Conditionality and the EU–ACP Partnership: A Misguided Approach to Development?

The Cotonou Partnership between the states of Africa, the Caribbean and the Pacific (ACP) and the European Union (EU) provides a case in which the human rights approach to development is being put into practice. This article uses the partnership to address broader questions regarding the effectiveness of the new approach to development. The EU–ACP partnership is innovative because it reflects the changing international consensus on development, but it is not clear if the norms used in the Cotonou Agreement have achieved the consensus needed to comprise any real shift in development policy. Moreover, it can be argued that what diminishes the efficacy of the human rights approach in the EU–ACP relationship is political conditionality as this mechanism may lead to interventions that are counterproductive to the establishment of stable democracy.

In the last few decades, the promotion of democracy and human rights has become one of the key debates shaping relations between the European Union (EU) and the states of Africa, the Caribbean and Pacific (ACP). The EU-ACP partnership reflects the existing global norms and has been moulded by the changing conceptions of development within the international community. The last 40 years have demonstrated a progression from a principally economic and commercial partnership to cooperation that addresses political issues: the EU was the first entity to explicitly mention human rights, democracy and the rule of law into its agreements with external partners. The Lome´IV agreement signed in 1989 between the EU and the ACP states was the first-ever multilateral agreement that included political conditionality. The successor to the Lome´Conventions, the Cotonou Partnership Agreement, signed on 23 June 2000, defines as the 'essential elements' 1 of the ACP-EU partnership respect for human rights, adherence to democratic principles and the rule of law.
The political dimension of the Cotonou Partnership Agreement aims at fostering a continuous political dialogue on these issues, as stipulated in Article 8 of Title II of the Agreement. If flagrant violations of the essential elements Andris Zimelis is a graduate student in political science at the University of Illinois at Chicago. 1 occur, a consultation procedure under Articles 96 and 97 of the Cotonou Agreement can be initiated, aimed at finding a common solution to the political difficulties encountered by one of the parties. In practice, Articles 96 and 97 provide a mechanism for a consultation procedure and possible aid suspension if serious breaches of human rights and democratic principles occur.
As human rights and democratic principles become a norm in international development, the EU has fully embraced these normative instruments to foster the growth of democracy, human rights and good governance. The conditionality mechanism, however, will only be effective if applied consistently and objectively by the EU. The question is whether the EU has applied the conditionality clause consistently.
The reliance on the essential elements, especially democracy, remains the key element of the partnership. By stating that these elements underpinned relations between ACP states and the EU and all provisions of the Cotonou Agreement, the EU made it clear that they were an integral part of the development concept. The potential problem however, is the lack of a consensus on the definitions of the essential and fundamental elements included in the Cotonou Agreement. For example, no single constitutional framework of democracy exists; legitimate democracies can and do take many forms. What concept of democracy should the EU promote and defend via political conditionality in the ACP states? Moreover, human rights and democracy principles should ideally align with the EU's and individual member-states' foreign policies. In reality, it is possible that the essential principles compete with other national and EUwide goals, thus rendering the conditionality clause less effective.
The purpose of this article is to assess the effectiveness of political conditionality by looking at the European Union's suspension of development cooperation with ACP states perceived to have violated the principles of human rights, respect for the rule of law, democratic principles and good governance as laid out in the ACP-EU Partnership Agreement. Given the reliance on ambiguous concepts and the diverging interests, I expect that the EU's application of the conditionality clause will be inconsistent and the potential for the EU's conditionality policies will remain limited. This article will help to discover how effectively the policy of aid suspension works in changing the processes considered in violation of the essential or fundamental elements in the recipient state. As an ancillary benefit of this research, we will observe what factors have contributed to the EU's suspension of development cooperation, which may help us discern some patterns to the organisation's decisions to invoke Article 96 or 97.
This article fills a gap in the development literature by studying the consistency and effectiveness of conditionality aid. It also has straightforward policy implications, because the successful development of a large segment of the world depends on the effectiveness of the new approach. Prior research has been scant on the practical application and effects of conditionality. Therefore, a better understanding of the application and implications of the suspension clause (Articles 96 and 97) is needed for both policymakers as well as social scientists.
The next sections will review the history of EU development policy before the Cotonou Agreement, and the political conditionality in the context of the Cotonou Agreement before moving on to the case studies. The final section outlines the findings and offers concluding remarks.

Historical Overview of the EU Development Policy
The concept of development has undergone a profound change. What today may appear as a totally antiquated model may in the past have been an approach fully responsive to the highest demands of the field. One has to be prudent therefore, in assessing whether the whole process has been efficient in achieving the ultimate goal since this goal has shifted continuously and is subject to ongoing redefinition (Hilpold 2002, 54). For decades development has been seen mainly as economic growth. The assumption was that growth induced by industrialisation and investment in the centres of developing states would trickle down to the poor. Development is now understood more as sustainable human development, addressing people in relation to resource management and participation.
The beginning of the EU-ACP partnership can be found in a speech on 9 May 1950, when Robert Schuman declared that 'Europe would, with increased resources, be able to pursue one of its essential tasks -the development of the African continent' (Europa, Gateway to European Union 2003), and the Treaty of Rome signed on 25 March 1957 set up a provision for the association of the overseas countries and territories with the embryonic European Economic Community (EEC). Since then, through two Yaounde´Conventions and four Lome´Conventions, a new EU-ACP partnership has been negotiated in Cotonou. The previous partnerships between the EU and ACP states did not seem to function well, which is why the Cotonou Agreement became necessary. By looking at what Cotonou does differently and why, we can ascertain why the old partnerships failed and how the new approach is rectifying the 'old' shortcomings.
The first comprehensive agreement between the EU and ACP states -LomeĆ onvention I -was signed in 1975, providing for the association of 46 ACP states with nine European Union member-states. In its summary review of past development policy, the European Commission (EC) concluded that the principles of partnership, contractuality, predictability and security outlined by Lome´produced an innovative and unrivalled development framework. Under this agreement, each state had the right to determine its own policies and nonreciprocal trade preferences were established. Agricultural preferences were limited to exports from the ACP to the EU because of the Common Agricultural Policy's protection for European producers.
When Lome´I was negotiated in the 1970s, conditionality was not included in the scheme. In 1975, member-states did not see the need to provide for the possibility of termination or suspension of financial flows. Furthermore, ACP states were unwilling to endow the EEC with the right to unilaterally revoke aid at their discretion, considering it a violation of their sovereignty. The Uganda crisis in 1977, notorious for its human rights abuses, proved that such a position was untenable. The gross human rights violations committed by Idi Amin were certainly grounds for the suspension of EEC aid, but Lome´I was ill-equipped to deal with this type of situation (Hilpold 2002, 57).
The Uganda situation demonstrated that the EEC needed a formalised structure for dealing with humanitarian crises, but Lome´II was unable to rectify the problem. The attempt to protect human rights through conditionality was met with strong resistance from ACP states. The pledge to protect Western-style human rights was perceived as too costly, and in opposition to the political priorities of ruling elites.
Several other factors precluded EEC member-states from placing pressure on the ACP states about this issue. A stark ideological divide over the definition of human rights existed between East and West. Western criticisms of human rights situations often sparked retorts from the opposing ideological camp. It was clear to the EEC members that pushing this issue would jeopardise strategic and economic ties (Hilpold 2002, 58). In addition, the relationship between development, democracy, and human rights had not yet been explored (Hamm 2001(Hamm , 1009. Both sides were somewhat relieved that the issue of conditionality had been abandoned, but it became clear that this was only temporary.
As a result, human rights were raised at the Lome´III Convention . The issue, however, was approached in an apprehensive manner, and differing opinions existed on how conditionality should be shaped. Human rights were only mentioned explicitly in Annex I, which attributed a high priority to human rights. In effect, human rights were not a high-priority issue at the time. The second paragraph of this Annex identifies the full enjoyment of economic, social and cultural rights through development as essential to the dignity of people. 2 The discussion of human rights in Lome´III was far too vague to have an immediate impact on the human rights situation in ACP states, but the member-states obviously acknowledged the evolving nature of the arrangement. The EU-ACP relationship was not simply an economic partnership; instead, Lome´III established the need for member-states to achieve a consensus on the role of non-economic issues such as human rights and good governance.
Lome´IV marked an important transition from the previous three arrangements. The Agreement was signed in 1989 between 68 ACP states and 12 European states. For the first time Lome´policy became genuinely political, with the inclusion of human rights as one of the fundamental clauses of the Agreement. Not only did Lome´IV provide preferential trade access for ACP products, it also established the commitment to human rights, democracy and the rule of law as one of the major objectives of ACP-EC development cooperation (Martenczuk 2000, 463). Many small adjustments were agreed upon, essentially refining the provisions mentioned in the previous Lome´agreements. Two substantial changes, however, differentiated Lome´IV from the predecessor agreements.
First, a suspension clause was introduced to remedy the problems of human rights violations. Procedures were established to deal with cases of violation of Lome´'s 'essential elements'. Under the 'Mid-Term Review' (MTR) of Lome´IV, these three essential elements were incorporated in Article 5, paragraph 1: Respect for human rights, democratic principles and the rule of law, which underpins relations between the ACP States and the Community and all provisions of the Convention, and governs the domestic and international policies of the Contracting Parties, shall constitute an essential element of this convention (in Arts 1997, 77).
The MTR also included a consultation clause to deal with violations of these three essential elements, which was a concession to the ACP states. If one party 2 Annex I, paragraph 2 of Lome´III reads: 'The Contracting Parties proclaim that ACP-EU cooperation must help eliminate the obstacles preventing individuals and peoples from actually enjoying in full their economic, social, and cultural rights and that this must be achieved through the development which is essential to their dignity, their well-being and their self-fulfilment'(ACP Group 1984). determined that another failed to fulfil the obligations referred to in Article 5, Article 366 a(2) created a consultation period that lasted no longer than 45 days after the initial complaint. If following consultation no solution was found, the party concerned was subject to full or partial suspension of the benefits established by the Convention (Arts 2000, 192). Article 366 a(2) established a specific timeframe to deal with consultations before cooperation or suspension, but it did not address which measures should be taken to resume cooperation. In the framework established by Lome´IV, the European Community had complete discretion over when measures were lifted (Arts 2000, 193). This left ACP states powerless, and endowed Europe with a considerable amount of room for manoeuvre.
In addition to the suspension clause, Lome´IV also delivered important procedural changes regarding the distribution of aid. Under the first three agreements, the EC allocated a certain amount of aid to each ACP state at the beginning of each new Lome´Convention. This money was spent on the implementation of the National Indicative Programme (NIP), which established the important sectors for Lome´support in a particular state. Until Lome´IV, the financial aid was initially determined regardless of the program's effect on performance. Following the MRT, the amount of money needed to implement the NIP was split into two allocations. Seventy per cent of the funds were distributed initially, and the remaining 30 per cent were only distributed following a three-year evaluation of the program (Arts 2000, 131). These types of adjustments clearly demonstrate the nature of the ACP-EU relationship was changing fundamentally. Unlike the initial focus on unilateral aid giving, LomeÍ V marked the transition towards a development scheme that incorporated political factors and an increased adoption of political conditionality and the decline of preferential trade relations.

EU Development Policy: Toward Political and Economic Conditionality
The Lome´Conventions symbolically acknowledged the importance of human rights in development policy, but the actual performance of ACP states in terms of economic, social and political development did not improve significantly. As a result, the status of Lome´was challenged. The European Commission's 1997 Green Paper on the future of Lome´identified supply-side inadequacies including the absence of sound microeconomic and macroeconomic policies and good governance as the main factors precluding ACP economic and political development (European Commission 1997). Signed in June 2000, the new Cotonou Partnership Agreement reflected the transition from purely economic cooperation to more inclusive political agreements in development policy. The broad objectives of this Agreement were defined in Article 1: 'To promote and expedite the economic, cultural, social, and social development of ACP states, with a view to contributing to peace and security and to promoting a stable and democratic political environment' (in Holland 2003, 164). Table 1 demonstrates the EU's transition toward economic and political conditionality in EU Development Policy.
In order to dismiss the idea that EU-ACP relations are merely a relic of colonialism, Article 2 of the Agreement outlined four 'fundamental principles' to govern relations: 1. Stress on the equality of the partnership and local ownership of the development attempts to avoid the criticisms of paternalism that were waged against Lome´. The text states, 'the ACP states all determine the development strategies for their economies and societies in all sovereignty'. 2. 'Dialogue' between ACP states and the EU was a crucial aspect of relations. 3. The new partnership also tried to broaden involvement in economic and political life by explicitly incorporating civil society and the private sector. 4. In order to acknowledge varying levels of development, the concept of 'differentiation' distinguished between the least-developed states and more competitive economies in the ACP region (European Commission 2000, 2).
Most importantly, however, Article 9 of the Agreement also incorporated the three 'essential elements' that were previously outlined in Lome´IV: respect for human rights, democratic principles and the rule of law. Breaches of these three elements could lead to suspension of the Agreement, and Cotonou firmly entrenched conditionality. Holland (2003) notes that the new Agreement also addressed supply-side constraints by incorporating 'good governance' as a 'fundamental element'. Unlike the three essential elements, good governance was not simply duplicated from the previous Lome´agreement. The ACP states and EU members had difficulty in defining good governance, but the text defines it as, 'the transparent and accountable management of human, natural, economic, and financial resources for the purpose of equitable and sustainable development' (European Commission 2000, Article 9.3). Serious cases of corruption, including bribery, were grounds for suspending cooperation, according to Article 97. The Cotonou Agreement included good governance because democratic institutions guarantee stable and continuous participation in civil society, while simultaneously discouraging dependence on paternalistic goodwill. The Commission's concern over supply-side constraints in the 1997 Green Paper (European Commission 1997) mirrored the concerns of the World Bank about the lack of strong and accountable government agencies in developing states. The Agreement aimed to reduce corruption, increase political freedom, promote free and fair elections, and eradicate human rights abuses. In the words of Der-Chin Horng (2003, 695), the 'EU has successfully extended its European Idea of human rights to international agreements and has developed its external relations based on human rights'. The conditionality clause purported to help achieve these goals.
The Cotonou Agreement contains a revision clause (Article 95) that foresees that the Agreement can be adapted every five years (with the exception of the economic and trade provisions, for which there is a special review procedure). In accordance with Article 95, at the end of February 2004, the EC and ACP partners noted the provisions of the Agreement they wished to amend. Negotiations began at the ACP-EC Council of Ministers in Gaborone in May 2004 and were concluded at a ministerial meeting in Brussels on 23 February 2005. The revised Agreement was signed in Luxembourg on 25 June 2005. The revision process did not question the fundamental axis of the Agreement. The main objectives were to enhance the effectiveness and quality of the ACP-EC partnership, to ensure its consistency with the new international agenda for development, to deepen the political dimension of the partnership and to broaden the cooperation framework in order to include new important security and anti-terrorism issues. Similarly, the ongoing 2010 review of the Agreement emphasised the same aspects and made minor changes in the Agreement's language (EU Development Agency 2011).

Conditionality and the Cotonou Agreement
Conditionality is commonly defined as 'a mutual arrangement by which a government takes, or promises to take, certain policy actions, in support of which an international financial institution or other agency will provide specified amounts of financial assistance' (Killick 1998, 6). Thus, aid conditionality represents an attempt to use aid as an incentive for developing states to reform their policies and institutions. Generally, conditionality has been deemed inefficient in attaining its desired objectives and favourably changing economic and political developments. Craig Burnside and David Dollar (1997) discovered no relationship between aid flows and policy reform. This conclusion is complemented by a similar study that demonstrates that aid does not promote positive reforms and that the conditionality attached to international loans does not produce policy change (Devarajan, Dollar and Holgren 2001).
Since the 1990s, the European Union has included a human rights clause in all agreements with developing states (Holland 2002). As mentioned before, Article 9 of the Cotonou Agreement incorporates the three 'essential elements' of Lome´IV: respect for human rights, democratic principles and the rule of law, and a new 'fundamental' element -good governance. Breaches of the three essential or the fundamental elements may lead to suspension of the agreement under Articles 96 and 97. Article 9(1) of the Agreement lists democratic principles based on the rule of law together with a regard for human rights and good governance. 3 If ACP states fail to fulfil an obligation based on the definition of democratic principles, the EU can begin the process of inquiry and suspension. 4 The Convention provides for consultation in the event of a serious breach of the terms of the Agreement, enabling the EU to respond to regressions or interruptions of the democratisation process, persistent violations of human rights, and endemic corruption. A party accused of violating the founding principles of the convention is invited to consult with the Commission. 5 Article 96 also attempts to define what constitutes a special urgency, which it refers to as 'exceptional cases of particularly serious and flagrant violation of one of the essential' political conditions. It also attempts to define 'appropriate measures' which must be taken in accordance with the principles of international law and proportional to the violation. Parties according to Article 96(2)(c) are also expected to notify the party against whom the measures are being taken, and the Council of Ministers. It appears, however, as if the language of the Agreement remains ambiguous, giving some leeway to the EU. After all, parties will have different definitions of what constitutes a situation of special measures and are likely to disagree over matters such as what constitutes 'special measures'. The principle reasons for consultations under Article 96 are coups d'e´tat, flawed or non-transparent electoral processes, and violations of democratic principles, the rule of law, respect for human rights and fundamental freedoms. The principal reason for consultations under Article 97 (good governance) is corruption. These articles are designed to be invoked after political dialogue and other measures have failed.
Martin Holland (2002) lists 15 cases when the EU has imposed sanctions based on violations of democracy, human rights, the rule of law and good governance. The list covers only the time period from 1990 to 1999 and includes states outside of the EU-ACP agreement. The highlighted states are the cases that occurred after 1999 that are added to Holland's table. Table 2 displays the cases when the EU has been involved in consultations under Article 96 and where the consultation and suspension mechanisms of the Cotonou Agreement 3 According to Article 9(2) of the Agreement: 'Democratic principles are universally recognized principles underpinning the organization of the Country to ensure the legitimacy of its authority, the legality of its actions reflected in its constitutional, legislative and regulatory system, and the existence of participatory mechanisms. On the basis of universally recognized principles, each country develops its democratic culture.' 4 According to Article 96 (2)a, consultations are to begin no later than 15 days after the request is made by either party, shall continue for a period of time established by mutual agreement and shall not last longer than 60 days. The revised Cotonou Agreement added that consultation would begin no later than 30 days after the invitation to discuss the breach. have been used. It also illustrates the reasons for consultations or sanctions. According to the European Centre for Development Policy Management, since 1996, the ACP states listed in Table 2 have been involved in consultations with the EU under Article 366 of the Lome´Convention and Article 96 of the Cotonou Agreement (2011). In some cases the EU suspended aid during consultations (in the case of Fiji, for example), and in some the ACP state promised to remedy the situation to avoid sanctions (for example, Guinea-Bissau). EU sanctions currently apply to the following ACP states: Republic of Togo, Republic of Zimbabwe, Republic of Guinea, Madagascar, Niger and Fiji.

Research Cases
As mentioned earlier, the primary goal of conditionality is to provide incentives for reforms and to rectify violations in the recipient state. The case selection is crucial in this study because we need to control for various extraneous factors that may affect the outcome of aid suspension in the ACP states. For a complete analysis we need to look at all cases of invocation of Articles 96 and 97 regardless of whether aid was suspended or only negotiations occurred in order to assess the effect of conditionality. What made those states comply with the EU's demands? Perhaps the EU has other motives not to impose sanctions. Why do some consultations have positive outcomes and some negative?
The states are chosen based on the general idea of the positive method of agreement proposed by John Stuart Mill. That is, the states all have a common outcome or dependent variable -invocation of Article 96 or 97. The strength of this method is that: 'Its true value is in its function to eliminate alternative explanations . . . no factor can explain an outcome satisfactorily that is not common to all occurrences of that outcome' (Savolainen 1994(Savolainen , 1218. This method helps us discover the common independent variables across all cases. An underlying hypothesis is that the EU initiates consultation procedures in cases of flagrant violations of at least one of the essential or fundamental elements, particularly in cases of a coup d'e´tat. Thus, we expect to find that the chosen states have breached at least one essential or fundamental element. For analytical purposes we also look at states that have in common the invocation of Article 96 or 97 and aid suspension. This helps us isolate the factors that are present in the cases that led to aid suspension and the ones that did not. Finally, to obtain a more complete comparative analysis of the efficacy of conditionality, further research should also explore the states of Eastern and Central Europe (for example Croatia and Turkey) in order to see how these states implemented the changes stipulated by the EU.
The sources of the research predominantly come from official EU information sources such as the EU Directorate General for Development, EU-ACP Courier, European Parliament, European Council, and European Commission. This information is likely to have an EU bias, which is why information gathered from the sources offered by the ACP states' secretariat will be useful as well.

Case Studies: Applications
The success or failure of a consultation procedure under Article 96 or 97 of the Cotonou Agreement is difficult to evaluate. Each case is judged on its own merits, according to the conditions of the state and the circumstances of the consultation procedure. We can, however, tentatively assess the effectiveness of EU conditionality by analysing the invocation of Articles 96 and 97 and the ensuing implications. For the current purposes, a brief look at three coterminous West African states against which the EU invoked Articles 96 and 97 will suffice to provide an idea as to how effective conditionality is. The states are chosen because of their geographic proximity and history of aid suspension.

Republic of Coˆte d'Ivoire (Ivory Coast)
In December 1999, long-time President Kona Bedie was deposed by a military coup led by General Robert Gue´i; the suspension of the constitution followed. A National Committee of Public Safety was established to restore the authority of the state and the new 'leader' officially stated that: 'Democracy and the rules of democracy will be scrupulously respected and I will personally ensure that they are respected' (BBC News 2002). In response, the EU Parliament called on the Council and the Commission to review cooperation with Ivory Coast and, 'at all events, to open immediate consultations with its government, according to the procedure pursuant to Article 96 of the Cotonou Partnership Agreement, in order to agree a timetable for the return to democracy' (EU Bulletin 2000).
Because the Ivorian de facto authorities pledged to restore democracy and agreed to a timetable to presidential, legislative and local elections by 31 October 2000, the EU decided not to suspend cooperation but to adopt 'appropriate' steps. The Commission would monitor compliance with the electoral timetable and the adoption of measures to guarantee the impartiality and credibility of the elections. Both parties emphasised the role of consultations as a means to remedy the problems and the new government seemed to be interested in cooperating with the EU.
Consequently, the EU continued giving aid for existing programs but adopted a conditional approach for new projects, on the basis of support for the restoration of constitutional democracy. The EU demanded a national, multi-party dialogue, national reconciliation and legal proceedings against those accused of human rights abuses, and stipulated that the resumption of full cooperation would depend on progress achieved in these areas. The EU normalised cooperation and discontinued the consultation process in February 2002 following what it deemed free and fair elections.
In 2004, however, the Commission expressed its discontent with the processes occurring in Ivory Coast and again instigated consultations with the main concern being human rights. Thereafter, progress in the consultation process has been negligible but the EU has not imposed sanctions on Ivory Coast. In September 2002 a troop mutiny escalated into a full-scale rebellion, voicing the ongoing discontent of northern Muslims who felt they were discriminated against in Ivorian politics. The situation did not improve much and remained extremely tense as the EU recognised that 'the persistent situation of conflict makes it impossible to implement a traditional program of co-operation' (ACP- EU Courier 2000, 17). A period of French bilateral aid activity followed, while EU assistance remained de facto suspended and other EU states steadily reduced their bilateral programs. The volatile political situation and bloody fighting between the forces of internationally recognised President Ouattara and the incumbent president Gbagbo in April 2011 prompted the EU to reaffirm its decision to authorise the countermeasures on the grounds that Ivory Coast has violated its obligations under the Article 96 of the Cotonou Agreement.

Guinea
In July 2004, the Commission opened consultations under Article 96 with Guinea because of its worries about the process of democratisation: Commission considers that the gradual deterioration in the democratic environment in Guinea, notably the dubious referendum of November 2001, the undemocratic parliamentary elections in June 2002, and the lack of positive signs of imminent change in the situation amount to non-respect of the essential elements set out in Article 9 of the Cotonou Agreement (EUR-Lex 2003a).
In response, the government of Guinea pledged to: . return to democracy through resumption of dialogue with the traditional opposition and civil society, including revision of electoral arrangements; . hold parliamentary elections based on new electoral arrangements in June 2007; . uphold the Constitution and the law, so guaranteeing respect for human rights and fundamental freedoms, including the rights of political parties to organise, meet, demonstrate and speak in public; . launching discussion on a legal framework for liberalisation of the airwaves; and . enhance macroeconomic management and implementing sectoral reforms, promoting decentralisation.
Thus, the Commission perceived violations of the essential elements of Article 9 of the Cotonou Agreement and opened a dialogue with the Guinean government. According to the Commission, the Guinean authorities 'showed great willingness both to continue and step up talks and to facilitate the EU mission' (EUR-Lex 2003a). In April 2005, however, the Commission found that: 'Important measures concerning essential elements of the Cotonou Agreement have not yet been taken [by the government of Guinea] and the Commission concludes the dialogue with Guinea and adopts specific measures adopted as appropriate measures within the meaning of Article 96(2)(c) of the Cotonou Agreement' (EUR-Lex 2003a). The appropriate measures include: -Programs to strengthen civil society (including non-organised forms), respect for and reinforcement of democracy, human rights and good governance and the emergence or consolidation of free media may be supported. -Humanitarian operations, trade cooperation and trade-linked preferences will be continued. This decision was based on the fact that on a number of occasions, since its arrival on power in 1997, the Government had acted in ways that amounted to a failure to fulfil its obligations stemming from the essential elements of Article 9 of the Agreement; that it had also acted in violation of good governance requirements as serious corruption is made possible by the lack of transparency of the management of public utilities and resources (EUR-Lex 2003b).
The Commission spelled out the exact demands for Liberia: -ensure the independence of an efficient and effective judiciary; -conduct an external independent audit of government financial institutions and parastatal agencies; -effectively guarantee the personal security and freedom of movement of opposition leaders in Liberia; -establish an independent and efficient human rights commission; -establish an independent and efficient reconciliation commission, in charge of organising and supervising a reconciliation forum; -implement decisions taken to enlarge access to short-wave broadcasting; -establish an independent and efficient election commission; CONDITIONALITY AND THE EU-ACP PARTNERSHIP 401 -dismantle the monopoly on fuel import; -constant promotion and guarantee of freedom of press; and -enhanced transparency in the way public concessions and licences are run and on the fiscal revenue derived therefrom.
In practice, the EU would review the situation every six months and offer electoral assistance, and there would be no impact on the implementation of humanitarian projects, contributions to regional projects, humanitarian operations, trade cooperation and trade-related preferences. The Eighth European Development Fund (EDF) NIP, however, was divided into two instalments: a first instalment would cover institution building and direct support to populations, and a second one more structured aid (EUR-Lex 2003b). The implementation of the first instalment was linked to actual progress made in restoring efficient democratic structures and in improving public financial management. The implementation of the second instalment was conditional on the holding of elections to international standards in 2003 in a wider context of improvement of the political and governance situation. The EU also agreed to an arms embargo and prohibited the provision of technical training connected with armaments, imposed a visa ban, and a ban on importing uncut diamonds from Liberia.
In 2003, the EU decided to provide financial support to a peace-keeping operation in Liberia and to make funds available for other measures accompanying the peace process. At the same time, the EU announced that 'the balances remaining for Liberia from the 8th EDF shall be available forthwith for implementation and notification of the 9th EDF allocation will be made once the comprehensive peace agreement is in force and the signatory parties to the agreement have shown commitment to implement the agreement as foreseen' (EUR-Lex 2006a). In 2004, the EU announced: 'The current conditions in Liberia do not yet ensure respect for democratic principles, governance and the rule of law' and, therefore, the decision was made to extend the validity period of the measures provided for in 2003 (EUR-Lex 2004). These measures expired in 2006 and the EU lifted the 'appropriate measures' adopted under Article 96 because of the commitment of the Liberian government to 'bring about social, economic and political reconstruction in the country' (EUR-Lex 2006b). Some restrictions on military assistance and weapons transfers to Liberia were imposed in 2006, 2008 and 2010 but they were not part of Article 96 or 97.

Conclusion
On the basis of these case studies, we draw several tentative conclusions about the effectiveness of EU conditionality. The main finding is that the suspension of aid is not an effective tool for promoting or restoring breaches of the 'essential' or 'fundamental' elements in ACP states. What may hinder positive change in ACP states experiencing EU aid suspension is the inflexibility of the approach. The suspension of aid to Guinea was to expire only in 2008 despite the cyclical review periods. The reviews did not appear to be as useful as the EU claims given that the decision to suspend aid until 2008 remained in place. Likewise, the suspension of aid to Liberia expired only in the summer of 2006, despite the situation in Liberia improving dramatically. This is probably a time when the EU's aid could have been used to rebuild Liberia. Instead, the aid suspension adversely affected ordinary Liberians. The suspension of aid denied funds to development projects such as road construction and water treatment plants which, in the end, affect ordinary citizens (ACP- EU Courier 2000). One could postulate that success or failure of suspending aid depends on whether some ACP states are more dependent on aid than others. For example, the EU is likely to have more leverage over a state that does not possess national resources. A state like Nigeria or Botswana that has a great amount of natural resources might resist EU demands more fiercely. Moreover, authoritarian or semi-authoritarian regimes such as Guinea may resist any stipulations given by outsiders. The EU admits that no breakthroughs have been detected in Guinea (EUR-Lex 2011). The aid suspension and pressure from the EU have not induced noticeable positive changes in this state and can be regarded as ineffective. Table 2 shows that most states had two common independent variables -a breach of the rule of law and democratic principles -that led to the initiation of dialogue and aid suspension by the EU. The inherent ambiguity in defining 'democratic principles' can contribute to the confusion and inconsistent application of the conditionality clause. The EU's emphasis on democratic elections as a manifestation of democratic progress is also problematic. Elections do not necessarily lead to democratic principles in practice. Does the EU count every election as a sign of progress? If not, what are the defining characteristics of 'democratic' elections? It seems that vague definitions of what constitutes 'democracy and human rights' allow for the selective application of Articles 96 and 97. Clear measures of democratic progress would help ensure a more coherent application of the suspension clause of the Cotonou Agreement.
Besides democratic principles and the rule of law, each state had different 'problems' in the eyes of the EU. Interestingly, Liberia is the only state against which the breach of the 'fundamental' element -good governance -has been invoked.
Coups are the most egregious basis for aid suspension, but other criteria for initiating negotiations remain opaque. One cannot help but notice that consultation procedures are initiated for rather vague reasons and, in some cases, may not reflect the real situation. Thus, aid suspension via Articles 96 and 97 officially still applies to Guinea but not to Ivory Coast and Liberia. Ivory Coast is not subject to official sanctions, but its situation can hardly be described as optimistic, and the unfavourable situation in that state even precludes a coherent dialogue with the EU. How exactly and why a state is considered as having breached the 'essential' or 'fundamental' elements remains to be specified. Why suspend aid to Guinea and Liberia (until 2006) and not to Ivory Coast given that human rights, democratic principles and the rule of law are not better observed in the latter? According to Amnesty International (2010), the human rights situation is worse in Ivory Coast than in Guinea, but only Guinea is subject to the appropriate measures.
According to Transparency , Ivory Coast is more corrupt than Liberia. Again, what motivated the EU to suspend aid to Liberia on the grounds of corruption but not Ivory Coast? The usefulness of conditionality is also unclear because the process is slow due to 'the EU undertaking rounds of formal consultations with the offending regime and they [sanctions] have generally been removed before anything resembling genuine democracy is (re-)installed' (Youngs 2010, 6).
The new Cotonou Agreement provides for consultation in the event of a serious breach of the terms of the Agreement, but the consultative procedure between EU and ACP states is still vague and the appropriate measures are at the discretion of the EU. It is difficult to conduct political dialogue and apply the suspension clause coherently and systematically. This can generate misunderstandings and friction between the EU and its partners and within the EU itself. For example, the United Kingdom made Sierra Leone a foreignpolicy priority and requested a suspension of aid to Liberia on the grounds that the Liberian government perpetuated war in Sierra Leone (ACP- EU Courier 2000). Unsurprisingly, this caused friction among the EU members. It is reasonable to presume that former colonial ties between EU members and the ACP states in part determine the severity of sanctions (Crawford 2001;Youngs 2010). External factors may also hinder the effectiveness of aid suspension. For example, China's 'no strings attached' aid to some ACP states may render EU aid suspension ineffective.
Lack of consensus on essential elements can also hinder the effective application of the conditionality clause. For example, definitions of democracy are varied. Despite the waves of democratisation across the world in recent years, there are different views on the means and structures required to achieve democracy, let alone on the general applicability of such means and structures. What concept of democracy is the EU promoting via political conditionality in the ACP states? Merely having a 'democratic' constitution does not guarantee adherence to its principles.
Another example is good governance. Again, how exactly is the EU defining 'good governance'? The difficulty is in deciding how 'good' is good enough. One could easily argue that the EU's own standard of good governance is far from perfect given the findings of corruption that led to the resignation of the Commission in 1999, or the election of extreme-right coalition governments. 'Third World' states can easily identify flaws in European 'good governance'. Unless both the 'essential' and 'fundamental' principles are unequivocally defined, respected and promoted, the quality and purpose of political dialogue will be marginalised and regarded as little more than cosmetic conditionality to be applied in a selective manner.
Finally, the EU's double standards undermine cooperation with the ACP states. The EU condemned human rights abuses in Guinea and imposed sanctions under Article 96 but still concluded a fishing agreement with that state (ACP-EU Joint Parliamentary Assembly 2004). Sectoral agreements like fishing or steel are beyond the purview of the Cotonou Agreement, and do not in theory contradict the Agreement. One must wonder why the EU chooses to contradict its own moral stance and undermine the efficacy of conditionality. Suspending aid to a state, while at the same time concluding a trade agreement with that state, will not make the aid suspension effective. The EU needs to clarify the vague concepts found (e.g. 'democratic principles') in the Cotonou Agreement and muster the political determination free of hypocrisy to successfully tackle underdevelopment and conflict in the ACP states.