Access to work and vocational training for applicants for asylum, Strategic Committee on Immigra-
tion, Frontiers and Asylum, 28 and 29 January 2002.
Crawford SES, Ostrom E (1995) A Grammar of Institutions. American Political Science Review 89, 582–600.
Dimitrova A (2010) The New Member States of the EU in the Aftermath of Enlargement: Do New European Rules Remain Empty
Shells? Journal of European Public Policy 17, 137–148.
Dimitrova A, Steunenberg B (2000) The Search for Convergence of National Policies in the European Union: An Impossible Quest?
European Union Politics 1, 201–226.
Dimitrova A, Steunenberg B (2013) Living in Parallel Universes? Implementing European Movable Cultural Heritage Policy in
Bulgaria. Journal of Common Market Studies 51, 246–263.
Drucksache_16/5654, 13. 06. 2007, Bericht des Innenausschusses zu dem Gesetzentwurf der Bundesregierung Entwurf eines Gesetzes
zur Umsetzung aufenthalts‐und asylrechtlicher Richtlinien der Europäischen Union.
Drucksache_16/5527, 30. 05. 2007, Unterrichtung durch die Bundesregierung Entwurf eines Gesetzes zur Umsetzung aufenthalts‐
und asylrechtlicher Richtlinien der Europäischen Union.
Drucksache 16/5621, 13. 06. 2007, Beschlussempfehlung des Innenausschusses des Bundestages zu dem Gesetzentwurf der
Bundesregierung, Entwurf eines Gesetzes zur Umsetzung aufenthalts‐und asylrechtlicher Richtlinien der Europäischen Union.
Duina F (1997) Explaining Legal Implementation in the European Union. International Journal of the Sociology of Law 25, 155–180.
DVB (2005) 5371296/05/ Ministerie van Justitie Directoraat‐Generaal Internationale Aangelegenheden en Vreemdelingenzaken
Directie Vreemdelingenbeleid, Letter from the Minister to the Parliament, Kamerstuk kst‐27062‐43‐b1.
Ellermann A (2006) Street‐level Democracy: How Immigration Bureaucrats Manage Public Opposition. West European Politics 29,
293–309.
Epstein D, O’Halloran S (1994) Administrative Procedures, Information, and Agency Discretion. American Journal of Political Science
38, 697–722.
Epstein D, O’Halloran S (1999) Delegating Powers: A Transaction Cost Politics Approach to Policy Making under Separate Powers. CUP,
Cambridge, MA.
Eurostat (2007) Eurostat Products Statistics in Focus Asylum Applications in the European Union, 110/2007.
Falkner G, Hartlapp M, Treib O (2007) Worlds of Compliance: Why Leading Approaches to the Implementation of EU Legislation
are Only ’Sometimes‐True Theories. European Journal of Political Research 46, 395–416.
Franchino F (2007) The Powers of the Union: Delegation in the EU. CUP, Cambridge, UK.
George AL, Bennett A (2005) Case Studies and Theory Development in the Social Sciences, 1st edition. MIT Press, Cambridge, MA.
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
14 © 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
82

transposition actors maintained constraining laws, even though policy conflict between implementers and
decisionmakers was absent and transaction costs were moderate.
The phenomenon that in contrast to domestic lawmaking, transposition does not necessarily include actors who
advocate changes in the legal status quo, can explain this surprising finding. With this inductive finding, we add to
the domestic delegation and the EU implementation literature. With regard to the domestic delegation literature, we
add that this literature’s implicit assumption that domestic actors want to change the status quo may not hold in the
transposition context. Domestic actors confronted with EU legislation may want to maintain the domestic legal status
quo. As a result, in the transposition context, policy conflict and transaction costs are only necessary conditions for
constraining transposition if the costs of modifying pre‐existing levels of discretion and transposition actors’prefer-
ences for the legal status quoare incorporated. This aspect may apply more broadly to domestic lawmaking. So far, del-
egation theories have focused on major legal revisions where agenda‐setters can be assumed to be proponents of change
(Epstein & O’Halloran 1999; Huber & Shipan 2002). However, federal systems may force lower‐level agenda‐setters to
introduce legislation, just like in the EU setting. In addition, the eagerness of agenda‐setters to invest in minor legal
revisions, which constitute a large part of domestic lawmaking, may not always be pronounced.
Moreover, our study contributes to the EU implementation and compliance literature. Thisliterature devoted some
attention to the status quo when relying on the goodness‐of‐fit to explain timely and correct transposition (Héritier
1996; Duina 1997). However, the assumption that member states always aim to keep the status quo received consider-
able criticism (Haverland 2000; Mastenbroek & Kaeding 2006; Falkner et al. 2007). Our theoretical model goes beyond
the original goodness‐of‐fit literature by allowing for both conservative and reformist preferences.
Finally, from a methodological point of view, this study showed that the coding of discretion with the help of the
IGT provided much richer information on the substance of discretion than existing quantitative measures. The tool
revealed that the deontic, which is the most widely used aspect to evaluate level of discretion in quantitative studies,
is not the only relevant component that determines the level of discretion in legislation. This study showed that trans-
position actors might also pass the buck to implementers by maintaining or extending discretion along the aim and
condition components. This can have considerable implications for the outcomes of implementation and needs further
attention when studying discretion.
Having said this, reflections about external validity are in place: we tested the delegation theories in the specific pol-
icy area of asylum. Accordingly, the findings of this study might be limited to the presence of several scope conditions.
First of all, through our case selection we controlled for policy uncertainty and non‐statutory mechanisms. Because of
the high costs of asylum reception and the political sensitivity of the topic, the costs of modifying pre‐existing levels of
discretion were relatively high in all three cases, which strengthened preferences regarding the status quo. While this
mechanism might be typical for many transposition processes, there is in principle also the possibility that national
legislators are in favor of changing national laws because of EU directives (Mastenbroek & Kaeding 2006). In these
instances, transposition will resemble classical delegation theories more closely.
Consequently, a next step would be to apply the IGT in a quantitative way to contribute to our understanding of
how states decide to delegate EU discretion in the transposition phase for a wider population of directives, member
states, and over time. Such follow‐up studies would allow for more fine‐grained understanding of the origins of the
EU’s regulatory patchwork (Héritier 1996), by testing if the peculiarities of the transposition context found inductively
in this study apply more widely.
Acknowledgements
The authors would like to thank Dr. Asya Zhelyazkova and Professor Jale Tosun for comments on a previous version of
this paper. A special thanks extends to all respondents for their time.
Notes
1 They considered data available to be too limited for quantitative analysis.
2 It must be noted that we do not go as far as the original goodness of fit argument, which holds that domestic actors involved in
EU‐induced legislative processes are necessarily inclined to maintain the policy status quo. Instead, we take into account the
possibility that they want to do so, a situation that is not recognized in the domestic delegation literature (Mastenbroek &
Kaeding 2006).
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 13
3 We took 2007 as the cut‐off point for the end of transposition, which was one year before the EU Commission initiated a recast of
the RCD based on the transposition results.
4 Two sub‐articles overlapped substantively and were combined.
5 For example, national law often has no explicit deontic and gives rights directly to asylum seekers. We deal with this complication
by assuming that phrases imply “shall”deontics for practical implementers. If the law explicitly mentioned that authorities have a
choice or that these rights “can”be granted, only then did we treat this as a “may”deontic.
6 Article 14.2b RCD constitutes an exception. This literal transposition provides for a higher likelihood of the Netherlands to keep
discretion in the national regulation.
7 This made the application of the IGT to the French caseparticularly challenging as often a range of diverse national provisions in
combination transposed individual directive provisions and had to be coded in combination.
8 Conditions were different for asylum seekers in the accelerated producer and Dublin cases without access to the centers.
9 During negotiations, a coalition of SPD and Green party governed Germany.
10 The more recent renegotiation of German reception law in response to a changed asylum situation shows the high transaction
costs of reopening lawmaking in this field.
References
Aida Asylum Information Database (2013) National Country Report France. Available from URL: http://www.asylumineurope.org/
reports (accessed April 2016).
Basurto X, Kingsley G, McQueen K, Smith M, Weible CM (2010) A Systematic Approach to Institutional Analysis: Applying
Crawford and Ostom’s Grammar. Political Research Quarterly 63, 523–537.
Bendor J, Meirowitz A (2004) Spatial Models of Delegation. American Political Science Review 98, 293–310.
Blom‐Hansen J (2005) Principals, Agents, and the Implementation of EU Cohesion Policy. Journal of European Public Policy 12,
624–648.
Calvert RL, McCubbins MD, Weingast BR (1989) A Theory of Political Control and Agency Discretion. American Journal of Political
Science 33, 588–611.
Council doc 5791/02 (ASILE 5) Proposal for a Council Directive laying down minimum standards for the reception of applicantsfor
asylum in Member States‐Access to work and vocational training for applicants for asylum, Strategic Committee on Immigra-
tion, Frontiers and Asylum, 28 and 29 January 2002.
Crawford SES, Ostrom E (1995) A Grammar of Institutions. American Political Science Review 89, 582–600.
Dimitrova A (2010) The New Member States of the EU in the Aftermath of Enlargement: Do New European Rules Remain Empty
Shells? Journal of European Public Policy 17, 137–148.
Dimitrova A, Steunenberg B (2000) The Search for Convergence of National Policies in the European Union: An Impossible Quest?
European Union Politics 1, 201–226.
Dimitrova A, Steunenberg B (2013) Living in Parallel Universes? Implementing European Movable Cultural Heritage Policy in
Bulgaria. Journal of Common Market Studies 51, 246–263.
Drucksache_16/5654, 13. 06. 2007, Bericht des Innenausschusses zu dem Gesetzentwurf der Bundesregierung Entwurf eines Gesetzes
zur Umsetzung aufenthalts‐und asylrechtlicher Richtlinien der Europäischen Union.
Drucksache_16/5527, 30. 05. 2007, Unterrichtung durch die Bundesregierung Entwurf eines Gesetzes zur Umsetzung aufenthalts‐
und asylrechtlicher Richtlinien der Europäischen Union.
Drucksache 16/5621, 13. 06. 2007, Beschlussempfehlung des Innenausschusses des Bundestages zu dem Gesetzentwurf der
Bundesregierung, Entwurf eines Gesetzes zur Umsetzung aufenthalts‐und asylrechtlicher Richtlinien der Europäischen Union.
Duina F (1997) Explaining Legal Implementation in the European Union. International Journal of the Sociology of Law 25, 155–180.
DVB (2005) 5371296/05/ Ministerie van Justitie Directoraat‐Generaal Internationale Aangelegenheden en Vreemdelingenzaken
Directie Vreemdelingenbeleid, Letter from the Minister to the Parliament, Kamerstuk kst‐27062‐43‐b1.
Ellermann A (2006) Street‐level Democracy: How Immigration Bureaucrats Manage Public Opposition. West European Politics 29,
293–309.
Epstein D, O’Halloran S (1994) Administrative Procedures, Information, and Agency Discretion. American Journal of Political Science
38, 697–722.
Epstein D, O’Halloran S (1999) Delegating Powers: A Transaction Cost Politics Approach to Policy Making under Separate Powers. CUP,
Cambridge, MA.
Eurostat (2007) Eurostat Products Statistics in Focus Asylum Applications in the European Union, 110/2007.
Falkner G, Hartlapp M, Treib O (2007) Worlds of Compliance: Why Leading Approaches to the Implementation of EU Legislation
are Only ’Sometimes‐True Theories. European Journal of Political Research 46, 395–416.
Franchino F (2007) The Powers of the Union: Delegation in the EU. CUP, Cambridge, UK.
George AL, Bennett A (2005) Case Studies and Theory Development in the Social Sciences, 1st edition. MIT Press, Cambridge, MA.
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
14 © 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 83

Goertz G, Starr H (2003) Necessary Conditions: Theory, Methodology, and Applications. Rowman and Littlefield, Lanham, MD.
Hartmann JM (2016) A Blessing in Disguise?! Discretion in the Context of EU Decision‐making, National Transposition and Legitimacy
Regarding EU Directives. Amsterdam University Press, Amsterdam.
Haverland M (2000) National Adaptation to European Integration: The Importance of Institutional Veto Points. Journal of Public
Policy 20, 83–103.
Héritier A (1996) The Accommodation of Diversity in European Policy‐Making and Its Outcomes: RegulatoryPolicy as a Patchwork.
Journal of European Public Policy 3, 149–167.
Huber JD, Shipan CR (2002) Deliberate Discretion? The Institutional Foundations of Bureaucratic Autonomy. CUP, Cambridge, UK.
Huber J, Shipan C (2011) Politics, Delegation, and Bureaucracy. In: Goodin RE (ed) The Oxford Handbook of Political Science,
pp. 1–13. OUP, Oxford.
Huber J, Shipan C, Pfahler M (2001) Legislatures and Statutory Control of Bureaucracy. American Journal of Political Science 45,
330–345.
Knill C, Lenschow A (1998) Coping with Europe: The Impact of British and German Administrations on the Implementation of EU
Environmental Policy. Journal of European Public Policy 5, 595–614.
Lijphart A,Crepaz MML (1991) Corporatism and Consensus Democracy in EighteenCountries: Conceptual and Empirical Linkages.
British Journal of Political Science 21, 235–246.
Mahoney J, Goertz G (2006) A Tale of two Cultures: Contrasting Quantitative and Qualitative Research. Political Analysis 14,
227–249.
Mastenbroek E, Kaeding M (2006) Europeanization Beyond the Goodness of Fit: Domestic Politics in the Forefront. Comparative
European Politics 4, 331–354.
McCubbins M, Noll R, Weingast B (1987) Administrative Procedures as Instruments of Political Control. Journal of Law, Economics,
and Organization 3, 243–277.
McCubbins M, Schwartz T (1984) Congressional Oversight Overlooked: Police Patrols versus Fire Alarms Source. American Journal
of Political Science 28, 165–179.
Moe TM (1982) Regulatory Performance and Presidential Administration. American Journal of Political Science 26, 197–224.
Moe TM (1984) The New Economics of Organization. American Journal of Political Science 28, 739–777.
Odysseus_Report_FR (2006) National report done by the Odysseus Network for the European Commission on the implementation
of the directive on Reception Conditions for Asylum Seekers in: France by Labayle H. (accessed through request).
Odysseus_Report_general (2006) ‘Comparative overview of the implementation of the Directive 2003/9 of 27 January laying down
minimum standards for the reception of asylum‐seekers in the EU member states’, [Last accessed March 2017.] Available from
URL: http://ec.europa.eu/dgs/home-affairs/elibrary/docs/pdf/odysseus_synthesis_report_2007_en_en.pdf
Odysseus_Report_Ger (2006) National report done by the Odysseus Network for the European Commission on the implementation
of the directive on Reception Conditions for Asylum Seekers in: Germany by Hailbronner K, Peek M. (accessed through request)
Odysseus_Report_NL (2006) National report done by the Odysseus Network for the European Commission on the implementation
of the Directive on Reception Conditions for Asylum Seekers in: the Netherlands by Franssen K, Groenendijk K. (accessed
through request).
Oosterwaal A, Payne D, Torenvlied R (2012) The Effect of Political Disagreement on Discretion: Mechanisms and Conditions.
Administration & Society 44, 800–824.
Parliamentary proceeding 47 (2005) Handelingen der Tweede Kamer, February 8th, vergaderjaar 2004‐2005, nr. 47, p. 3006‐3014.
Pollack MA (1997) Delegation, Agency, and Agenda Setting in the European Community. International Organization 51, 99–134.
Siddiki S, Basurto X, Weible CM (2012) Using the Institutional Grammar Tool to Understand Regulatory Compliance: The Case of
Colorado Aquaculture. Regulation & Governance 6, 167–188.
Steunenberg B (2006) Turning Swift Policy‐making into Deadlock and Delay: National Policy Coordination and the Transposition of
EU Directives. European Union Politics 7, 293–319.
Steunenberg B, Voermans WJ (2006) The Transposition of EC Directives: A Comparative Study of Instruments,Techniques and Processes
in Six EU Member States. Leiden University, Leiden/ Research and Documentation Centre (WODC) of the Ministry of Justice.
Thomann E (2015) Customizing Europe: Transposition as Bottom‐up Implementation. Journal of European Public Policy 22,
1368–1387.
Thomson R, Torenvlied R (2011) Information, Commitment and Consensus: A Comparison of Three Perspectives on Delegation in
the European Union. British Journal of Political Science 41, 139–159.
Toshkov D (2011) The Quest for Relevance: Research on Compliance with EU Law. Working paper. [Last accessed 03 March 2017]
Available from URL: http://www.dimiter.eu/articles/Compliance%20review.pdf
Toshkov D (2013) Discretion Is Fractal’. Keynote Presentation held January 25, 2013 at Leiden University, Netherlands. [Last
accessed 1 Apr 2016.] Available from URL: http://rulesofreason.files.wordpress.com/2013/02/discretion‐is‐fractal.pdf
Treib O (2014) Implementing and Complying with EU Governance Outputs. Living Reviews in European Governance 9 (1), 1–47.
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 15
Versluis E (2007) Even Rules, Uneven Practices: Opening the “Black Box”of EU Law in Action. West European Politics 30, 50–67.
Zaun N (2016) Why EU Asylum Standards Exceed the Lowest Common Denominator: The Role of Regulatory Expertise in EU
Decision‐making. Journal of European Public Policy 23 (1), 136–156.
Zhelyazkova A, Torenvlied R (2009) The Time‐dependent Effect of Conflict in the Council on Delays in the Transposition of EU
Directives. European Union Politics 10, 35–62.
Laws Cited
Council Directive 2003/9/EC of 27 January 2003 laying down minimum standards for the reception of asylum seekers
Rva (2005) Regulation on the provisions for asylum seekers and other categories of aliens (Regeling verstrekkingen asielzoekers en
andere categorieën vreemdelingen) 26 January 2005, Nr. 5332253/05/DVB Directie Vreemdelingenbeleid, published in the
Government Gazette of February 3rd 2005, no. 24, p. 17.
Appendix
Supporting information
Additional Supporting Information may be found online in the supporting information tab for this article.
Table S1: Transposition table: Wording and coding.
Table S2: Transposition table.
Table A1 Coding results
France Germany Netherlands
Topic Directive
article
Aim Deontic Condition Aim Deontic Condition Aim Deontic Condition
Access to labor
market
Art 11.1 −00−0+−0−
Art 11.2 −00−0+−0−
Art 11.4 Na 0 0 −Na
Material
Conditions
Art 13.1 −0−+0 −00 −
Art 13.2 −0 Na −0 Na −0−
Art 13.3 Na Na −− −
Art 13.4 −− − −(0/−)−−0/−−
Art 13.5 −− − +0 0 −− −
Accommodation Art 14.1 0 0 0 + 0 + −0−
Art 14.2
a)
+0++0+00+
Art 14.3 −0−0 (+/0) −−00
Art 14.4 Na Na 0 + +
Art 14.5 Na Na −0−
Art 14.7 + 0 −−+ 0 00 −
Art 14.2
b)
Art 14.8 + −0 Na + 0 +
Healthcare Art 15.1 −0−−0+−0−
Art 15.2 −0 + 0+ −+0 −
Art 17.1
Na, not applicable.
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
16 © 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
84

Goertz G, Starr H (2003) Necessary Conditions: Theory, Methodology, and Applications. Rowman and Littlefield, Lanham, MD.
Hartmann JM (2016) A Blessing in Disguise?! Discretion in the Context of EU Decision‐making, National Transposition and Legitimacy
Regarding EU Directives. Amsterdam University Press, Amsterdam.
Haverland M (2000) National Adaptation to European Integration: The Importance of Institutional Veto Points. Journal of Public
Policy 20, 83–103.
Héritier A (1996) The Accommodation of Diversity in European Policy‐Making and Its Outcomes: RegulatoryPolicy as a Patchwork.
Journal of European Public Policy 3, 149–167.
Huber JD, Shipan CR (2002) Deliberate Discretion? The Institutional Foundations of Bureaucratic Autonomy. CUP, Cambridge, UK.
Huber J, Shipan C (2011) Politics, Delegation, and Bureaucracy. In: Goodin RE (ed) The Oxford Handbook of Political Science,
pp. 1–13. OUP, Oxford.
Huber J, Shipan C, Pfahler M (2001) Legislatures and Statutory Control of Bureaucracy. American Journal of Political Science 45,
330–345.
Knill C, Lenschow A (1998) Coping with Europe: The Impact of British and German Administrations on the Implementation of EU
Environmental Policy. Journal of European Public Policy 5, 595–614.
Lijphart A,Crepaz MML (1991) Corporatism and Consensus Democracy in EighteenCountries: Conceptual and Empirical Linkages.
British Journal of Political Science 21, 235–246.
Mahoney J, Goertz G (2006) A Tale of two Cultures: Contrasting Quantitative and Qualitative Research. Political Analysis 14,
227–249.
Mastenbroek E, Kaeding M (2006) Europeanization Beyond the Goodness of Fit: Domestic Politics in the Forefront. Comparative
European Politics 4, 331–354.
McCubbins M, Noll R, Weingast B (1987) Administrative Procedures as Instruments of Political Control. Journal of Law, Economics,
and Organization 3, 243–277.
McCubbins M, Schwartz T (1984) Congressional Oversight Overlooked: Police Patrols versus Fire Alarms Source. American Journal
of Political Science 28, 165–179.
Moe TM (1982) Regulatory Performance and Presidential Administration. American Journal of Political Science 26, 197–224.
Moe TM (1984) The New Economics of Organization. American Journal of Political Science 28, 739–777.
Odysseus_Report_FR (2006) National report done by the Odysseus Network for the European Commission on the implementation
of the directive on Reception Conditions for Asylum Seekers in: France by Labayle H. (accessed through request).
Odysseus_Report_general (2006) ‘Comparative overview of the implementation of the Directive 2003/9 of 27 January laying down
minimum standards for the reception of asylum‐seekers in the EU member states’, [Last accessed March 2017.] Available from
URL: http://ec.europa.eu/dgs/home-affairs/elibrary/docs/pdf/odysseus_synthesis_report_2007_en_en.pdf
Odysseus_Report_Ger (2006) National report done by the Odysseus Network for the European Commission on the implementation
of the directive on Reception Conditions for Asylum Seekers in: Germany by Hailbronner K, Peek M. (accessed through request)
Odysseus_Report_NL (2006) National report done by the Odysseus Network for the European Commission on the implementation
of the Directive on Reception Conditions for Asylum Seekers in: the Netherlands by Franssen K, Groenendijk K. (accessed
through request).
Oosterwaal A, Payne D, Torenvlied R (2012) The Effect of Political Disagreement on Discretion: Mechanisms and Conditions.
Administration & Society 44, 800–824.
Parliamentary proceeding 47 (2005) Handelingen der Tweede Kamer, February 8th, vergaderjaar 2004‐2005, nr. 47, p. 3006‐3014.
Pollack MA (1997) Delegation, Agency, and Agenda Setting in the European Community. International Organization 51, 99–134.
Siddiki S, Basurto X, Weible CM (2012) Using the Institutional Grammar Tool to Understand Regulatory Compliance: The Case of
Colorado Aquaculture. Regulation & Governance 6, 167–188.
Steunenberg B (2006) Turning Swift Policy‐making into Deadlock and Delay: National Policy Coordination and the Transposition of
EU Directives. European Union Politics 7, 293–319.
Steunenberg B, Voermans WJ (2006) The Transposition of EC Directives: A Comparative Study of Instruments,Techniques and Processes
in Six EU Member States. Leiden University, Leiden/ Research and Documentation Centre (WODC) of the Ministry of Justice.
Thomann E (2015) Customizing Europe: Transposition as Bottom‐up Implementation. Journal of European Public Policy 22,
1368–1387.
Thomson R, Torenvlied R (2011) Information, Commitment and Consensus: A Comparison of Three Perspectives on Delegation in
the European Union. British Journal of Political Science 41, 139–159.
Toshkov D (2011) The Quest for Relevance: Research on Compliance with EU Law. Working paper. [Last accessed 03 March 2017]
Available from URL: http://www.dimiter.eu/articles/Compliance%20review.pdf
Toshkov D (2013) Discretion Is Fractal’. Keynote Presentation held January 25, 2013 at Leiden University, Netherlands. [Last
accessed 1 Apr 2016.] Available from URL: http://rulesofreason.files.wordpress.com/2013/02/discretion‐is‐fractal.pdf
Treib O (2014) Implementing and Complying with EU Governance Outputs. Living Reviews in European Governance 9 (1), 1–47.
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 15
Versluis E (2007) Even Rules, Uneven Practices: Opening the “Black Box”of EU Law in Action. West European Politics 30, 50–67.
Zaun N (2016) Why EU Asylum Standards Exceed the Lowest Common Denominator: The Role of Regulatory Expertise in EU
Decision‐making. Journal of European Public Policy 23 (1), 136–156.
Zhelyazkova A, Torenvlied R (2009) The Time‐dependent Effect of Conflict in the Council on Delays in the Transposition of EU
Directives. European Union Politics 10, 35–62.
Laws Cited
Council Directive 2003/9/EC of 27 January 2003 laying down minimum standards for the reception of asylum seekers
Rva (2005) Regulation on the provisions for asylum seekers and other categories of aliens (Regeling verstrekkingen asielzoekers en
andere categorieën vreemdelingen) 26 January 2005, Nr. 5332253/05/DVB Directie Vreemdelingenbeleid, published in the
Government Gazette of February 3rd 2005, no. 24, p. 17.
Appendix
Supporting information
Additional Supporting Information may be found online in the supporting information tab for this article.
Table S1: Transposition table: Wording and coding.
Table S2: Transposition table.
Table A1 Coding results
France Germany Netherlands
Topic Directive
article
Aim Deontic Condition Aim Deontic Condition Aim Deontic Condition
Access to labor
market
Art 11.1 −00−0+−0−
Art 11.2 −00−0+−0−
Art 11.4 Na 0 0 −Na
Material
Conditions
Art 13.1 −0−+0 −00 −
Art 13.2 −0 Na −0 Na −0−
Art 13.3 Na Na −− −
Art 13.4 −− − −(0/−)−−0/−−
Art 13.5 −− − +0 0 −− −
Accommodation Art 14.1 0 0 0 + 0 + −0−
Art 14.2
a)
+0++0+00+
Art 14.3 −0−0 (+/0) −−00
Art 14.4 Na Na 0 + +
Art 14.5 Na Na −0−
Art 14.7 + 0 −−+ 0 00 −
Art 14.2
b)
Art 14.8 + −0 Na + 0 +
Healthcare Art 15.1 −0−−0+−0−
Art 15.2 −0 + 0+ −+0 −
Art 17.1
Na, not applicable.
N. Dörrenbächer and E. Mastenbroek Discretion after transposition
16 © 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd.
Discretion after transposition N. Dörrenbächer and E. Mastenbroek
© 2017 The Authors. Regulation & Governance published by John Wiley & Sons Australia, Ltd. 85