Partnership on paper, pragmatism on the ground: the European Union’s engagement with civil society o

This article analyses the evolution of the European Union’s development policy in relation to civil society. Based on a review of overall policy trends, strategies and practices in Central Asia, it demonstrates how the EU’s development policy has gradually moved from a focus on European NGOs towards civil society organisations, broadly defined and increasingly associated with the private sector and local authorities. While the EU’s policy recognises the intrinsic value of civil society in all its diversity and promotes partnership, its operational practices show a pragmatic preference for working with professionalised organisations in service delivery roles. ARTICLE HISTORY Received 16 January 2020 Accepted 26 May 2020


Introduction
In the mid-1970s, the European Commission was among the first development cooperation providers to create a dedicated funding line for supporting development projects implemented by European NGOs. During the following four decades, the EU's support to and through NGOs was a key feature of its international development policy and grew in financial volume, as new member states joining the expanding Union brought their own NGOs into the mix. CONCORD, the Brussels-based platform representing more than 2,600 organisations from all EU member states and the United Kingdom, reflects the breadth of the EU's development policy engagement with NGOs.
December 2019 saw the introduction of a new team of European Commissioners for the period 2019-2024 under the leadership of Commission President von der Leyen. Her mission letter instructed the new Commissioner on International Partnerships to give priority to the EU's cooperation with civil society (Von der Leyen 2019, 5; emphasis in original): I want you to have a dedicated focus on supporting civil society around the world. As the experts on the ground, they often lead the way on sustainable development. We should ensure they have a far greater role in designing and implementing European policies, programmes and projects.
Besides committing to an increase in funding to civil society compared to the previous multi-annual financial period, the President's instruction indicates that the Commission seeks a greater operational role for civil society in designing and implementing EU development policy. 1 Given the involvement in terms of contributing to policy design and managing development projects using EU funding, this in practice implies a focus on civil society organisations (CSOs), which EU law defines as "an organisational structure whose members serve the general interest through a democratic process, and which plays the role of mediator between public authorities and citizens" (EUR-Lex n.d.) The mission letter, in contrast to this definition, emphasises an official and managerial role over a mediating one.
Important changes in the EU's wider external relations policy occurred during 2014-2019, including in its development policy. The EU's Global Strategy on Foreign and Security Policy called for reforms in the EU's development policy to become more flexible and aligned to the EU's interests (EU 2016). The strategy specifically committed to deepening "partnerships with civil society and the private sector as key players in a networked world." A subsequent EU development policy statement adopted in 2017 emphasised the need for an enabling environment for CSOs and outlined the various roles that CSOs play (EU 2016;2017).
Against this backdrop, this article analyses how the EU's development policy has evolved over time in relation to NGOs and CSOs. It observes a discrepancy between the EU's development policy, which recognises the intrinsic value of civil society in all its diversity, and its operational practices which indicate a more pragmatic orientation towards working with professionalised organisations in service delivery roles. A case study on patterns of cooperation in Central Asian states confirms that the EU's primary interest remains to work with CSOs as an aid delivery channel. The EU's sizeable development budget, relatively rigid financial management rules and the considerable disbursement pressure this creates are important factors explaining this discrepancy. Another key factor is the involvement of developing country authorities that, as illustrated by the Central Asian states, influence the conditions under which CSOs can operate officially and, linked to this, can gain access to EU funding. The resulting predominantly pragmatic and project delivery-oriented nature of the EU funding of CSOs may result in relevant and effective development interventions. Yet, the extent to which the nature of this funding creates unintended trade-offs in relation to the core mandates and societal functions of CSOs tends not to be scrutinised by the EUsince this was not the primary focus of the funding provided. 2 Investing in such broader evaluation inquiry would be key for the EU to manage expectations in relation to its own ambitious policy statements.
The article is structured as follows. It starts with an overview of literature on CSOs and NGOs in international development cooperation, followed by an analysis of how both emerged and feature in the EU's development policy. This general policy outlook is then contrasted with an analysis of the EU's engagement with CSOs in Central Asia. A final section presents concluding reflections and identifies opportunities for further research.

Civil society's NGOs
Development policy debates during the 1990s prominently featured the term civil society, which reflected broader government policy trends promoting a smaller role of the state and a stronger role for the market. Between market and state, a "civil society" would provide an associational space that would both strengthen as well as promote democracy (Lewis 2015). NGOs welcomed this stronger emphasis on civil society in development policy since it " … embedded the idea of NGOs within established political theory and provided theoretical justification for their role" (Lewis 2015, 4). Following these changes in development policy and funding, many NGOs began to increasingly self-identify as CSOs and members of civil society. 3 During the late 1990s and the first decade of the new millennium, NGOs gradually claimed a more prominent place in both the development cooperation narrative and in practice. During this period the European economy grew considerably, leading to corresponding increases in ODA budgets and grants to NGOs. The start of the global economic and financial crisis in 2008 did not counteract this trend. World Bank projects involving NGO participation or other CSOs increased from 21% in 1990 to almost 90% in 2015 (Brass et al. 2018). Whereas these and other statistics underline a view of NGOs as efficient development cooperation implementers, other scholarly contributions consider and analyse NGOs as "grassroots challengers to entrenched power structures" (Brass et al. 2018, 136). Yet it is not straightforward to be a grassroots challenger and reliable project implementer at the same time. External support to CSOs in relation to the first role can be framed as a "principled" approach, as it involves providing funding to CSOs to support them in advancing their mission and mandate. External support to the role of project implementer can be characterised as a "pragmatic" approach, as it entails providing funds to CSOs to implement development cooperation. The pragmatic approach views CSOs as a potential aid delivery "channel", with other such channels including multilateral organisations, national implementing agencies or the private sector.
The direction chosen by many NGOs to "professionalise" as development actors and scale up project management functions is generally regarded to having weakened their legitimacy and effectiveness in promoting social justice and transformation (Banks, Hulme, and Edwards 2015;Pearce 2000). Because of their ability to function as conduits (or "channels") from rich to poor communities, NGOs were criticised of being close to or even directly promoting the former's agendas (Eade 2008). In contrast to the EU's legal definition of CSOs, the legitimacy risk is illustrated by the EU development policy's more pragmatist emphasis on CSOs being a "partner" in promoting this policy (see Von der Leyen 2019).
The absence of a common (or sufficiently consensual) definition of NGOs and CSOs is not unusual in the broad field of development cooperation, given similar broad concepts such as partnership, ownership and capacity. Yet, research suggests that development practice has contributed to depoliticising the concepts of NGOs and CSOs as a wide range of actors that contribute to furthering official development policy outcomes, as opposed to their representative and "watchdog" roles as interfaces between state and society (see Banks, Hulme, and Edwards 2015;Pearce 2000). The following section analyses how the role and place of NGOs and CSOs in the EU's development policy has evolved in previous decades.

Four decades of EU development policy: from NGOs to CSOs
The European Commission's engagement with NGOs in the context of its development policy started in 1976 with the creation of a co-financing budget line, which provided financial assistance to European NGO initiatives (Bossuyt 2004;Carbone 2008). The Commission's official justification for proposing the budget line was the complementary nature of NGO projects and the Community's own development cooperation, as well as the positive attitudes of European NGOs towards the European project (EC 1975). 4 In 1979, the budget line was broadened from financing projects in developing countries to also promoting awareness of development cooperation among European citizens. In March that year, the European Commission presented its first report on relations with NGOs to the European Council, and declared the cooperation "fully operational" (EC 1979).
The main driver behind the creation of the budget line was the European Parliament, which at that time had only limited influence and formal scrutiny rights over EU development cooperation. The introduction of the budget line was followed by European NGOs creating a "NGO Liaison Committee", which in addition to federating and representing its European NGO members also maintained a development policy dialogue with the European institutions (Carbone 2008;Bossuyt 2004). 5 By 1998, the committee represented 15 national platforms with a collective membership of over 800 European NGOs (DAC 1998, 62).
In 2000, the publication of a European Development Policy Statement showed a European Commission at a crossroads. The inclusion of a dedicated sub-section on "working with civil society" in a chapter on partnerships expressed the Commission's policy preference of civil society over NGOs (EC 2000). This was noteworthy, since during the same year negotiations were concluded for the Cotonou Agreement, an international agreement between the EU and African, Caribbean and Pacific states. This agreement referred to Non-State Actors (NSAs) as one of the key actors of the partnership, representing a broad range of actors encompassing NGOs, trade unions as well as the private sector. European NGOs questioned the use of the term NSAs, since not only did it produce confusion as other EU policy areas continued to refer to CSOs, they also objected to be treated together with social partners and the private sector (Carbone 2008).
The Development Policy Statement was one of several policy documents produced around the turn of the millennium reflecting a shift from (European) NGOs to a broader focus on civil society. This shift followed international development policy trends, yet also reflected increased dissatisfaction in the Commission with the budget line's established focus on and provision of earmarked finance to European NGOs. Concerns over the limited development effectiveness of financed projects were raised, while both NGOs and the Commission had to invest considerable time respectively developing and appraising project proposals, many of which were not granted (Carbone 2008;Bossuyt 2004). A seminar co-organised by the EC and European NGOs in 2003 identified five key pressures on the co-financing budget line: (1) a growing number of proposals while budgetary resources saw no substantial increase since 1998; (2) new demands created by the EU's enlargement process and associated development awareness needs; (3) doubts about the overall identity and strategy of the budget line, as well as ultimately its impact; (4) the increasing use of programmebased approaches in EU development cooperation such as budget support, as well as the emergence of strong civil society organisations in developing countries; and (5) reforms of the EC's financial management overall. The fifth reform included the requirement of calls for proposals on all grants, stricter contractual rules but also the "deconcentration" of development cooperation management to the Commission's delegations abroad (Bossuyt 2004).
The revision of the conditions for co-financing sought to further emphasise partnership and capacity development between European NGOs and NGOs based in developing countries, while also seeking to promote longer-term cooperation relations (DAC 1998, 63). Yet, it also followed a period of contentious relations between the Commission and the NGO Liaison Committee, the latter perceiving the policy change as part of a hostile strategy due to their increased critical views towards the Commission. In 1999-2001, a Commission-initiated external audit observed cases of insufficiently documented or ineligible expenditure by the NGO liaison committee, which led to its orderly closure and the subsequent creation of the Confederation of European NGOs for Relief and Development (CONCORD) in 2003 (Carbone 2008;Kenety 2001). Although the budget line and the associated practice of demand-driven NGO project financing (i.e. EU financing of selected projects as proposed by the NGOs) was phased out, CONCORD was among those successfully calling for continued earmarking of parts of the EU's development cooperation budget to CSOs. CONCORD's membership today covers 28 national associations, 25 international networks and four associate members that together represent more than 2,600 NGOs (CONCORD n.d.).
In 2005 a joint policy statement on EU development policy was adopted by the member states, Commission and Parliament. This statement, the European Consensus on Development, did not contain a single reference to NSAs, concentrating instead on civil society. A few years later, under the leadership of a new Development Commissioner, the EU proposed an "Agenda for Change" that anticipated a further broadening understanding of civil society. The proposed policy agenda included commitments to increasing cooperation with "non-state actors and local authorities", as well as to "civil society and local authorities", the latter's narrative description also including social partners (EC 2011). A subsequent 2012 policy proposal titled "The roots of democracy and sustainable development: Europe's engagement with Civil Society in external relations" referred to civil society as including, among others, NGOs, foundations, research institutions and business associations (EC 2012). In effect, the Commission's policy statements simply relabelled the previousand contentious definition of NSAs as civil society.
Parallel to these policy changes, the EU also continued reforming its approaches to providing funding to CSOs, following the phasing out of the budget line. In 2008 the Commission created a Potential Applicant Data Online Registration (PADOR) system in 2008 to allow organisations interested in participating in calls for proposals to register and regularly update information about themselves. The information registered is used by the Commission to determine the eligibility of organisations for those calls in which they participate. 6 Such calls are then published in an online portal ("eCalls PROSPECT"). Registration at PADOR requires prior registration with the European Commission Authentication Service (ECAS), a service that allows the user to access a wide range of Commission information systems. The system is formally open to all, and also allows for "offline" registration when organisations that for technical reasons cannot register with PADOR wish to respond to calls for proposals. Nevertheless, the registration process may make it challenging for some CSOs to access EU development finance (see EC 2016).
In 2017, a successor policy statement to the 2005 European Consensus on Development again prominently acknowledged the "multiple roles that CSOs play as promoters of democracy and defenders of rightsholders and of the rule of law, social justice and human rights" (EC 2016, 8). It also lists civil society several times alongside other actors such as the private sector and multilaterals in having a role in promoting the 2030 agenda (i.e. implementing development projects), a contradiction that perhaps is unavoidable in a wordy 57-page statement formulated between Commission, Parliament and member states. However, while acknowledging the various roles that CSOs play, the EU is predominantly interested in CSOs' ability to act as project implementers. This is in no small part due to the aforementioned financial procedures and approaches with which the EU operates, as well as the considerable volume of development finance it manages.
In June 2018, the European Commission proposed merging several existing external financing instruments into a single instrument for its neighbourhood, international cooperation and development policies. The proposed legal text for the instrument includes the following definition of civil society in its preamble (EU 2018, 21): Civil society organisations should embrace a wide range of actors with different roles and mandates which includes all non-State, not-for-profit structures, non-partisan and non-violent, through which people organise to pursue shared objectives and ideals, whether political, cultural, social or economic. Operating from the local to the national, regional and international levels, they comprise urban and rural, formal and informal organisations.
While CSOs can be involved as development project implementers under the instrument as a whole, in addition a dedicated thematic programme of €1.5 billion was set aside for them. The European Parliament's position, which guides its negotiations with the co-legislating European Council, proposed to increase this amount to €2.39 billion (2018 prices) and increase the eligibility of the thematic programme to include "civil society and local authorities." By means of several substantive amendments, the Parliament proposes to additionally provide support for state actors at local levels to create an enabling environment for civil society, as well as to provide dedicated support to civil society and local authority networks. The broad definitions applied and proposed make it far from straightforward to appreciate the Commission leadership's desire to increase the involvement of CSOs in EU development cooperation (see Von der Leyen 2019), while also accentuating the contrast with the realities of the approaches to funding cooperation with them.
After more than two decades of structured cooperation between the European Commission and NGOs in the field of development policy, relations evolved considerably in the late 1990s. Successive EU policy changes saw a shift from a focus on NGOs to CSOs, interpreted broadly so as to include actors such as trade unions and business associations, with more recent policy statements linking civil society to local authorities. Yet, these broader concepts are typically presented as composites of different actors, including CSOs. Although policies generally allude to the need to strengthen civil society as an end in itself, the EU's evaluations to date by and large focus on these actors' performance in delivering EU financed interventions. Comparatively limited evidence is available on the EU's policy principles on supporting and working with civil society are translated into practice.

EU engagement with CSOs in Central Asia
To scrutinise to what extent these discursive developments have played out on the ground, we focus on the EU's engagement with CSOs, including NGOs, in Central Asia, where the EU has gradually increased its foreign aid since the 1990s. The EU is currently the second-largest aid donor in the region, with only China providing more financial assistance (Axyonova and Bossuyt 2016;Bossuyt 2019).
In their engagement with local civil society, Western donor organisations active in post-communist Central Asia in the 1990s mostly tended to support local NGOs. 7 Embedded within the transition paradigm that shaped Western assistance to the former communist countries at the time, Western governments began to offer grants and training to local NGOs in order to empower them in relation to the state (Babajanian, Freizer, and Stevens 2005, 211). More recently, international donors have started engaging with communal civil society in Central Asia, premised on a less politicised and more developmental take on the role of civil society. 8 While the former trend links civil society support to a broader democratisation agenda, whereby NGOs receive support targeted at empowering them vis-à-vis the state (the principled approach), the shift towards communal grassroot initiatives emphasises the role of CSOs as development agents which provide development-related services (the pragmatic approach). Recent studies have indicated that the EU's engagement with civil society in Central Asia has also shifted towards approaching local CSOs more as development cooperation agents rather than as agents of political change (Aydin 2018;Pierobon 2019).
While the EU provided technical and financial assistance to post-Soviet Central Asia in the 1990s, its engagement with Central Asian CSOs started a decade later, in the early 2000s. The EU has relied on three external financing instruments to engage with civil society in the region: the Institution Building and Partnership Programme (IBPP), the Non-State Actors and Local Authorities in Development (NSA/LA) programme, and the European Instrument for Democracy and Human Rights (EIDHR).
The IBPP was introduced in 2001 and financed through the EU's Technical Assistance for the Commonwealth of Independent States (TACIS) instrument. The programme sought to support an institution-building process through partnership cooperation of NGOs, local authorities and not-forprofit professional organisations in the post-Soviet countries, including Central Asian states, with their counterparts in the EU. The partnership concept implied that the respective partner organisations should have similar goals and should be faced with similar challenges, be it in different contexts. The relationship established through the project was expected to proceed beyond EU financial support. One specific aim was to improve the capacity of local civil society through small-scale projects establishing "direct contact to the local communities" (Giffen et al. 2007, 4). Between 2001 and 2007, over 40 projects were funded under the IBPP in Kyrgyzstan, Kazakhstan, Tajikistan and Uzbekistan, with grants ranging between €100,000 and €200,000. The programme ended in 2007, except for Uzbekistan, where it continued until 2015.
The NSA/LA was established as a follow-up programme of IBPP as part of the newly established Development Cooperation Instrument (DCI) in 2007. It operated from 2007 until 2014, when it was replaced by the Civil Society Organisations and Local Authorities (CSO-LA) programme. Both programmes represent a pragmatic approach to external support and seek to strengthen CSOs' capacities as service providers, especially for the poorest segments of society, as well as to foster their participation in policy-making processes. The focus is thus less on institution-building for post-socialist transition and more on poverty reduction and basic services provision in achieving the MDGs (and subsequently SDGs), in line with the goals set out in the European Consensus of Development (Pierobon 2019). Nevertheless, the programme also served a political purpose, as it sought to facilitate CSOs' participation in defining and implementing governmental measures for poverty reduction and government strategies for sustainable development.
The EIDHR concerns a dedicated legal instrument in the EU's budget for providing financial support to CSOs working on issues of human rights and democratic development. What distinguishes the instrument from the aforementioned ones is that it operates without the need for consent from the target countries' governments. Central Asian countries became eligible for EIDHR support in 2004, and have since received funds for regional (macro-) projects as well as country-specific micro-projects. Macro-projects are multi-country projects mostly implemented by large European NGOs or international organisations and are managed from the Commission's headquarters in Brussels, while micro-projects are country-specific projects and are managed by EU delegations in partner countries.
With respect to IBPP, the indicative programmes envisaged different types of CSOs to benefit from the programme. The indicative programme outlining the assistance for Central Asia under TACIS for 2005-2006, for instance, stated that the IBPP was supposed to offer "a comprehensive thematic approach towards civil society issues … through a direct dialogue and cooperation at the grassroots level with a wide range of partners: civil servants, professionals, young people, women, consumer groups, trade unions, farmers associations, regional or communal bodies" (European Commission 2004b, 35-36). However, the IBPP's calls for proposals and the conditions set for potential beneficiaries significantly narrowed the range of potential beneficiaries. According to the programme's operational procedures, only officially registered legal entities could submit their project proposals to be selected for IBPP grants, automatically excluding community-based initiatives and self-help groups that were not formally institutionalised. Moreover, the calls for proposals were open exclusively to partnerships between European and Central Asian non-state actors. The project proposals for IBPP came mainly from European CSOs that had previously established contacts with Central Asian counterparts, while local initiatives not involving cooperation with European partners were not considered.
IBPP beneficiaries included partnerships between NGOs, trade unions, local municipalities and educational and cultural institutions. However, the types of actors supported varied depending from country to country. In Tajikistan and Kyrgyzstan, IBPP funds were primarily granted to western-style NGOs, while in Uzbekistan funds often went to partnerships with pro-governmental or government-organised NGOs (GONGOs). In Kazakhstan, both state-independent and state-led types of CSOs were supported via the IBPP. In Uzbekistan, which only started allowing independent CSOs to operate after the death of President Karimov in 2017, the EU has faced the dilemma of either not engaging with local CSOs at all or supporting state-led organisations.
Similar to the IBPP programmatic documents, the NSA/LA Strategy Article for 2007-2010 specifies that "all types of non-profit making non-state actors and local authorities" originating from the EU and beneficiary countries are eligible for funding (EC 2006, 11). At the same time, NSA/LA officially aimed to enhance a participatory approach to development assistance, prioritising local actors' projects and "own initiatives" in the target countries (see European Parliament and Council of the European Union 2006). A particular focus was on the empowerment of "vulnerable" groups. However, the practice of the NSA/LA implementation in Central Asia revealed that some categories of local CSOs were excluded from this support, since under EU funding rules only officially registered organisations are eligible for EU funding. Among the excluded CSOs are informal groups in rural areas that are only in rare cases registered with official bodies, thus making them ineligible for NSA/LA grants.
While community-based rural development was a focal area for the NSA/LA in Kyrgyzstan and Tajikistan, projects were implemented mainly by foreign NGOs. Rural communities benefited from project activities, but only indirectly as they were not the principal implementers or receivers of funding. NSA/LA-funded projects that aimed at empowerment of vulnerable groups, such as youth, the elderly, people with disabilities and women workers, were also mostly implemented by foreign NGOs, although sometimes in partnership with local NGOs or formal community-based organisations. In a few cases they were implemented solely by officially registered local organisations. Although our analysis does not cover the implementation of projects under the CSO-LA programme, it appears that these tendencies can also be observed in Central Asia under this programme (Bossuyt 2018).
EIDHR funding is generally reserved for NGOs working on human rights and democratisation issues to which the countries' governments would be reluctant to consent. The allocation of funds to certain regions and countries reflects the level of political liberalisation of the respective states and needs of local civil society. For instance, the EIDHR programming for 2005-2006 asserts that "[t]he selection of regions and countries for a particular campaign is based, above all, on an assessment of the human rights and democratisation concerns in the region/countries and a deliberate choice of thematic priorities to be addressed" (European Commission 2004a, 20). For the macroproject allocations, the selection process "will seek to obtain a balance between implementing NGOs or other civil society organisations based in the European Union and those from the eligible region, or eligible countries within that region. The calls for proposals will encourage EU-based NGOs to establish partnerships with NGOs from the eligible region/countries and to submit joint project proposals" (European Commission 2004a, 33).
More recent EIDHR programme documents reiterate that EIDHR funding remains tailor-made to the specific context of the countries and continue to promote cooperation between EU and local NGOs (see European Commission 2014). However, by also supporting networking between local CSOs, the EIDHR instrument as a whole has now become more concerned with "reaching out to under-represented groups, disenfranchised CSOs and community-based organisations" (European Commission 2014, 6).
Nevertheless, so far, EIDHR country-specific funding to the Central Asian countries has been predominantly provided to western-style CSOs. Similar to the IBPP and NSA/LA programmes, from which it formally seeks to distinguish itself, the primary beneficiaries have been international or foreign NGOs and their local branches (e.g. Soros Foundation, Eurasia Foundation of Central Asia and Pro NGO) and well-established local organisations with a considerable history of foreign funding. While in some cases, local authorities and government-supported NGOs have been involved in dialogue or capacity-building activities as part of the funded projects, the EIDHR operational procedures and calls for proposals excluded them from directly applying for funding. EIDHR projects aiming at capacity building and training for civil societywith the majority implemented in Kyrgyzstanhave also largely targeted state-independent NGOs. In the case of Kazakhstan, several CSOs that are related to state structures or are largely dependent on the government's funding have also received EIDHR grants. As in the case of all other EU funding mechanisms, semi-official or unregistered local groups and community-based initiatives could not be considered for EIDHR grants.
The practice of EU engagement with CSOs in Central Asia has gradually shifted from a focus on principled support to the organisations' mandates in the first two decades after the fall of the Soviet Union, to a more pragmatic approach, whereby CSOs function as implementing agents for EU development projects. Direct beneficiaries of the EU funding mechanisms focusing on CSOs in Central Asia include mainly the "Western-style" type of NGOseither larger European NGOs and their partners, or local organisations that often receive Western funding. This is particularly the case for Kyrgyzstan and Tajikistan. In Kazakhstan and Uzbekistan, EU funding has also gone to partnerships between European NGOs and state-led organisations, mostly because these two countriesand especially Uzbekistanhave fewer state-independent CSOs. As Turkmenistan continues to be a very closed country, it has largely remained out of the remit of the EU's funding mechanisms for civil society support to the region.
It is worth highlighting that only formal, well-established organisations in Central Asia have received grants from the EU. No direct grant support has gone to communal civil society that is represented through informal community-based groups and self-help initiatives. While programme documents and calls for proposals explicitly call for projects that involve cooperation with community-based groups or support to community-based initiatives, operational procedures of the calls for proposals limit the circle of grant applicants to officially registered organisations. Community-based groups become eligible for the EU funding only after they have developed a certain internal structure and institutionalised their activities by registering as NGOs. Moreover, the organisations that have newly evolved from grassroot initiatives often lack the capacity to compete for EU funding with established CSOs, whose members have been previously trained in writing grant applications and have years of experience in acquiring foreign funding.

Concluding reflections
This article has shown how the EU's development policy has, over several decades, gradually moved from a focus on cooperating with European NGOs to a focus on CSOs, broadly defined and increasingly associated with the private sector and local authorities. Although overarching strategies continue to emphasise the importance of civil society in all its diversity and the need for an environment conducive to its functioning, the EU's operational practice shows a continued reliance on either international (European) NGOs or western-style local CSOs, since these have the professional systems and processes needed for managing EU funding. Put differently, the EU's policy towards working with CSOs is principled on paper, yet pragmatic in practice.
The Commission's new leadership has expressed a high ambition when it comes to working with civil society in the EU's development policy, which today is characterised by broadening objectives, many of which reflect an approach to furthering a more integrated EU foreign policy. This article's analysis of the current wide-ranging policy priorities and description of the EU's engagement in Central Asia underlines that the EU's policies recognise the intrinsic value of civil society in all its diversity, but that its funding practices reflect a priority for involving them in a service delivery role. The EU's sizeable development budget, the relatively rigid financial management rules and the considerable budget pressure this creates is an important factor explaining this discrepancy. Another key factor concerns developing country authorities that determine conditions under which CSOs can operate officially and, by virtue of formal registration, may access EU funding. The anticipated increase in cooperation to and through civil society and the ambition to involve them in planning decisions is expected to further increase these tendencies. Hence, further empirical analysis is needed to identify more systematically where and why the EU's principled strategic focus contradicts its pragmatic approach on the ground. Empirical research and independent evaluation is also needed to assess whether EU funding creates unintended trade-offs in terms of the CSOs' capacity to deliver on their core mandates and objectives, as well as to manage the EU's expectations of its own policy and ambitions in relation to CSOs.