Two Variations on a Theme: Different Logics of Implementation Management in the EU and the ILO(*)
Miriam Hartlapp
European Integration online Papers (EIoP) Vol. 9 (2005) N° 7;
http://eiop.or.at/eiop/texte/2005-007a.htm
Date of Publication in EIoP: 14.6.2005
| Abstract | Back to homepage | PDF |
| This paper's comments page | Send your comment! to this paper |

Contents:


1

1. Introductionto contents list

In recent years compliance with international rules and agreements has gained importance in society as well as in research. It is widely assumed that growing international interdependencies are leading to an increase in international institutional arrangements such as the International Court of Justice and the continued transfer of competence to Brussels. The theoretical-conceptual framework of ‘legalization’ (Goldstein et al. 2000) attempts to capture these developments by assessing the degree of institutionalization of international relations along three dimensions: obligation, thus binding states by rules or commitments; the precision of these rules; and the delegation of implementation authority to third parties. In the words of Abbott and others (2000, 401), ‘legalization’ is intended ‘to show how law and politics are intertwined’ in different institutional arrangements, defined as rules and procedures, and to look at its consequences for compliance with international rules (for further debate and development of the concept see e.g. Keohane, Moravscik and Slaughter 2000; Finnemore and Toope 2001; Goldstein et al. 2001; Alter 2003; Zangl and Zürn 2004b; Zürn 2005)

 
 
 
 

2

From this perspective the European Union (EU)(1) is characterized as highly legalized, scoring high in obligation, precision and delegation. The International Labour Organization (ILO), as a specialized United Nations agency, is considered to have low obligation, low precision and (moderately) high delegation (Abbott et al. 2000, 406, Table 1) .(2) The observation that in both cases Member States have delegated implementation powers to a comparably high degree to an international organization (IO) is taken as a starting point for this article. Taking into account that ‘the delegation of authority includes such different tasks as fact-finding, dispute settlement, and rule development in the process of rule application’ (Zürn 2005, 23-24), I have chosen the delegation of implementation powers as the subject of further analysis.(3) The actual patterns of compliance management need to be explained in order to understand inter-category differences with respect to the use that IOs make of the powers delegated to them when addressing non-implementation or incorrect implementation of commonly agreed standards.(4) I therefore take ‘legalization’ as a starting point for an empirical analysis that borrows from other explanatory concepts in order to build a bridge between compliance and legalization.

What are the benefits of such an analysis? First, the assessment of similarities and differences between the EU and the ILO is an attempt to make comparable specific aspects of the two most important actors in social policy beyond the nation-states. Currently, despite apparent similarities between the EU and the ILO in the degree of delegated authority, many authors studying compliance within the EU realm (e.g. Mbaye 2001; Mendrinou 1996) implicitly or explicitly understand the EU as a political system sui generis, thus preventing a broader comparison (for a critique see e.g. Hix 1998, 54–55). Knowledge about the differences and commonalities of the two IOs is an important contribution to a better understanding of the (partial) success or failure of policy implementation. Second, the analysis will advance our understanding of delegation as a specific aspect of ‘legalization’ through empirically driven differentiation. But why should we know more about delegation and thus about legalization at all? The answer is a very practical one. Much hope has been placed on international or transnational legalization as a central pillar of global governance. There are more and more international rules; if they are to be effective, we need to know more about mechanisms that can ensure that Member States comply with these rules. And legalization is considered to be beneficial in improving compliance with international rules and agreements (on this line, see Zangl and Zürn 2004a, 20; for a more critical view, see Mayer 2004).

To this end the article first presents a theoretical framework offered by international relations research on the question of how to make states comply. Three different logics in particular (enforcement, management and persuasion) have received scholarly attention in recent years. Against the background of these abstract categories, the article then provides an empirical assessment and categorization of EU and ILO policies to bring their respective policies into practice. This assessment includes the formal means and powers for improving the implementation of their instruments, as well as other mechanisms and strategies used to alter the behaviour of member countries. For each institution I show that it uses all three logics—albeit to differing degrees. With respect to formal sanctioning powers the EU is able to exert relatively more pressure, whereas management is more widely used in the ILO. Finally, I discuss in a comparative context the use of (hard) enforcement, (supportive) management and/or (soft) persuasion in the two IOs.

 
 
 
 

3

2. Theoretical background and operationalization to contents list

A review of the theoretical frameworks in the field of compliance with international rules shows that three main schools of thought can be distinguished: enforcement, management and persuasion.(5) The enforcement approach (Hart 1968; Olson 1965; Downs, Rocke and Barsoon 1996) focuses on pressure and sanctions when it comes to explaining the behaviour of states. It assumes that Member States choose not to comply on the basis of their own cost-benefit calculations. Hence, in order to ensure compliance, the possible losses when found in breach of a rule must be greater than the potential gains obtained from non-compliance. The level of compliance thus depends on the probability and severity of punishment. From such a perspective, IOs with substantial sanctioning power should be in a good position to bring about compliance.

In contrast, the management approach (Chayes and Handler Chayes 1993; Haas, Keohane and Levy 1994) assumes that non-compliance is above all due to financial, administrative or technical shortcomings (i.e. to a lack of resources or expertise), not to opposition to norms. Thus, the imposition of high fines does not improve compliance rates. From this perspective non-compliance is seen as a problem to be jointly solved by the IO and the state. Capacity-building through the transfer of knowledge and resources is the key to changing the behaviour of non-compliant states. In this sense, the functioning and well-equipped administrative structures of an IO should help improve compliance through management.

The enforcement and management approaches have competing assumptions about why actors do not comply, but they both exert positive (carrot) or negative (stick) instrumental influence in order to change Member State behaviour. The third approach, persuasion (Checkel 2001; Risse 2000; on reflexivity/responsiveness, see Neyer and Wolf 2005, 59–60), is distinctly different in that it aims to change the underlying norms and values that drive Member State action through a logic of appropriateness.(6)

These are analytical categories rather than adequate descriptions of reality. Although the theoretical concepts depict different understandings of the functioning of interaction, in practice the logics are not exclusive. And according to Alter (2003, 56), ‘It would be silly [for an IO] not to use all of these levers to encourage compliance.’ It must be determined empirically when and how an IO follows the three logics in order to deal with a failure to implement or incorrect implementation. What observable implications support the argument that the IOs follow one (or several) of the three logics in its implementation management?

The implementation policy (also referred to as implementation management) of an IO is the product of its institutional setting and policy instruments on the one hand and the use that actors make of the powers assigned to them on the other. Hence, the implementation policy of an IO is its action that aims at furthering ratification or transposition and application of international rules in the Member States. By looking at the instruments that an IO uses in its implementation policy, we can assess whether its main logic of changing Member State behaviour is enforcement, management or persuasion. If an instrument is created to raise the costs of non-compliance through the imposition of (financial) sanctions or the discrediting of a Member State in the arena of the IO,(7) I categorize the instrument as one following a logic of enforcement. An instrument that aims at lowering the barriers to implement a specific policy by promoting knowledge about how to solve problematic conflicts in the political system or by means of financial support to build up or reform administrative structures (concerned with either rule-making or application) is considered to follow a logic of management.

 

4

According to the persuasion concept, interaction between the decisive actors should generate compliance through a logic of appropriateness. Neyer and Zürn (2001; also see Neyer and Wolf 2005, 60), who have comparatively studied the preconditions for compliance by states or societies which goes beyond the nation-state, argue that a greater degree of political interaction—in both positive (cooperation) and negative (infringements) form—leads to a higher political, legal and societal internalization of the rules. I consider instruments that create interactions aiming at an exchange of ideas, such as country visits, formal or informal meetings, and institutionalized contacts, to be ones following a logic of persuasion.

One point needs to be stressed here. Empirically it is often difficult to make a clear-cut distinction between management and persuasion (a meeting about a technical assistance project will at the same time most likely generate understanding and adherence to the norms conveyed in the project). Analytically, however, the two approaches reflect a fundamental divide about the underlying reason for compliance. If the EU or the ILO can convince a member country that following a rule makes the whole society better off, this is a normative logic, the logic of persuasion (e.g. abolition of child labour enhances the level of education in the long run). At the same time, the provision of resources to build a school under the condition that working children are enrolled in educational activities is an instrumental logic, the logic of management.

Having presented the prevailing logic for individual instruments, I can now address two aspects that are important for the overall assessment of implementation policies in IOs: First, with respect to relative importance, what is the dominant logic within an IO? Relative importance refers to the incidence with which the three logics are used. It can be assessed through a quantitative comparison of the use the organization makes of instruments following the logic of enforcement, management or persuasion. The second question refers to the absolute strength of the three logics in the implementation instruments studied. For example, even though management might be the relatively most important logic, it can still be weak in absolute terms. In such a case the implementation power of an IO would be rather weak. To assess absolute strength, two aspects should be separated: formal and actual strength. An implementation logic is considered to be formally strong when foreseen rules and procedures allow for changing Member State behaviour (see the discussion of legalization of secondary law in Zangl and Zürn 2004a, 22). It is moderate when the formal rules only allow for compliance under specific conditions or up to a certain degree. Finally, it is considered to be weak when its instruments are not sufficient to make non-compliant countries comply.(8) In practice the strength of foreseen rules and procedures might be weakened through institutional characteristics, such as (factual) veto points in decision-making (on the concept in general see Tsebelis 1995, on factual veto points Héritier 2001, 12–13) or a lack of resources. Whenever such factors significantly impede the use of formal powers, a markdown is applied, so that the respective actual strength is then determined to be moderate or weak. For both relative importance and absolute strength, the explanations are expected to lie in the IO’s institutional features.

Table 1

 
 
 
 
 
 

5

3. Empirical material: Data and casesto contents list

The social policy area will be considered in the case of the EU. This part of the article builds on two sources: (a) the results of a multi-annual research project that analyses the national transposition, enforcement and application of six EU social policy directives(9) in all 15 Member States and (b) my dissertation written in the context of this project (Hartlapp 2005).(10) The directives concern written information on employment conditions (91/533/EEC), parental leave (96/34/EC), working time (93/104/EC), and the protection of pregnant (92/85/EEC), young (94/33/EC) and part-time workers (97/81/EC). Hence, they cover all important EU social policy directives from the 1990s.(11) Detailed knowledge about compliance and implementation policies from these 90 case studies will be combined with a broad perspective on the quantity and quality of the instruments that the European Commission uses to make Member States comply.

The analysis of ILO implementation policies is mostly based on data on member countries’ (non-)compliance, published regularly in reports and on the web page of the ILO (http://webfusion.ilo.org/public/db/standards/normes/index.cfm), as well as on expert interviews. For individual cases information on the effect of the ILO implementation policies is also provided.

On the issue of data, it should be emphasized that there is a fundamental difference between the level of compliance in a country and the manner and frequency that non-compliant behaviour is responded to through coercion. There are two reasons for this circumstance: First, there may be cases where the IO does not know about non-compliance, and, second, there may be situations in which the IO does not wish to exert pressure, even though it is aware of the infraction. Compliance research often does not account for this bias. One explanation is that it is very cumbersome to generate data about the de facto level of compliance, because it must be determined at the micro level. For the same reason we still know little about the specific effect of supervision and enforcement instruments on the level of compliance in a country. In this respect, I wish to stress that the aim of this article is not to assess the level of compliance with international rules, but to systematically and comparatively analyse the instruments that IOs use to bring their policies into practice.

4. Differing institutional featuresto contents list

Within the field of social policy, the EU and the ILO(12) are without doubt the most important international actors, and it is therefore especially interesting to compare their implementation policies. It is widely known that the EU, as a highly legalized international institution, has vast formal enforcement powers. At the same time, many scholars question whether the ILO can at all exert pressure to ensure compliance. Before the supervision and enforcement instruments of the EU and the ILO are described in detail, it is necessary to make some introductory remarks on their institutional features.

A major institutional difference between the two organizations is that Member States of the ILO are not required to ratify ILO conventions (the degree of obligation to rules is low; see Abbott et al. 2000, 406). There are big differences between the ratification records of member countries and conventions. Once a Member State has ratified, its government is responsible for the correct implementation of the respective international labour standards.(13) In the ILO there are currently 185 ILO conventions (only part of them are still actively promoted; see ILO 2000)(14) with 7,262 ratifications in 177 member countries differing tremendously in development, economic strength and political systems. The supervision and enforcement of these ratifications is carried out in a complex interplay between the three constitutive bodies: the International Labour Conference (ILC), the Governing Body (GB) and the International Labour Office.


6

The ILC is the annual meeting of delegates from all Member States and the legislative body of the ILO. The GB is the executive council of the ILO and is composed of 28 government members, 14 employer members and 14 worker members. The International Labour Office is the permanent secretariat of the ILO, with headquarters in Geneva and 42 field and liaison offices around the world. The most important arenas for supervision and enforcement are the International Labour Standards/Standards and Fundamental Principles and Rights at Work Department in the secretariat, the GB and specialized sub-bodies (see below).

The ILO is a uniquely tripartite organization in which the complicated interplay between governments, employers’ and employees’ interests, and other cleavages, such as those between developing and industrialized countries, offers multiple possibilities for the obstruction of supervision and enforcement policies. Moreover, because these political decisions are generally taken on a consensual basis, only the fairly uncontested cases make it up the enforcement ladder. The concrete differences between the formal and the actual strength of implementation policies will be analysed below.

In the EU, the implementation management of commonly agreed rules lies with the European Commission, the ‘Guardian of the Treaties’ (Article 211 ECT). The Secretariat General overviews the entire procedure, the responsible Directorate General sets out the specifics of the procedure and the College of Commissioners makes decisions (for details, see Hartlapp 2005, 184–186). The question of how to proceed with violations has to be agreed unanimously, thus national interests cannot be fully excluded. Here, too, differences in the formal and actual strength of implementation are possible. But in general the procedure is blocked only in cases of explicit opposition.(15) To fulfil its role as Guardian of the Treaties, the Commission has various formal and informal instruments at its disposal.

5. Implementation policies in the EUto contents list

5.1. Infringement procedure and other enforcement instruments

Through the infringement procedure the European Commission probably exerts more direct pressure on defecting Member States than any other IO. When a Member State does not follow commonly agreed rules, whether as a consequence of late or incorrect transposition or the non-application or incorrect application of a directive, the Commission can start an infringement procedure (Articles 226 and 228 ETC). It consists of four different steps: ‘Letter of Formal Notice, Reasoned Opinion, Referral to the European Court of Justice (ECJ) and Judgment of the ECJ’. They are all initiated by the Commission or the ECJ, but reflect an interaction of the supranational and the national level of governance. In cases of remaining opposition to the ECJ judgment, the procedure can be started over again. Since 1997 this second round can lead to moderate financial sanctions.(16) In many cases the Commission’s announcement that it is demanding that the ECJ impose sanctions leads to hectic activity at the national level (e.g. Falkner et al. 2005, 109–110). So far there have been only two cases in which sanctions were definitely imposed – both outside the area of social policy.(17)

 
 
 
 

7

Having described the legal basis of the infringement procedure, I shall now turn to the use the European Commission makes of this instrument. Overall there are 74 EU social policy directives that had to be transposed by the end of 2003. For 96.58% of these directives, notification of national transposition measures had been given (http://europa.eu.int/comm/secretariat_general). However, this does not mean that Member States did so on time, nor that the measure met the standards of the directive. Cross-sectoral data on EU infringement procedures running in 2003 indicate a substantial implementation deficit (1,552 Letters of Formal Notice, 553 reached the state of a Reasoned Opinion whereof 215 were referred to Court; Commission of the European Communities 2004, Annex II).

For the smaller sample of six directives in 15 Member States examined in this article, 65 Letters of Formal Notice and 30 Reasoned Opinions were sent by 2003. Ten Referrals to the ECJ and four Rulings of the ECJ took place. This roughly fits the tendency seen in the cross-sectoral European Commission data above. But a comparison of this information with the much lower actual compliance in the Member States derived from our case studies (see Falkner et al. 2005, chapters 4–9) shows that the Commission only responded to the ‘tip of the iceberg’ of non-compliance. Taking the Commission’s own rules on how to follow up on non-compliant Member States as a benchmark,(18) we see that infringement procedures either were not carried through (59%) or were lacking altogether (20%). Differentiating between enforcement of late transposition and of incorrect transposition, the analysis further showed that the Commission makes more effective use of its formal enforcement power when Member States are late in transposing commonly agreed measures than when they transpose European standards incorrectly.(19) Even more remarkably, enforcement for non-application or incorrect application did not take place at all.

One could interpret these findings as a sign that the European Commission considers enforcement to be an adequate logic for increasing the timeliness of implementation, whereas it counts less on its positive effect for improving the correctness of transposition. However, the uneven use that the Commission makes of infringement procedures is more likely due to the bigger administrative burden that the enforcement of incorrect transposition measures carries. It is easy to assess whether a Member State has given timely notification of a piece of legislation. In contrast, to control for the fit or misfit between EU standards and the national transposition measure often requires translation and detailed legislative analysis by (external) experts – both of which demand considerably more time and resources. Furthermore, an infringement procedure against a Member State is in general only started after the legal situation for the relevant standard has been clarified for all other Member States, too (interview KOM1).(20) Hence, enforcement logic in the form of infringement procedures is an important and widely used part of EU implementation policy, but its formally strong power is constrained by cumbersome internal procedures and limited resources available for the task.

This observation is supported by the increasing importance that the European Commission attaches to another part of the EU enforcement policy: whistle-blowing. Third parties at the sub-national level (individuals or organized interests such as trade unions) make the supranational level aware of the (presumed) legal or actual non-compliance of a Member State by contacting the Committee of Petitions at the European Parliament or by sending a complaint letter to the EU Ombudsman. In recent years the communication of individuals’ knowledge from the sub-national level has been successfully nurtured by the Commission, for example by means of ready complaint forms available at the Secretariat General’s homepage.(21) Overall, breaches of commonly agreed rules by Member States are more often detected through individual complaints than through routine checks of reports by the Commission. The Commission (2003, 6) itself describes complaints from the sub-national level as ‘the chief source for detecting infringements’. The systematic use of whistle-blowers helps the Commission to broaden its information base. Whistle-blowing has been encouraged not to replace the infringement instrument, but to feed into the infringement procedure and to make efficient use of enforcement.


8

To complete the picture of formal instruments for enforcement, preliminary rulings (Article 234 ETC) should be mentioned. In a preliminary ruling a national court asks the ECJ to interpret EU legislation in the light of a specific national case. In combination with the mechanism of direct effect, it allows individuals (under specific conditions) to sue their state authorities for non-implementation or incorrect implementation of EU rules.(22) Preliminary rulings inform the European Commission about possible cases of non-compliance, but they also indirectly exert pressure for compliance. Preliminary rulings can thus be considered to follow an enforcement logic. Their use does not directly depend on the Commission’s interest in changing Member State behaviour, but rather more on national legal cultures and systems (on this issue, see Tesoka 1999; Alter and Vargas 2000). As for whistle-blowing, however, the Commission does indirectly encourage the use of this instrument through financial support for the training of national judges on Community law and through exchange programmes (e.g. GROTIUS).

Finally, there are two other noteworthy instruments that follow an enforcement logic. Both build on public opinion and some kind of peer pressure, and thus on discrediting Member States in the arena of the IO. Naming and shaming can develop into a powerful enforcement instrument when the European Commission decides to publicly pillory non-compliant Member States. The success of naming and shaming depends crucially on Member State sensibility. Under certain conditions, such as holding the presidency and thus attracting increased international media attention, sensibility is almost always high. In these cases, Member States have shown great interest in avoiding infringement proceedings or even a judgment by the ECJ. Belgium, for instance, reformed its administrative structures in the run-up to the EU presidencies of 1993 and 2001 by introducing the post of a European Coordinator and later a Commissioner for Interministerial Coordination, in order to avoid a presidency ‘that would be pointed at because we did not transpose a series of EU instruments’ (interview B1, translation MH; also see interview F7).(23)

By the same token, the scoreboard also exploits Member States’ sensibility. This instrument provides a direct comparison of Member States’ performance in giving notification of the transposition of EU directives in a specific sector. Although its regular format is probably attracting less media interest than highly visible cases before the ECJ, it stands for a more systematic use of peer pressure and increases competition among Member States. When introducing the instrument, then Commissioner Mario Monti explained: ‘Member States do not like to have the finger pointed at them.… It is part of the game to embarrass’ (FT 1997).(24)

Having looked at different formal and informal instruments to increase pressure on Member States to implement international rules, I shall now discuss the use that the European Commission makes of instruments more clearly following the logics of management and persuasion.

5.2. Management and persuasion strategies in the EUto contents list

During the transposition phase of a directive, bilateral or package meetings sometimes take place. Both aim at improving the implementation of EU directives in the Member States and are clear-cut examples of cooperation instruments that fall under the category of management logic. In bilateral meetings, officials from the national administration meet annually with Brussels officials from the European Commission’s General Secretariat to discuss general implementation difficulties.(25) Package meetings for single sectors unite high-ranking national and Brussels officials to discuss concrete implementation problems on ongoing as well as upcoming infringement procedures. Pressure can come into play in these potentially more conflictive cases. However, both sides endeavour to cooperate (see SEC (1999) 367:3).


9

Both types of meetings are most likely to succeed in cases in which incorrect or late transposition is based on a misunderstanding, a shortcoming in the administrative structure or procedural problems of a horizontal character. Thus, where there is a lack of expertise or resources. Similarly, the Commission (Commission of the European Communities, 2002:8) has announced that it will set up single coordination points responsible for the application of Community law in an attempt to improve administrative implementation capacities.(26) These instruments could also be considered to carry traces of the persuasion logic. The interactions create the mutual perception of problems and ease consensus through the common recognition of norms.

The instruments discussed so far focus on improving the timeliness and correctness of Member State transposition. But the implementation policy of the European Commission also includes application. The following instruments are examples of how it uses the logics of management and persuasion to improve compliance at the level of businesses and institutions. Exchange programmes for labour inspectors and financial incentives under the auspices of the Community strategy for Safety and Health 2002–2006 (see e.g. the call for tenders VP/2000/020 or VP/2000/024) directly aim to improve Member States’ capacities to ensure application. They are thus clear-cut management instruments. The founding of the European Agency for Occupational Safety and Health in Bilbao (OSHA) in 1994 as well as the founding of the intergovernmental Senior Labour Inspectors Committee (SLIC, a working group comprised of the directors from national labour inspectorates, which has met twice a year since 1995) builds on the concentration of expertise on health and safety issues and mutual learning. Experts from Southern European Member States in particular have stressed the relevance of knowledge transfer and cooperation (interviews GR7 and P6). But the determination of priority themes on specific sectors, groups of workers or risks also has elements of common norm generation.(27) Similarly, OSHA is described as ‘legitimating discourse’ (Smismans 2001, 8).

For EU social policy standards for which a successful application depends on active demand by the workers to whom they are addressed, EU implementation management is again different. The policy includes financial support of actors that promote new ideas and a change of underlying norms at the national level. The Directorate General for Employment and Social Affairs of the European Commission systematically strengthens gender-related networks and equality-promotion institutions across Europe; examples are the equal treatment think-tank KETHI in Greece (interview GR14) and the equal-treatment unit in the Spanish trade union CCOO (interview E8). The active promotion of, for example, the individual right to parental leave undertaken by such centres may influence the national discourse and in so doing change in the medium to long term what is considered appropriate.(28) In this way the Commission attempts to build up organizational structures that can then improve application through a logic of persuasion.

5.3. Overview of implementation policies in the EU systemto contents list

It has been demonstrated that the EU policies to make Member States comply with EU social policy standards have features of all three implementation logics. The use of these three logics represents an expansion of implementation policy with respect to Tallberg’s (2002) assessment of EU compliance policy as a combination of only two logics in a ‘management-enforcement ladder’. The infringement procedure that can include considerable financial sanctions is a strong instrument for exerting pressure on non-compliant Member States.

 

10

Despite the frequent use and power of this instrument, there are still many cases of infractions in which infringement procedures are not used or only inconsistently used to enforce EU directives in the Member States. I have argued that this situation is mainly due to the cumbersome internal procedures and the limited resources available for this task, that is, to a limitation of formal power through institutional characteristics. In addition, other formal instruments (preliminary rulings) and informal instruments (29) (scoreboard, naming and shaming) are used to enforce timely and correct transposition. Overall, enforcement is a widely used and highly visible logic in EU implementation policy.

Bilateral and package meetings are examples of instruments to improve the transposition of directives by means of knowledge transfer and capacity-building. Incentives for exchange programmes for labour inspectors, the Agency for Occupational Health and Safety in Bilbao and the regular meetings of Senior Labour Inspectors function along the same management logic. These interactions, as well as the work of equality-promotion centres, might also be conducive to the implementation of EU standards because they generate recognition of previously neglected norms.

As for change in the relative importance of the logics, it must be stated that in recent reforms of EU implementation policy it was expressly stated that infringement procedures should be used ‘unless the situation can be remedied more rapidly by other means.… Cases of lower priority will be handled on the basis of complementary mechanisms’ (Commission of the European Communities 2002, 12). Thus, although enforcement logic is still the most dominant in implementation policy, management seems to be gaining in importance. It is still more commonly used for cases of minor significance or cases in which non-compliance is still ‘at an early stage’, not in more deeply rooted conflicts.

Table 2

6. Implementation policies in the ILOto contents list

6.1. Formal instruments, competences and procedures

Similar to infringement procedure in the EU, ILO activities to ensure that Member States comply with the commonly agreed rules are carried out in different steps. I shall show that differences as well as similarities to the EU approach exist; they are visible in the parallel use of enforcement logic and management logic throughout all stages of the implementation policy. Table 3 summarizes the most important features of the three (not necessarily consecutive) steps.

Table 3

The regulatory supervisory process (Article 22-23 ILO constitution) provides the headquarters in Geneva with information on the implementation situation in the Member States. Member governments must regularly submit reports on the implementation of ILO conventions ratified in their country.(30) The Committee of Experts on the Application of Conventions and Recommendations (CEACR, 20 independent experts) judges the conformity with the respective ILO convention. They use additional sources of information, such as collective agreements, court cases or results from labour inspections. Preparatory work is carried out by the International Labour Standards Department in the ILO, whose officials ‘work effectively as brokers and, in this capacity, to safeguard their basic principles of policy in the solutions that are reached’ (Cox 1974, 112). The resources of this department with 74 regular staff members (thereof 52 legal experts) are limited when facing the continuously increasing number of reports.(31)


11

When the CEACR reveals difficulties concerning compliance, it can react either by making a direct request for clarification to the government (mostly for ‘technical’ problems) or by issuing an observation (cases of heavy and ongoing violations). Whereas direct requests are used to shed light on unclear situations (management), the aim of observations is to put pressure on the respective member government (enforcement). Finally, the instrument constitutes an important element in stimulating dialogue both with the governments and among the delegates in the GB, possibly leading to a more in-depth recognition of the international labour standards. This instrument has elements of all three logics.

The exertion of pressure becomes more important when 20 to 25 cases are picked to be examined again. Selection is made by the political Conference Committee on Application of Standards (150 members from the GB). The choice of cases is subject to consensus on which problems should receive specific attention and represents a balanced selection of issues and geographical areas. Workers’ representatives select a list of cases, and the employers’ side adds a few cases only (interview ILO14). Formal enforcement powers are reduced through the bargaining process between different interests in the ILO.

The same holds true for the ranking of the cases in different groups (failure to comply with reporting obligations, cases of progress, certain special cases and continued failure to comply). The ranking is discussed and adopted by the plenary of the GB on a consensus basis. Unlike in the EU, where early stages of the infringement procedure are mainly carried out by the administration, the ILO reporting procedure is dominated by a political logic and is highly contingent on the interests of many actors.

A less contingent instrument of coercion was introduced in 1998 on an initiative by the General Director of the ILO (very likely in reaction to the actual limitations of delegated implementation powers). For fundamental conventions a reporting obligation was introduced – disregarding whether a country has ratified.(32) Pressure can now be exerted swiftly without lengthy political decision-making. In two different types of reports experts and the General Director critically describe success or failure, name laggards as well as non-compliers (Annual Reviews), present countries in scoreboards according to their performance on implementation and reporting duties and explicitly list manifest violators (Global Reports; e.g. ILO 2004, 24). This approach can be considered an important shift in the implementation policy towards an enforcement logic in two ways. First, it establishes direct comparability between all member countries, and, second, by doing so it increases the (moral) pressure on those states that are lagging behind in ratification. The proper functioning of the new instrument was underlined by the criticism by those Member States that saw the finger pointed at them (see Elliott 2000, 3–4).

The second path along which the ILO can respond to non-compliance is called a representation procedure (Article 24-25 ILO constitution). National or international employers’ or workers’ organizations can claim that a given state has failed in implementation of a ratified convention by issuing a ‘representation’.(33) The GB decides about setting up an Ad Hoc Committee (three members) to deal with the question and about whether a critical report on the case should be adopted. More than the reporting procedure, this instrument follows an enforcement logic from the beginning through explicitly raising public awareness. But similar to the reporting procedure, the decision to increase pressure depends on the support of the GB. Even though no vote is taken, in actuality each of the three groups can exert veto power at various stages in the process. And thus only consensual cases make it.


12

These institutional constraints result in a rather scarce use of representations. Employers’ (5) and workers’ (100) organizations made uneven use of the enforcement power entitled to them. Noticeable is the progressive use over time; from its initiation in 1924 until the end of the 1980s, 37 cases were lodged, whereas the bulk of representations (78) reached the ILO in the last 14 years (1990–2003).(34)

The most powerful instrument in formal terms is the complaint procedure (Article 26-34 ILO constitution). It resembles an advanced EU infringement procedure. It is used to follow up cases of the regular supervisory process with ‘continued failure to comply’ or cases of the representation procedure. Like the comparable instrument in the EU, a complaint is usually invoked in the case of persistent violations and disregard for the decisions of the ILO bodies. The GB can appoint a Commission of Inquiry (three prominent, independent persons) that will prepare a report with recommendations and a deadline. This report will be discussed by the GB, but again without a formal vote and on a consensus basis. The member government has to reply within three months as to whether it is willing to accept the proposed steps to be taken. Alternatively, it can challenge the recommendations and the arguments put forward by the ILO before the International Court of Justice.(35)

Although the formal powers delegated to the EU under the infringement procedure and to the ILO under the complaint procedure seem quiet similar, practical use is more constrained in the ILO. Apart from a first attempt that never really went through in 1934, the complaint procedure was not used in the first 40 years of the ILO (Valticos 1994, 108). In the meantime there have been 26 cases. Similar to and most likely based on increasing cases in the reporting and representation procedure, the number of complaints lodged per decade increased from one in the 1960s to seven in the 1990s. Eleven Commissions of Inquiry were called into action by 2003, pertaining to less developed countries and industrialized countries, from the south as well as the east. Member States did not always follow the recommendations (e.g. Poland resisted the introduction of freedom of assembly and Germany continued to ban workers from professions for political reasons; Elliott 2000, 12), and they never made recourse to the International Court of Justice. Nevertheless, it was only in June 2000 that the ILC for the first time ever took further action against a non-compliant member country.

6.2.Capacity-building: A mandate of the ILO to contents list

In this section, I discuss capacity-building activities of the ILO which are used independently as well as in conjunction with instruments presented so far.(36) After the Second World War, technical assistance constantly gained in importance, and it is now at the level of standard-setting activities (Dufty 1972). It is implemented vertically through specialized departments that offer guidance and expertise in their field of competence (e.g. the drafting of national legislation or the training of administrative staff and social partners). It also is implemented horizontally in sub-regional and national offices (e.g. through country-specific implementation assistance). A department specialized in ‘development cooperation’ coordinates the projects funded by the ILO or donors. In addition to the improved implementation of ILO conventions, ratification is often at the heart of such projects. For this instrument, management through knowledge transfer and financial support is the dominant logic.

Recent studies in the area of child labour (Liese 2003) or with respect to the financial support of small enterprises to enable cooperation in employers’ associations (Senghaas-Knobloch 2004, 154) have concluded that such management approaches are often linked to a logic of persuasion as well. Further research is needed to systematically determine when and how financial and technical assistance is used to make nations comply with international labour standards and, by contrast what might have changed in the perception or judgement of a country independent of incentives or punishment.(37)


13

Institutional leeway is greater in capacity-building than in the reporting, representation or complaint procedures, where formally delegated powers are often hampered by the institutionally entrenched need for consensus. In principle, groups of actors within the ILO also can oppose a technical assistance project. But, contrary to decision-making in the GB or ILC, where positive selection is a precondition for continuation of the procedure, in technical assistance an explicit and substantial opposition is required to stop the procedure.

Another implementation instrument following a management logic is direct contacts. They complement cases under the supervisory and enforcement procedures. Long before the EU first held bilateral and package meetings (in the 1980s), ILO representatives started to visit the country in question in order to talk with high government representatives about application difficulties for a specific convention or group of conventions. Characteristic of these meetings is that the procedure is comparatively ‘lean’ and that there is an atmosphere of confidentiality – far from the political discussions in Geneva – which often leads to fast and straightforward solutions (Valticos 1981, 479–480 and 488).(38)

With respect to the importance of direct contacts, they were used 42 times from 1987 to 2003, or an average of 2.6 times per year. The number varies from year to year, without a clear-cut tendency towards increase or decline.(39) Similarly, there are ‘study groups’, ‘inquiry missions’ and ‘multidisciplinary advisory missions’, which often undertake visits to countries that are suspected of having compliance problems. Their assignment is less directly linked to an ongoing enforcement procedure.

In the following section I change the perspective to a single case for country-level analysis. I do so to draw out a specific stage in the ILO enforcement procedure.

6.3. The Myanmar case: Maximum pressure with continuous managementto contents list

The Myanmar case reveals concrete features of the ILO implementation policy and the simultaneous use of instruments that follow a logic of management or enforcement. From the current context, this is an extraordinary response to an extraordinarily serious violation of one of the most fundamental ILO standards. Depending on future developments, it might also indicate a trend towards a more thorough use of enforcement logic in the ILO.

Myanmar (Burma) was found guilty of neglecting the ILO principles on forced labour despite having ratified C29 in 1955. Ever since a complaint procedure was started in 1996, pressure on the country has continuously increased – but always along with efforts to help the country stop forced labour. After a critical report by a Commission of Inquiry in 1998 and the neglect of recommendations on how to improve the situation, the ILC decided in 1999 to further tighten measures. Technical cooperation and assistance from the ILO was cut; only the right of direct assistance in immediately implementing the recommendations of the Commission of Inquiry was upheld.

 
 
 
 
 

14

Formal adherence took place in response, but it proved to have little impact on events on the ground (ILO 2001). In June 2000 the ILC for the first time called for sanctions on a Member State found to be in continuous breach of international labour standards.(40) The ILO has no direct sanctioning power like that of the ECJ in the EU system; thus, it called on member countries and other IOs to reconsider any cooperation that might contribute to forced labour. The ILO itself withdrew stocks of its pension fund from Myanmar’s financial market even though returns were still high. The reaction from other IOs was rather weak. Trade unions took a more proactive stance. At the international level, they had started campaigning for disinvestment and trade rupture even before the adoption of the resolution (publishing a ‘red list’ of companies linked with Myanmar). Now highly unionized dockworkers in India and Bangladesh have started to refuse loading ships from Myanmar. Overall, the approach taken to the resolution was one of ‘wait and see’ (interview ILO1).(41)

As in the proceeding steps, this increase in pressure went hand in hand with continued cooperation in order to promote the full implementation of the recommendations by the Commission of Inquiry. Since spring 2002 an ILO Liaison Officer works on capacity-building in the country in order to effectively address the root causes of forced labour. In May 2003 a joint plan of action between the ILO and the government of Myanmar was negotiated (ILO 2003a) but then suspended after the detention of opposition leader Daw Aung San Suu Kyi. In February 2005 a High-Level Team visited Myanmar. Although the governmental interest to defect is obvious, this ILO policy reflects an awareness that even if the government were willing compliance would still depend on cooperation and capacity-building.

Compared to ECJ rulings imposing sanctions, the pressure exerted by the ILO by means of Article 33 remained weak. Unlike the EU, the ILO has to rely on others ‘to bite’.(42) Significant is that the ILO for the first time used all of its formally delegated powers to press for the application of its policies. Factual veto points were no longer used. It will be of interest to see whether such actions will be taken for other cases, too – and whether consensus of action can be reached in less clear-cut cases or in cases accusing bigger or more developed countries.

On the basis of the Myanmar case, we should expect that the affiliation with a group of countries with generally similar interests and its behaviour in the run-up will be more important to predict the likeliness of such action than the sheer size of a country or employer-worker cleavages on the issue. A faster and more decisive approach against Myanmar was limited by the stance taken by its East Asian neighbour states, whereas both employers’ and workers’ representatives wanted to move more quickly.(43) Moreover, the ILC and the GB both were reluctant to go ahead with sanctioning measures each time Myanmar showed willingness to cooperate. These are very tentative expectations because they are based on one case only. Likewise, future developments will have to provide an empirical answer to the question of whether the ILO implementation policy makes increasing use of enforcement logic.(44)

6.4. Overview of implementation policies in the ILO systemto contents list

The ILO implementation policy, like that of the EU, has features of all three implementation logics. Capacity-building and knowledge transfer seem to be the implementation policies most often employed in the ILO. Examples are general technical assistance projects, which are at the core of ILO activities, and direct contacts in reaction to the detection of non-compliance. In reporting, representation and complaint procedures, enforcement (e.g. observations) and management (e.g. direct requests) co-exist in different degrees. In all cases in which the procedure provides for increased interactions through meetings and exchanges between the Member State and ILO officials, persuasion also can play a role in advancing the implementation of international norms. However, persuasion is not the driving logic behind the use of these implementation instruments.


15

Table 4

The ILO does not use powers delegated by the Member States to the extent formally possible. There are two explanations: First, Member States are not bound to ratify ILO conventions (low obligation to rules; see Abbott et al. 2000, 406). In these cases, exerting pressure is not of much use, although financial or technical assistance can lead to ratification. This point supports the argument that compliance with international rules can be better understood if, in addition to the absolute level of delegation and legalization, the profile of legalization is taken into account (Zangl and Zürn 2004b, 254; also see Zürn and Neyer 2005, 200). Second, the limited use of enforcement also can be explained by the necessity of a broad political consensus to carry out the different steps leading to (indirect) sanctions. Even though no formal vote is taken (until the procedure reaches Article 33), the tripartite GB and the ILC have to come to an agreement on which cases they wish to pursue further. Here, actual strength is weaker than formal strength as a result of institutional constraints. Probably in reaction to these contingencies, another alternative to exert pressure was introduced. Reporting duties now also apply to unratified conventions if they are fundamental conventions. Moreover, Member States are named and shamed when in breach. Overall, the ILO is more constrained by political cleavages in its use of the implementation powers delegated to it than the EU is.

7. Conclusionto contents list

The categorization of the ILO and the EU as IOs in which Member States have delegated implementation powers to a comparably high degree (Abbott et al. 2000, 406) was taken as a starting point. The article then analysed differences with respect to these delegated powers and the use that the ILO and the EU make thereof to address non-implementation or incorrect implementation of commonly agreed standards. It did so along the three implementation logics of enforcement, management and persuasion. This approach was considered valuable for two reasons: It makes specific aspects of EU activities comparable with other IOs, and it advances our understanding of delegation as a specific aspect of ‘legalization’ through empirically driven differentiation.

The first important result is that implementation instruments of the EU and the ILO both showed features of all three implementation logics: enforcement, management and persuasion. From the perspective adopted in this article, differences in the implementation policies of the two IOs can thus be considered questions of degree.(45) A second result refers to the relative importance of the implementation logics within the IO. The use of instruments following an enforcement, management or persuasion logic was assessed comparatively. Furthermore, the increasing use of specific instruments over time was taken into account in order to identify change in relative importance. On these grounds I have argued that within the EU system enforcement is most developed, whereas the logic most widely employed in the ILO system is management. This can be explained by differences in the institutional settings of the IOs. In the ILO the power to exert pressure is more constrained through a need for consensus in the decision-making bodies than in the EU. At the same time, administrative capacities to provide technical assistance are greater in the ILO than in the EU.

 
 
 
 
 

16

Furthermore, my findings indicate the successive use of instruments dominated by a management or an enforcement logic in the EU; in the ILO, they are more likely to be used in parallel or even complementarily. The Myanmar case depicted above clearly shows that in the ILO system, management continues along all stages of an enforcement procedure. This approach might reflect awareness that in many ILO Member States non-compliance stems from a lack of interest and, at the same time, a substantial lack of capacity to implement the international standards. This situation does not hold true to the same extent for the EU Member States. (This point does not exclude the possibility that in the EU chronological use may also be due to the role of the European Commission as Guardian of the Treaties.) Finally, with respect to the absolute strength of a logic, formal and actual powers were taken into account.

Table 5

Enforcement: In the EU system the infringement procedure provides for a strong framework. Even though cumbersome procedures and limited resources constrain the use of financial penalties, they are still a realistic scenario once an infringement procedure has been initiated. This Sword of Damocles has a positive influence on Member State compliance. Actual enforcement strength is moderate. The ILO can exert more pressure than is widely assumed, but sanctioning power remains indirect. In order to fully use its enforcement power, the ILO needs a broad political consensus among its stakeholders. So far such a consensus has been reached only once (Myanmar). A tentative expectation derived from the Myanmar case is that the consensus needed to carry out enforcement will depend on (a) the degree and clearness of the breach, (b) the country’s position within the ILO and (c) an uncooperative stance of the Member State. Future developments will show whether this case is the beginning of an increasing use of enforcement logic. So far, formal strength is moderate, and actual strength is moderate to weak.

Management: With respect to instruments that help to solve problems hindering the implementation of international standards, the ILO can be regarded as strong. Its organizational structure and procedures have been explicitly set up to fulfil the tasks of capacity-building and knowledge transfer (e.g. Dufty 1972, 490). In the absence of more rigorous enforcement policies, and in the face of Member States that often lack basic administrative or financial capacities to put the conventions into practice, it seems logical for the ILO to make recourse to management rather than to enforcement. Because Member States are not bound to ratify ILO conventions, there is little use in exerting pressure here. At the same time, financial support in the form of projects or technical assistance can give a boost to the ratification of international labour standards.

The European Commission uses positive incentives and knowledge (e.g. bilateral and package meetings, exchange programmes and expert networks). But, its opting for an extensive use of management instruments when facing non-compliance could – at least in the past – collide with its mission as Guardian of the Treaties and hinder consistent implementation of the rules. In recent reforms of EU implementation policies, management seems to have gained in importance. Nevertheless, it is still more commonly used for cases of minor significance or cases in which non-compliance is still ‘at an early stage’, not for more deeply rooted conflict.

 
 
 
 
 

17

Persuasion: Improvement of compliance with international standards through recognition of the appropriateness of a norm was visible in both the EU and the ILO. The analysis, however, did not find that either IO provides for explicit and strong instruments that could exploit the persuasion logic more systematically. The ILO seems to be more inclined to use persuasion. There is no obligation for Member States to implement ILO conventions, and countries often have to be convinced to ratify a convention in the first place. I therefore consider the logic of persuasion to be weak in the EU and moderate to weak in the ILO. The way an IO uses the implementation power delegated to it is influenced by the (low) level of obligation – a separate aspect of legalization as discussed by Abbott and colleagues (2000). This last point links up with recent developments in the debate on legalization. For example, Zangl and Zürn (2004b, 254; also see Zürn and Neyer 2005, 200) argue that compliance with international rules can be better understood if, in addition to the absolute level of legalization, the profile of legalization is taken into account.

Having systematically analysed and categorized the formal and informal supervision and enforcement instruments in the EU and the ILO as well as their relative importance, this article has prepared the ground for further research on the effectiveness of different implementation logics. Can more enforcement power ensure better compliance? If so, the EU system would be more likely to generate compliance with its standards. Enforcement might be less able, however, to influence the application of international rules at the national ground level.


Referencesto contents list

Abbott, Kenneth W./Robert W. Keohane/Andrew Moravscik/Ann-Marie Slaughter/Duncan Snidal, 2000: The Concept of Legalization. In: International Organization 54(3), 401-420.

Alter, Karen, 2003: Do International Courts Enhance Compliance with International Law? In: Review of Asian and Pacific Studies 25, 51-78.

Alter, Karen J./Jeannette Vargas, 2000: Explaining Variation in the Use of European Litigation Strategies: European Community Law and British Equality Policy. In: Comparative Politics 33(4), 452-482.

Chayes, Abram/Antonia Handler Chayes, 1993: On Compliance. In: International Organization 47(2), 175-205.

Checkel, Jeffrey T., 2001: Why Comply? Social Learning and European Identity Change. In: International Organization 55(3), 553-588.

Commission of the European Communities, 2002: Commission Communication. Better Monitoring of the Application of Community Law, COM(2002) 725 final/4. Brussels: Commission of the European Communities.

Commission of the European Communities, 2003: 20th Annual Report on Monitoring the Application of Community Law, COM(2003) 669 final. Brussels: Commission of the European Communities.

Commission of the European Communities, 2004: 21th Annual Report on Monitoring the Application of Community Law, COM(2004) 839 final. Brussels: Commission of the European Communities.

Cox, Robert W., 1974: ILO: Limited Monarchy. In: Robert W. Cox/Harold K. Jacobson (eds.), The Anatomy of Influence. Decision Making of International Organizations. New Haven: Yale University Press, 102-138.

Downs, George/David M. Rocke/Peter N. Barsoon, 1996: Is the Good News about Compliance Good News about Cooperation? In: International Organization 50(3), 379-406.

Dufty, Norman F., 1972: Organizational Growth and Goal Structure: The case of the ILO. In: International Organization 26(3), 479-498.

Elliott, Kimberley A./Richard B. Freeman, 2003: Can Labor Standards Improve Under Globalization? Washington DC: Institute for International Economics.

Elliott, Kimberly Ann, 2000: The ILO and Enforcement of Core Labor Standards: Institute for International Economics (July 2000).

Falkner, Gerda/Treib Oliver/Hartlapp Miriam/Leiber Simone, 2005: Complying with Europe. Theory and Practice of Minimum Harmonisation and Soft Law in the Multilevel System. Cambridge: Cambridge University Press.

Finnemore, Martha/Stephen J. Toope, 2001: Alternatives to 'Legalization': Richer Views of Law and Politics. In: International Organization 55(3), 743-758.

FT, 1997: Brussels Move To Shame EU States Over Single Market. In: Financial Times, 16.2.1997, 1.

Goldstein, Judith/Miles Kahler/Robert O. Keohane/Anne-Marie Slaughter, 2000: Introduction: Legalization and World Politics. In: International Organization 54(3), 385-399.

Goldstein, Judith/Miles Kahler/Robert O. Keohane/Anne-Marie Slaughter, 2001: Response to Finnemore and Toope. International Organization. In: International Organization 55(3), 759-760.

Gravel, Eric/Chloé Charbonneau-Jobin, 2003: La Commission d'Experts pour l'Application des Conventions: Dynamic et Impact. Genève: Organisation International du Travail.

Haas, Peter M./Robert O. Keohane/Marc A. Levy (eds.), 1994: Institutions for the Earth. Sources of Effective International Environmental Protection, Global Environmental Accords Series, Nazli Chooucri. Cambridge, MA: MIT Press.

Hart, H. L. A., 1968: Punishment and Responsibility. Essays in the Philosophy of Law. Oxford: Clarendon.

Hartlapp, Miriam, 2005: Die Umsetzung europäischer Sozialpolitik. Die Kontrolle der nationalen Rechtsdurchsetzung durch die Europäische Kommission, Politik, Recht, Verbände: Bedingungen und Grenzen der Umsetzung der europäischen Sozialpolitik. Vol. 3. Frankfurt/M.: Campus.

Héritier, Adrienne, 2001: Differential Europe: The European Union Impact on National Policymaking. In: Adrienne Héritier/Dieter Kerwer/Christoph Knill/Dirk Lehmkuhl/Michael Teutsch/Anne-Cécile Douillet (eds.), Differential Europe: The European Union Impact on National Policymaking. Lanham: Rowman and Littlefield, 1-21.

Hix, Simon, 1998: The Study of the European Union II. The 'New Governance' Agenda and Its Rival. In: Journal of European Public Policy 5(1), 38-65.

ILO, 2000: Follow-up on the recommendations of the Working Party (GB.277/LILS/WP/PRS/1/1). Geneva: International Labour Office.

ILO, 2001: ILO team finds limited impact of new legislation against forced labour in Myanmar. Suggests action for further progress (press release 01/48). Geneva: International Labour Organization (7 November 2001).

ILO, 2003a: ILO, Myanmar agree on facilitator to help end forced labour (press release 03/21). Geneva: International Labour Organization (14 May 2003).

ILO, 2003b: Report of the Committee of Experts on the Application of Recommendations. Geneva: International Labour Office.

ILO, 2004: Organizing for social justice. Global Report under the Follow-up to the ILO Declaration on Fundamental Principles and Rights at Work 2004. Geneva: International Labour Office.

Keohane, Robert W./Andrew Moravscik/Ann-Marie Slaughter, 2000: Legalized Dispute Resolution: Interstate and Transnational. In: International Organization 54(3), 457-488.

Liese, Andrea, 2003: Internationale Regulierung im Politikfeld Kinderarbeit. In: Eva Senghaas-Knobloch/Jan Dirks/Andrea Liese (eds.), Internationale Arbeitsregulierung in Zeiten der Globalisierung. Politisch-organisatorisches Lernen in der Internationalen Arbeitsorganisation (IAO). Münster: LIT Verlag, 39-125.

March, James G./Johan P. Olsen, 1989: Rediscovering Institutions. The Organizational Basis of Politics. New York: Free Press.

Mayer, Peter, 2004: Verantwortungsethische Grenzen der Verrechtlichung? Zur realistischen Kritik des Legalismus in den Internationalen Beziehungen. In: Bernhard Zangl/Michael Zürn (eds.), Verrechtlichung - Baustein für Global Governance? Bonn: Dietz, 220-238.

Mbaye, Heather A.D., 2001: Why National States Comply with Supranational Law: Explaining Implementation Infringements in the European Union 1972-1993. In: European Union Politics 2(3), 259-281.

Mendrinou, Maria, 1996: Non-Compliance and the European Commission's Role in Integration. In: Journal of European Public Policy 3(1), 1-22.

Neyer, Jürgen/Dieter Wolf, 2005: Analysis of compliance with international rules: Definitions, variables, and methodology. In: Michael Zürn/Christian Joerges (eds.), Law and Governance in Postnational Europe. Cambridge: Cambridge University Press, 40-64.

Neyer, Jürgen/Michael Zürn, 2001: Compliance in Comparative Perspective. The EU and Other International Institutions. InIIS-Arbeitspapier 23. Bremen: Institut für Interkulturelle und Internationale Studien.

Olson, Mancur, 1965: The Logic of Collective Action. Public Goods and the Theory of Groups. Cambridge: Harvard University Press.

Pollack, Mark A., 1997: Delegation, Agency and Agenda Setting in the European Community. In: International Organization 51(1), 99-134.

Risse, Thomas, 2000: 'Let's Argue!': Communicative Action in World Politics. In: International Organization 54(1), 1-39.

Romano, Cesare P.R., 1996: The ILO System of Supervision and Compliance Control: A Review and Lessons for Multilateral Environmental Agreements: International Institute for Applied Systems Analysis.

Sécretariat Général, 1998: Amélioration des méthodes de travail de la Commission relatives aux procédures d'infraction. SEC(98) 1733. Bruxelles: Commission Européenne.

Senghaas-Knobloch, Eva, 2004: Zwischen Überzeugen und Erzwingen - Nachhaltiger Druck für Geltung und Wirksamkeit internationaler Arbeits- und Sozialstandards. In: Bernhard Zangl/Michael Zürn (eds.), Verrechtlichung - Baustein für Global Governance? Bonn: Dietz, 140-158.

Smismans, Stijn, 2001: The legitimacy of social partner's participation in European occupational health and safety policy. Conference paper, ECPR Joint Session of Workshops (6. - 11.04.2001), Grenoble.

Södermann, Jacob, 2001: Der Bürger, das Rechtsstaatsprinzip und die Offenheit. Conference paper, Europarechtskonferenz (10.-12.6.2001), Stockholm.

Spencer, David, 1994: Structure, Functions and Procedures in the Commission. In: Geoffrey Edwards/David Spence (eds.), The European Commission. London: Longman, 97-116.

Tallberg, Jonas, 2002: Paths to Compliance: Enforcement, Management, and the European Union. In: International Organization 56(3), 609-643.

Tesoka, Sabrina, 1999: Judicial Politics in the European Union: Its Impact on National Opportunity Structures for Gender Equality. MPIfG Discussion Paper 99/2. Cologne: Max Planck Institute for the Study of Societies.

Tsebelis, George, 1995: Decision Making in Political Systems: Veto Players in Presidentialism, Parliamentarism, Multicameralism and Multipartism. In: British Journal of Political Science 25, 289-325.

Valticos, Nicolas, 1981: Une nouvelle forme d'action internationale: Les 'contacts directs' de l'O.I.T. en matière d'application de conventions et de liberté syndicale. In: Annuaire Francais De Droit International 27, 477-489.

Valticos, Nicolas, 1994: Once More About the ILO System of Supervision: In What Respect is it Still a Model? Towards More Effective Supervision by International Organizations. Essays in Honour of Herny Schermers. Vol. I: Martinus Nijhoff Publications, 99-113.

Zangl, Bernhard, 2001: Bringing Courts Back In: Normdurchsetzung im GATT, in der WTO und der EG. In: Swiss Political Science Review 7(2), 49-80.

Zangl, Bernhard/Michael Zürn, 2004a: Make Law, Not War: Internationale und transnationale Verrechtlichung als Baustein für Global Governance. In: Bernhard Zangl/Michael Zürn (eds.), Verrechtlichung – Baustein für Global Governance. Vol. 18. Bonn: Dietz, 12-45.

Zangl, Bernhard/Michael Zürn (eds.), 2004b: Verrechtlichung - Baustein für Global Governance, EINE Welt – Texte der Stiftung Entwicklung und Frieden. Vol. 18. Bonn: Dietz.

Zürn, Michael, 2005: Introduction: Law and compliance at different levels. In: Michael Zürn/Christian Joerges (eds.), Law and Governance in Postnational Europe. Cambridge: Cambridge University Press, 1-39.

Zürn, Michael/Jürgen Neyer, 2005: Conclusions - the conditions of compliance. In: Michael Zürn/Christian Joerges (eds.), Law and Governance in Postnational Europe. Cambridge: Cambridge University Press, 183-217.


Endnotesto contents list

(*) Special thanks to the participants of the workshop on policy implementation and international organizations at the ECPR Joint Sessions of Workshops 2004 in Uppsala, as well as to Armin Schäfer and two anonymous referees, for helpful comments. The research was supported by a Brückenprogramm Scholarship of the Volkswagen Foundation. I am grateful to the ILO for hosting me during this scholarship.

(1) Even though this study focuses on the first pillar and thus, strictly speaking, European Communities (EC) would be the correct designation, I employ European Union (EU) as the more widely used term.

(2) Zangl and Zürn (2004a) have developed a different concept of legalization that is supported by a range of case studies on different international organizations. The concept takes ‘adjudication’, ‘institutionalized enforcement’ and ‘deliberative law making’ as decisive categories. However, with respect to the ILO, their categorization remains controversial, perhaps in part due to the argument that the legalization concept coined on interstate conflicts is not easily transferable to the ILO, which is based on a concept of cooperation (see Senghaas-Knobloch, 2004, 150, in the same volume). Notwithstanding this objection, Zangl and Zürn conclude that the ILO as an institution has reached a high degree of legalization (2004a, 240). This is astonishing because the assessment of the ILO in terms of their three categories points to a different direction: The ILO scores ‘medium’ in deliberative law making. Institutionalized enforcement and adjudication are described as ‘low’ (Zangl and Zürn 2004a, 241 and 247; contradicting Table 1, p. 244 categorizes adjudication as ‘medium’).

(3) I shall return to possible relationships between the different categories at the end of this article.

(4) Implementation is understood as the process of, first, taking international regulations over to the national level and, second, applying these rules at the ground level. Non-compliance can occur in both phases of this process. When a state successfully implements an international rule, it is considered to be in compliance. Thus, implementation is a process, whereas compliance is a possible outcome thereof.

(5) Adjudication is often considered to be another, separate approach (e.g. Alter 2003). The role of courts is valuable in explaining inter-country conflicts in trade (Zangl 2001). However, with respect to the cases of social policy implementation through IOs under examination here, the courts only come into play at a late stage (if at all). Then they can either serve to increase pressure, for example through imposing sanctions (enforcement logic), or create certainty with respect to international rules (management logic). Therefore, adjudication will not be treated as a separate implementation logic in this article.

(6) Starting from the assumption that actors (here: states) do not want to comply, the enforcement concept builds on a logic of consequences to reach compliance, whereas the concept of persuasion emphasizes that compliance with norms is achieved through the recognition of norms following a logic of appropriateness (March and Olsen 1989). This distinction is difficult to apply to the management concept. The management concept does not differentiate whether the willingness to comply is based on interest or on a feeling of appropriateness. Its main interest lies in the lack of capacity as an explanation for non-compliance. However, because in this article I focus on the logics of instruments that are used in implementation policy, I argue that enforcement and management instruments both function along an instrumental logic.

(7) This approach would in turn lower the state’s credibility, and thus its ability to win support for future pursuits, commitments, etc.

(8) Such an assessment does not exclude that changes in behaviour can be attained on a minor scale.

(9) EU directives set binding goals, but they also leave room for the form and means by which the Member States ensure that these goals are met at the national level within a given time period.

(10) For more information on the research project “New Governance” and Social Europe: Theory and Practice of Minimum Harmonization and Soft Law in the Multilevel System, please consult our homepage (http://www.mpi-fg-koeln.mpg.de/socialeurope/).

(11) Genuinely supranational topics such as the directive on European Works Councils (94/45/EC) were discarded from our sample because we wanted to study areas in which EU regulation supersedes national regulation. We also discarded directives that only update or reform older ones, as well as directives that are too closely related to some other EU laws to be studied individually.

(12) The ILO was founded in 1919 through the Treaty of Versailles and was integrated into the United Nations system after the Second World War (1946), becoming the first specialized UN agency.

(13) From a legal point of view this makes a fundamental difference. Once a convention is ratified, a ‘contract’ exists between the Member State and the IO (this holds true for regulations and directives in the EU system, too).

(14) Eight of these conventions are considered ‘fundamental’. They cover freedom of association and the right to collective bargaining (C87 and C98), elimination of forced and compulsory labour (C29 and C105), abolition of child labour (C100, C138 and C182) and elimination of discrimination in the workplace (C100).

(15) Such a strategy can be understood, for example, through the principal-agent approach, which argues that states anticipate paying high costs in the long run for breaking agreements in order to achieve immediate gain (Pollack 1997). In a similar way, such a strategy may be in the interest of governments to increase the credibility of a commitment (see Alter 2003, 59–60).

(16) Financial sanctions were formally introduced in the Maastricht Treaty in 1993. But the rules on how to calculate the fine from a lump sum and different coefficients were only adopted in 1997. They amount to between 500 euro and 791.293 euro per day.

(17) The penalization of Greece for a dump in Korupitos, Crete (case C-387/97) and of Spain for the insufficient quality of its bathing water (case C-278/01).

(18) These rules in internal Commission documents state that non-compliant Member States must be sent a Letter of Formal Notice at the latest a year after the end of the transposition period or a year after incorrect measures have been notified. Thereafter, no more than a year shall pass for each subsequent step of the infringement procedure (SEC (93) 1288, Sécretariat Général 1998, 4).

(19) ‘Correct’ enforcement for failure to notify took place in 65% of the cases (31 of 48 late transpositions), whereas enforcement was ‘incorrect’ in nine cases (19%) and lacking altogether only two times (4%). A focus on formal pressures in response to incorrect transposition was used in less than half of the cases (49%; 23 of 47), and only once was enforcement carried out in accordance with its own rules. In the remaining 22 cases, enforcement pressure was exerted only late or stopped before compliance was ensured (47%; for details, see Hartlapp 2005, 192–196).

(20) Recent improvements of the notification procedure announced by the European Commission nicely fit this observation (e.g. concordance tables to facilitate conformity judgements; Commission of the European Communities 2002, 9).

(21) The average number of annually registered complaints grew from 536 (1983–1989) to 1,047 (1990–1998) and reached 1,346 in 2002 (Commission of the European Communities 2002, 13–14).

(22) Established in ECJ ruling Van Gend en Loos (C-26/62) and, more specifically, for directives in Van Duyn (C-41/74).

(23) On the other hand, there are also cases in which increased Member State sensibility during the presidency made the Commission give in, as the following quotation shows: ‘To save the British government possible embarrassment and to help establish good relations with London during the UK Presidency in the second half of 1992, for instance the Commission repeatedly requested the Court to postpone a number of highly publicized environmental law cases against the UK’ (Spencer 1994, 111).

(24) Until 2001 benchmarking on implementation performance was limited to the Common Market and to environmental issues. In Nice, France, the Council introduced a scoreboard for social policy.

(25) They were first installed in 1987 and have been more systematically used to improve implementation after an internal Commission reform in 1993 (SEC (93) 216/7:10).

(26) SOLVIT is another example in which the transfer of knowledge is systematically enhanced in order to improve compliance with EU standards through out-of-court decisions and ‘lean’ solutions – though only for the Common Market area and not for social policy (http://europa.eu.int/comm/internal_market/solvit/).

(27) There are definitely also ‘hard interests’ in equal compliance with EU rules in order to prevent social dumping and unfair competition between production locations (interview DK5). This point is not to disregard that SLIC meetings can generate the recognition of previously neglected norms.

(28) Of course, the overall success of such measures crucially depends on additional factors such as general developments in employment and, in particular, pay differences.

(29) Informal instruments are not explicitly laid down in the Treaty on European Union or provided for by secondary legislation.

(30) This practice follows a standardized formula for each convention, covering aspects such as incidence of related court rulings and statistics. Reports have to be submitted in two- or five-year intervals, depending on the subject of the convention. Employers’ and workers’ representatives can comment on these reports in order to provide a more balanced picture.

(31) In 1927, when there were 23 conventions, 180 reports had to be examined (Gravel and Charbonneau-Jobin 2003, 2). Since then the annual number of reports rose to an average of 1,267 in the 1990s (ILO 2003b, 744–745).

(32) Prior to 1998 such an obligation could be pronounced by the GB only for individual cases (Article 19).

(33) The formal ‘access’ of the social partners to the supervisory and enforcement system in the ILO constitutes a main difference to the EU system, but flows logically from their greater involvement in the standard setting. The reporting procedure might look similar to the whistle-blowing in the EU system at first glance. They both qualify as ‘transnational dispute resolution’ (Keohane et al. 2000, 458) open to individuals and groups in society. In consideration of the tripartite nature of the ILO, representations raise constituents’ interests into a publicly visible sphere. Albeit conflicting cases are unlikely to be made a point of in the GB, the procedure is in principle open and democratic. In the EU the use of the individual complaint depends on the unclear preferences of the EU Commission ( for a critical review, see Södermann 2001; Sécretariat Général 1998).

(34) Above all, this development is probably the result of increased acceptance of the procedure over time – stakeholders have learned how to use this mechanism to their advantage. Partly it is also a result of the above-mentioned growth in the number of ILO members and conventions, which has enlarged the number of potential cases of conflict. And partly it is due to political changes in the world which make representations on specific issues possible for a whole group of countries which were not likely before. We can distinguish ‘waves’ of representations, for example in dealing with colonial or post-colonial conflicts (1930s), on issues of the repression of workers and workers’ rights under Latin American military regimes (1980s) and the growing strength and self-confidence of trade union movements in the former Soviet countries (from the early 1990s on).

(35) Similarly, under the freedom of association procedure the tripartite Committee on Freedom of Association deals with complaints breaching the ‘freedom of association principle’ (C87 and C98, even if not ratified). Equivalent to the Commission of Inquiry, a Fact-Finding and Conciliation Commission can be established to follow up on the cases. Since its founding in 1951, the procedure was set in motion 1,831 times (by December 2003), but until 1995 only six Fact-Finding and Conciliation Commissions were installed (Romano 1996, 18).

(36) I am not aware of systematic inter-linkage of the legal departments preparing the CEACR and the policy departments implementing technical cooperation. It seems that cases of conflict usually have to reach advanced stages of the supervision and enforcement procedure to be taken into account in technical cooperation projects. For the EU the situation seems to be similar. There are, for instance, no signs that support from the European Social Fund is dependent on the correct implementation of EU directives.

(37) If one wished to give policy recommendations, one could stress that it would be in the interest of the ILO to also (or even exclusively) convince Member States of ILO norms and values, as technical assistance is a scarce resource and the sustainability of projects is at stake.

(38) As in the EU package meetings, persuasion may play a decisive role here, too. But, as I have argued, it is empirically difficult to separate this logic from knowledge transfer and capacity-building measures.

(39) Valticos (1981, 481) states that of the 222 direct contacts that took place to discuss non-application (1970–1979), 115 had positive outcomes in the form of application improvements (52%). However, it seems that he is referring to direct contacts as well as to advisory missions and regional advisors.

(40) Whereas the workers’ group called for immediate action, Myanmar’s East Asian neighbours argued for delay. Finally, a compromise put forward by the employers’ group was adopted despite dissenting votes from China, India, Malaysia and Russia (for a brief description of these developments, see Elliott and Freeman 2003, 105).

(41) Two types of argument were used by IOs or Member States to justify the weak implementation of the sanctioning decision: (a) A specific national policy vis-à-vis Myanmar did not contribute to forced labour and therefore there was no need for a reaction; (b) Myanmar would gradually improve the situation and, even though the situation was still far from satisfying, it would give the wrong signal to start sanctions at that moment.

(42) Elliott (2000, 9) has rightly argued that this approach is comparable to the WTO approach, which leaves it to member governments ‘to determine within prescribed limits, the cost of enforcement they are willing to bear’, with the Security Council the only organ that can impose direct sanctions. All of these processes are substantially different from horizontal sanctions in inter-country conflicts (see Neyer and Wolf 2005, 50).

(43) Another case that is even more supportive of this argument is the attempt of workers’ representatives to lodge an Article 33 procedure against Columbia (following up on a case under the freedom of association procedure), which is blocked by the position taken by other Latin American countries supporting Columbia.

(44) The developments would correspond with an increasing number of representations and the introduction of naming and shaming in the Global Report and Annual Review.

(45) Although the assessment of the overall degree of delegation was not the focus of this analysis, the results generally seem to support the categorization of the EU as having high delegation and the ILO as having moderately high delegation (see Abbott et al. 2000, 406, Table 1 and below).


©2005 by Miriam Hartlapp
formated and tagged by K.H., 7.6.2005