EU Rules Beyond its Borders: The Policy‐Specific Effects of Transgovernmental Networks and EU Agencies in the European Neighbourhood

The European Union (EU) is increasingly involved in its rule advancement outside its borders through a dense net of transgovernmental networks. However, we know little about the effects of these networks. This article assesses the impact of transgovernmental networks across various policy domains in the European Neighbourhood Policy (ENP) region. Building on a novel longitudinal dataset, we demonstrate that the effects of transgovernmental networks vary across policy sectors. Policies marked by a higher degree of mutual interdependence exhibit greater positive change as a result of denser networks. Meanwhile, the involvement of EU agencies stimulates more intense technical co&#8208;operation and broadens the range of policy areas covered. In doing so, EU agencies serve as bridges for the establishment of strong links between the EU and its neighbours.


Introduction
Transgovernmental networks have been described as functional bodies shaping Europeanization of the neighbouring states via acquis communautaire transfer (Shyrokykh, 2019). They are established to improve sectoral performance, facilitate convergence with the legislations of the EU, andwhen necessaryprovide capacity-building to support adjustments at the domestic level. The European Commission (hereafter, the Commission) describes such cross-border co-operation as capable of shaping the administrative culture of the beneficiary states and contributing to the consolidation of democratic change (European Commission, 2006).
In this transgovernmental co-operation, conducted by the means of technocratic networks consisting of civil servants, EU agencies play an important role. 1 They often operate as hubs of experts (Eberlein and Newman, 2008, p. 29;Lavenex, 2015, p. 838). The role of the agencies in such networks is to facilitate capacity building and ensure the sharing of best practices with third countries' regulatory bodies. Within this *An earlier version of this article has been presented at the TARN Conference on the 'External Dimension of EU Agencies and Bodies', University of Luxembourg, Luxembourg, 27-28 June 2017. The authors would like to thank the participants of the conference for their helpful and constructive comments. We are also very grateful to the EU agencies and the European Commission for sharing data with us. Lastly, we are grateful to the three anonymous referees for their valuable comments. 1 We use the terms 'technocratic networks' and 'transgovernmental networks' interchangeably. By these we mean networks of civil servants in highly specialized technocratic settings in which civil servants establish tight co-operation to jointly target corresponding issues. co-operation, public officials possess a considerable degree of independence from the member states or EU central administration and are to a lesser extent subjected to the bureaucratic chains of command (Buess, 2015;Eberlein and Newman, 2008, p. 29). This semi-independent format enables considerable flexibility on both endsthe provider of technocratic expertise and the beneficiary stateallowing for individually-tailored coorganization of joint activities (Zeitlin, 2015).
The third countries' involvement in the operations of EU agencies is often perceived as a technocratic form of integration into the EU, a sort of de facto technocratic membership. Unlike EU institutions, such as the Council or the Commission that are closed for third countries, EU transgovernmental networks, including some EU agencies, are open for sector-specific engagement with third countries. Thus, transgovernmental networks assist the extension of the EU's regulatory boundaries by offering European Neighbourhood Policy (ENP) countries a flexible form of regulatory integration. It is this flexibility in co-operation settings between third countries and EU agencies that leads some scholars to nest transgovernmental network research within the external differentiated integration literature (Holzinger and Schimmelfennig, 2012;Lavenex, 2015).
The growing regulatory outreach of the EU is acknowledged in the literature examining the external governance of the EU (Lavenex, 2015;Lavenex and Schimmelfennig, 2009;Rimkutė and Shyrokykh, 2017;Shyrokykh, 2019). However, the impact of such EU engagement on the actual practices of neighbouring states remains under-researched. Although the phenomenon of EU transgovernmental networks has received some attention (see, for example, Bosse, 2012;Freyburg, 2011Freyburg, , 2015Katsaris, 2016;Shyrokykh, 2019), the role of EU agencies in this co-operation has not been considered, which is surprising given that EU agencies play an increasingly important role in the external governance of the EU (Hofmann et al., 2019).
In 2007, the Council agreed that some EU agencies would be opened to participation by ENP countries in their core technical activities (European Commission, 2006), which was aimed at enhancing 'regulatory and administrative reform', as well as promoting 'the convergence of partners' policies with EU norms, standards and good practice' (Commission, 2010b). Twenty out of approximately forty EU agencies were declared eligible to co-operate with the European Neighbourhood Policy (ENP) countries. 2 Building on the argument of European integration theories, as well as on international public administration literature (Christensen and Yesilkagit, 2018) (ECHA, 2018). Likewise, the Community Plant Variety Office (CPVO) has been involved in the neighbourhood via TAIEX programmes (CPVO, 2017). Since 2013, the European Asylum Support Office (EASO) has been engaged in asylum-protection programmes in some of the ENP countries (EASO, 2014). impact of transgovernmental networks and the role of the corresponding EU agencies in cross-border co-operation. In so doing, the article contributes to the external differentiation literature illustrating the role that public administration can play in defining the extent of neighbouring states' inclusion into the EU's regulatory framework.
More specifically, this article provides the first systematic cross-policy assessment of the effects of EU regulatory activities in the neighbourhood. Furthermore, it appraises the role of EU agencies in this co-operation. We show that the effects of the EU's regulatory outreach vary across policy areas. Policy fields marked by mutual interdependence exhibit greater positive effects of transgovernmental networks. Meanwhile, EU agency involvement implies broader and more intense transboundary co-operation, bringing technical co-operation to new policy domains. In this way, EU agencies build robust bridges between the EU and ENP countries on which thicker transgovernmental networks evolve.
The remainder of this article is organized as follows. First, we discuss the literature on transgovernmental networks and EU agencies, as well as review the role they play in the ENP region. Then, we introduce our theoretical approach to the study of transgovernmental networks and their effects and derive hypotheses that are later tested using a novel dataset and method. The next section presents the results. In the concluding section, we briefly summarise the contribution of the study and indicate broader implications of the results.

I. Transgovernmental Networks and EU Agencies in the Neighbourhood
Transgovernmental networks refer to the sustained technical interactions across state boundaries and are limited to public servants working within a specific policy field. They are created to support public administrations with regard to the approximation, application and enforcement of EU legislation, as well as to facilitate the sharing of EU best practices. The external administrative support follows a needs-driven approach and is aimed at delivering appropriate tailor-made expertise to address issues that third countries face (European Commission, 2019).
Such networks operate outside centralized bureaucracy or formal bodies of government and are particularly focussed on addressing common problems and sharing specialised knowledge about successful solutions (Shyrokykh, 2019). The EU's transgovernmental networks aim to fulfil several roles. Firstly, they are meant to address functional needs and specific cross-border problems, as well as jointly tackle issues and share experiences of effective problem-solving. Secondly, such co-operation seeks to help third countries integrate into the EU legislative framework without granting them formal membership (European Commission, 2003). Lastly, they are designed to facilitate domestic transition processes and reforms.
Scholarly work has also demonstrated that EU transgovernmental networks may be capable of stimulating democratic governance in third countries. They suggest that sectorspecific transgovernmental co-operation can not only facilitate the transfer of technical standards and promote legislative convergence outside Europe (Langbein and Wolczuk, 2012;Katsaris, 2016), but also positively impact democratic attitudes of civil servants in the beneficiary countries (Freyburg, 2015) and, in turn, affect the actual democratic practices of the EU's southern and eastern neighbours (Freyburg et al., 2009;Shyrokykh, 2019).
Technocratic co-operation within the European Commission's Technical Assistance and Information Exchange (TAIEX) is the largest policy-specific framework in public administration, and we therefore focus on this particular tool in this study. TAIEX's main goals are to improve the quality of public services, support the development of third countries' capacity, provide targeted technical assistance in drafting legislation related to the Action Plans and help the third countries with implementation and enforcement. TAIEX instruments cover a number of policy areas including agriculture and food safety; freedom, security and justice; environment, energy, transport and telecommunications; internal market; and TAIEX GEGIO Peer 2 Peera tool bringing together cohesion policy experts (European Commission, 2019). Drawing on a peer-to-peer approach, projects like those provided within the TAIEX framework function with the support of experts from member states' public administrations by providing beneficiary states in the neighbourhood with relevant tools and advice bringing their national legislation in line with the Union acquis (European Commission, 2006).
Technocratic co-operation between experts from member states and third countries can take place in two formatswith involvement of EU agencies or solely via transgovernmental networks, such as the TAIEX capacity building projects at the level of state administration (European Commission, 2006). In several selected policy domains, EU agencies are an important integral part of those transgovernmental networks offering technical co-operation to third countries. In this co-operation, EU agencies often serve as hubs of experts which are seconded to the beneficiary administration. In this way, establishment and maintenance of co-operation between EU agencies and third countries can be best described as a bottom-up process that is tailored to assist the technical and scientific knowledge transfer to the corresponding regulatory authorities. EU agencies, therefore, serve as facilitators in the process of transferring the EU regulatory state principles to third countries by furthering their administrative capacity and helping adapt to EU standards.
It is widely agreed that one of the key features of the development of the European regulatory state has been agencification (Majone, 1997(Majone, , 1999. The core justification for creating EU agencies is a growing need for scientific expertise, expert knowledge, and technical assistance to EU institutions and member states (Rimkutė, 2018). While the role of agencies in EU member states has received much scholarly attention (for an overview, see Egeberg and Trondal, 2017), their involvement in EU external governance remains under-researched. Existing research in the field only scarcely addresses the extent to which the involvement of EU agencies in regulatory outreach to third countries brings their regulatory standards closer to EU principles (see, for example, Hofmann et al., 2019).
There are more than 40 decentralized EU agencies and bodies involved in technical, scientific or managerial tasks that assist EU institutions in making and implementing European regulations. Half of the agencies are a part of technical co-operation with the eastern and southern neighbours (for a detailed mapping of agencies' outreach, see Rimkutė and Shyrokykh, 2017). Involving agencies in technocratic co-operation with the neighbourhood is 'a longstanding agreed key objective of the ENP, supporting reform and convergence with EU legislation' (European Commission, 2011, p. 2). Engagement of EU agencies in technical co-operation with the ENP states is viewed as a means of promoting reforms and development in the neighbourhood, as well as 'strengthen[ing] administrative and regulatory convergence of partner countries with the EU' (EEAS, 2016). The integration of ENP countries into regulatory activities of EU agencies infers a de facto integration of the neighbourhood in the EU regulatory framework.
In this way, agencies contribute to the rising role of scientific and technocratic expertise in various policy domains, not only within the EU, but also beyond its borders. EU agencies are involved in collaboration with ENP countries by either signing formal working/technical agreements or by engaging in more informal ad hoc co-operation (European Commission, 2006, pp. 4-5 and pp. 7-8). The first format of co-operation, namely, the one based on formal co-operation agreements with third countries, builds on technical co-operation underscoring long-term capacity building, a format found in the context of EU enlargement. The second format implies that agencies can be involved in the ENP on an ad hoc basis, such as by arranging events aimed at short-term technical capacity building. EU agency involvement and forms of participation in the ENP vary considerably across EU agencies: some EU agencies have both ad hoc and co-operation-based agreements with the ENP states, while others only recently started to engage in informal ad hoc co-operation (see Rimkutė and Shyrokykh, 2017). However, we know relatively little about whether and to what extent agencies' involvement affects the transfer of the EU acquis to the ENP countries. This study contributes to closing this research gap in the literature by (1) theorizing when and under what conditions transnational networkswhich EU agencies are part ofare successful in furthering integration of third countries into the EU, and by (2) systematically testing the theoretical expectations using a novel dataset compiled for the purposes of this study.

II. Theoretical Framework
Transgovernmental networks consist of civil servants from beneficiary states and their counterparts from EU member states. Existing literature suggests that such networks can have a profound impact on public servants' performance (Freyburg, 2015;Shyrokykh, 2019). They might be a particularly valuable tool in the neighbourhood where direct democratization pressure might meet open resistance: transgovernmental networks can serve as an alternative indirect way of stimulating adjustments (Shyrokykh, 2019). Existing work also highlights the significant differences in the extent to which transgovernmental networks can impact various policy areas. Studies attribute this variance to the adjustment costs associated with the specific policy domain (Langbein and Wolczuk, 2012;Shyrokykh, 2019).
Existing works argue that civil servants play a key role in maintaining the everyday functioning of state institutions: they play a crucial role in implementing policies, sustaining state institutions, interpreting and implementing laws (Shyrokykh, 2019). They are the body entrusted with carrying out a regime's decisions, serving as the major point of contact between citizens and the state. Regular and sustained interaction within transgovernmental networks might directly impact civil servants' practices through information sharing and diffusion of best practices. As a result, such a collaboration might ultimately be reflected in states' overall performance in the corresponding policy field.
Building upon the idea of knowledge sharing in a professional environment and bringing together relevant actors to engage in sustained co-operation, transgovernmental networks are often described as being capable of fostering capacity-building and constituting successful instances of external governance (Lavenex, 2008). Cross-border networks have been recognized as being capable of spreading 'know-how' expertise, shaping democratic attitudes and transferring new practices even in non-democratic environments (Freyburg, 2015;Shyrokykh, 2019;Turkina and Kourtikakis, 2015). Some also suggest that the inclusion of local actors in transgovernmental networks results in growing support for convergence with the acquis among state officials (Langbein and Wolczuk, 2012). Therefore, we expect that intensive co-operation in transgovernmental networks is likely to be positively associated with the subsequent improvements across various policy sectors in the ENP countries: H1: The more intensive co-operation is, the more likely it is to positively influence the subsequent practices of a state in the corresponding policy sector.
Although limited in scope, some recent scholarly work provides early attempts to shed light on the emerging phenomenon of 'agentification' of the neighbourhood and the effects of EU regulatory governance in third countries (Hofmann et al., 2019;Lavenex, 2015;Rimkutė and Shyrokykh, 2017). They show that various agencies become open to participation of third countries at different time points and to different extents. At the same time, there is little understanding of the effects of agency involvement across different policy domains. It is fair to expect that agency involvement can magnify the effects of transgovernmental networks for a few reasons.
First, literature suggests that the character of EU foreign policy vis-à-vis third countries might depend on the depth of already established involvement into co-operation with them. For instance, Hazelzet (2005) shows that the EU is less likely to impose sanctions on countries with which it has trade agreements (2005, pp. 9-10). Likewise, when investigating the most favourable conditions for positive effects of cross-border co-operation, existing studies suggest that the formalization of links, such as inclusion into special frameworks of co-operation like Association Agreements, Deep and Comprehensive Free Trade Area, or Partnership and Cooperation Agreements may magnify the outcome of this co-operation (Schimmelfennig, 2012;Shyrokykh, 2017).
In a similar way, inclusion of ENP states in agency activities via ad hoc or agreementbased collaboration can establish more robust technocratic links between ENP states' and member states' public administrations. It also can create frameworks for joint sustained monitoring, reporting and problem detection. In addition, sustained co-operation can stimulate learning from previous experience and optimise efficacy of future co-operation.
Second, in the presence of agency involvement, effective capacity building is more likely than in a situation of a novel exposure of public servants from both EU member states and ENP states to joint problem-solving within programmes such as TAIEX. When co-operation is supported by formal organizational structures, such as EU agencies with relevant expertise, we expect that both the sender and the beneficiary are more likely to address existing problems in a systematic manner. Serving as hubs of expertise, agencies might accumulate knowledge and experience facilitating co-operation planning and might be more likely to involve partners that have already proved themselves the most reputable and effective.
Third, establishing ties with agencies may serve as a signal of a long-term commitment to co-operation for both sidescivil servants from member states and ENP states; it may establish sustained channels of communication and create bureaucratic and inter-personal links facilitating future interactions. We expect that agency involvement in transgovernmental co-operation with the ENP partner countries may stimulate third countries to treat such co-operation more seriously.
Such a magnifying effect of agency involvement can be expected as a result of two mechanisms. In the ENP region, the EU seeks to enhance stability and security, as well as spread its standards of good governance via both conditionality aka the more-for-more instrument, as well as socialization through the people-to-people instrument (such as research funding schemes, capacity building programmes and regular exchange of ideas) (European Commission, 2003). Involvement of EU agencies in transgovernmental networks may help maintain a policy dialogue, which in its turn may create the necessary conditions for both conditionality and socialization to yield an effect.
Given the short-term nature of events within the TAIEX framework, they, on their own, are unlikely to trigger socialization of civil servants into new norms. However, when they are enhanced by EU agencies, such networks may expose civil servants from beneficiary countries to new norms on a more regular basis. Therefore, in such settings, effects of transgovernmental networks are more likely to transpire. Existing works demonstrate that networks nested in more permanent structures (sustained long-term collaboration) can stimulate socialization of their participants (Freyburg, 2011;Turkina and Postnikov, 2012).
Likewise, the magnifying effect of agency involvement may also be expected from the perspective of cost-benefit calculations triggered by conditionality. Involvement of agencies in transgovernmental networks may assure the beneficiary countries of longer co-operation perspectives, as well as financial and technocratic assistance that would come with future joint projects. Therefore, we expect the following: H2: When EU agencies are involved, transgovernmental networks are likely to yield more positive effects on the subsequent domestic situation in the ENP countries in the corresponding policy areas.

III. Research Design
To test the hypotheses, a novel dataset has been created. It accounts for the annual number of TAIEX events in seven regulatory sectors in which EU agencies are regularly involved, and the corresponding adjustments in the ENP states. 3 The sectors accounted for are: food safety (contributing agency: EFSA), environment (EEA), aviation (EASA), health (EU-OSHA, ECDC), border control (Frontex), asylum (EASO), and police and justice (Europol, Cepol, Eurojust, EMCDDA). These seven sectors are chosen due to the active involvement of EU agencies in these policy areas (for more information, see Rimkutė and Shyrokykh, 2017), as well as the availability of systematic EU assessments of the ENP countries' progress in each of these policy areas (through ENP Progress Reports issued by the Commission, for instance).
The structure of the data is time-series cross-sectional (TSCS). The data consist of the repeated observations of ENP states over 11 years (2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016). The study includes the following ENP states: Armenia, Azerbaijan, Moldova, Ukraine, Egypt, Lebanon, Jordan, Georgia, Israel, Morocco, and Tunisia. The countries are chosen on the grounds of data availability (such as annual country-specific Progress Reports). The time-span included in the analysis is also dictated by the availability of data. 4 The data on the annual number of TAIEX events were obtained on request from the Commission. These data reflect the annual number of TAIEX events by sector.
To analyze the data, we use an ordinal probit model. To ensure robustness of the findings, models utilizing instrumental variables were adopted to capture the possibility of endogeneity in the models. 5 The models were specified in accordance with the results of the corresponding tests for the violations of statistical assumptions. 6 The dependent variable in this study is a four-level ordinal variable that reflects the progress of each ENP state separately in each given year (for the coding scheme, see Appendix 3 in the online supporting information). It is operationalized by evaluating ENP states' compliance with EU demands regarding both policy adoption and implementation. The variable captures the extent of improvement (since a year before) in regulation in the corresponding policy area: 0no improvement, 1limited improvement, 2some improvement (new strategies, new law drafts bringing domestic legislation in line with EU regulatory standards), 3good improvement (full adoption of new laws aligned with the EU acquis, full adherence to EU standards, creation of new institutions). 7 The dependent variable has been coded manually using country reports issued by the Commission (see Appendix 1 in the online supporting information for an overview of documents coded in the study). We acknowledge that the Commission's country reports may reflect some biases, such as the interests and preferences of the EU. However, relying on documentation of the ENP countries themselves could introduce even more biases that we cannot account for (political preferences of domestic actors and country-specific contexts, for example). To that end, we recognize that the Commission's reports may include certain predispositions, however, they allow us to trace compliance in corresponding policy areas in a systematic way across ENP countries. The Commission's reports systematically trace core changes in the ENP and summarize whether, and if so, to what extent, improvements took place in each of the seven sectors we list above.
The independent variable accounts for the annual number of TAIEX events in each policy sector. Altogether, between 2006 and 2016, there were about 3,000 events taking place in the ENP region in various sectors (Shyrokykh, 2019). Ukraine and Moldova are the most active in taking part in TAIEX projects (see Figures A2 and A3 in Appendix 2 online). Furthermore, the number of events steadily increased over the years (see Figure  A1 in Appendix 2 online). 8 Most EU agency involvement with the region took place via TAIEX projects coordinated by the Commission. For instance, co-operation with EASA, EEA, EFSA, ECHA, EMA, EMSA, ECDC, and Europol is based on requests by the ENP countries and tend to be implemented through the Commission's TAIEX tool. Therefore, we argue that by focussing on the TAIEX tool, we can capture transgovernmental co-operation. For the same reason, we focus on the policy areas where the EU regulatory governance is the most active. If transgovernmental networks can have any impact on the actual practices of third countries, we expect that in these policy areas.
Our second theoretical expectation is that when EU agencies are involved, transgovernmental networks are likely to yield more positive effects on the subsequent domestic situation in the ENP countries in the corresponding policy areas. To test this hypothesis, we create a binary variable that reflects whether the country has established either ad hoc or formal technical co-operation agreements with EU agencies (for instance, EMCDDA has co-operation agreements with Ukraine (since 2010) and Moldova (since 2012); Frontex has working agreements with Armenia (since 2012), Azerbaijan (since 2013), Moldova (since 2008), Ukraine (since 2007) and Georgia (since 2008)). To test this hypothesis, we include in the models an interaction term between the treatment variable (annual number of events in the corresponding policy area) and the binary variable capturing whether agencies are involved. The data on agency involvement were collected from official agency reports and verified by contacting individual agencies (see Appendix 1 online).
The models also include several control variablesfactors that might determine domestic progress. These are political stability (World Bank, 2016a), the rule of law (World Bank, 2016a), population (World Bank, 2016b), log GDP per capita (World Bank, 2016c). To isolate the effects of the strength of state institutions, we control for the level of democracy using Polity IV dataset (Marshall et al., 2018), as well as the level of governance effectiveness. The governance effectiveness variable reflects perceptions of the quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government's commitment to such policies. The data are taken from the World Governance Indicators (WGI) dataset (World Bank, 2016a). The rule of law variable reflects perceptions of the extent to which agents have confidence in and abide by the rules of society, and, in particular, the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence. The data are obtained from the WGI dataset. The political stability variable measures perceptions of the likelihood of political instability and/or politically-motivated violence; the data are also taken from the WGI dataset.
To exclude alternative explanations, we control for the possibility of a spillover effect from participation in other projects, that is we control for the number of the TAIEX events in all other sectors. Additionally, given that TAIEX co-operation is demand-driven, we introduce a variable that accounts for the number of requests for TAIEX co-operation 8 To exclude the trend from the data, we differentiate the independent variable.
(the data were received by request from the Commission). Variables included in the model are briefly summarized in Table 1.

IV. Results and Discussion
Each of the models separately assesses the effects of transgovernmental co-operation in seven policy areas: police, asylum, border control, health, food safety, aviation, and environment. The first hypothesis, suggesting that the more intense co-operation is, the more likely it is to positively influence subsequent alignment with the EU standards in the corresponding policy area, is confirmed (see Table 2). The effects of transgovernmental co-operation are statistically significant in Police, Asylum, and Aviation policy domains. The results are intuitive and relate to the objectives of the ENP policy but require some explanation.
Police co-operation has been referred as one of the most important policy areas within ENP (European Commission, 2003). Co-operation in this area is regarded as crucial for the member states' safety and, in fact, co-operation projects in this policy area are among the most numerous in comparison to co-operation in other policy areas (see Table 1). Likewise, co-operation in asylum policy is among the most crucial for the ENP framework. The EU regulation 439/2010 of the EU Parliament and the Council establishing   Notes: All models are estimated with regressor lagged by one year. Level of statistical significance: * p < 0.10, ** p < 0.05, *** p < 0.01.
the European Asylum Support Office (EASO) explicitly states that 'the Support Office shall be fully involved in the external dimension of the Common European Asylum System (CEAS)' (p. 14). Asylum policies in the ENP states directly affect the migration situation in the EU, since the ENP states often serve as transit zones for both legal and illegal migrants traveling to the EU member states. Co-operation between the EU and ENP countries on asylum policies is also referred to as one of the benchmarks against which countries' progress is evaluated in the context of the neighbourhood's regulatory integration. For instance, this policy was included in visa-free travel negotiations with three Eastern Partnership countries -Georgia, Moldova, and Ukraine (European Commission, 2013;2014). In a similar vein, co-operation in the policy area of aviation relates to safety and security. It is of interest for both the EU and the ENP states to maintain safety in this sector and regulate corresponding activities in a joint effort. Aviation represents an area where both the sender and the beneficiary state have common objectives and are interested in co-ordinating their activities, synchronizing regulations, improving air transport safety, as well as sharing information and best practices. The EASA plays an important role in ensuring smooth technical co-operation, as well as closely monitoring how the ENP countries adapt EU standards. For instance, the EASA regularly reports progress in the process of implementation of the Common Aviation Area Agreement by ENP countries (European Commission, 2015, p. 18). The Commission emphasizes that through such a transfer of EU safety regulations, ENP countries will gradually adapt to EU standards in the field of aviation and security (European Commission, 2010a, p. 17).
In this way, transgovernmental networks help the EU transfer its rules beyond its borders. They not only grant access to policy-specific expertise, but also provide the necessary financial and technical resources needed for the effective implementation of the corresponding acquis. Areas such as asylum, for instance, although comprising an essential element of domestic security, cannot be effectively regulated by states with a lower level of state capacity. Therefore, financial and technical assistance provided within the networks can significantly improve the likelihood of their positive effect. For instance, when preparing for visa-free travel arrangements, Ukraine had to comply with a number of asylum regulations, improving the quality of facilities and services for asylum seekers in (and passing through) Ukraine. Acknowledging that Ukraine has limited capacity to fully comply with such requirements, the EU provided Ukraine with the necessary financial and technical resources to improve both asylum facilities and related services (European Commission, 2013, 2014, 2016.
In sum, the results support the argument that the more intensive transgovernmental cooperation is, the more likely it is to positively influence the subsequent domestic situation in the corresponding policy sector. The positive effects are detected in the areas of vital common interest for both ENP countries and the EU; and failure to regulate these policy areas may directly lead to high safety-related risks for both.
Disregarding regime type and political situation in a country, co-operation in policy areas related to safety is of interest for both parties. Functionally, such co-operation serves common interests and adjustments in these policy fields are highly unlikely to directly yield any political risks for ruling elites. At the same time, there is variation in the size of the effects of technocratic co-operation across policy fields which merits separate investigation in future work.  The second theoretical expectation suggesting that the effects of technocratic cooperation might be larger when EU agencies are involved, is not confirmed ( Table 3). The results demonstrate that agency involvement does not impact the extent to which cross-border co-operation has an effect. At the same time, the analysis of the difference in the number of transgovernmental events when agencies are involved and when such involvement does not take place reveals that agency involvement is associated with a higher number of joint events (Table 4).
At a statistically significant level, however, agency involvement explains the extent of technocratic co-operation only in the policy areas of environment and food safety ( Table 5). The results suggest that agencies play an important role in strengthening transgovernmental networks; involvement of agencies implies a higher level of cooperation activities between the EU and third countries. In other words, agencies create a robust co-operation structure, on the basis of which tighter transgovernmental cooperation evolves.
Meanwhile, agency involvement does not imply a higher number of co-operation events in every policy area. Thus, in police, aviation, asylum, border control, and health policy areas, co-operation with agencies does not explain the difference in the intensity of technocratic networks (Table 5). These results follow a functionalist logicthe nature of the externalities associated with these five policy fields has a trans-boundary character. These five policy domains relate to cross-border issues that may directly or indirectly endanger safety and stability in EU member states as indicated in the Commission's 'Wider Europe' (2003) document laying a foundation stone of the ENP. Failure to effectively regulate policies in these fields might directly jeopardize the corresponding domains in the EU. Therefore, it is reasonable that tight co-operation in these areas is developed disregarding agency involvement. Meanwhile, in the policy areas of food safety and environment, the involvement of EU agencies -EFSA and EEA, respectivelyis associated with a higher number of joint events.
In sum, the results support earlier studies in showing that technocratic co-operation can positively affect domestic situations in the neighbourhood (Freyburg, 2015;   2019). At the same time, the involvement of EU agencies does not directly influence the effects of technocratic networks, agencies rather facilitate co-operation and, in turn, increase the overall number of joint events. In this way, EU agencies serve as facilitators of sustained co-operation between the EU and states in the neighbourhood, strengthening and broadening the scope of cross-border professional interactions.

Conclusions
Can the EU transfer its administrative practices to the neighbouring countries, and if so, what is the role of transgovernmental networks, of which EU agencies are a part, in this process? This article contributes to the existing literature by assessing the influence of EU regulatory transfer in the ENP countries across seven regulatory sectors. Relying on the TSCS analysis of 11 ENP states across the time period from 2006 until 2016, we illustrate that technocratic networks can foster ENP states' regulatory progress. Policy sectors marked by a higher degree of interdependence exhibit greater positive change as a result of transgovernmental co-operation. Although the involvement of EU agencies does not have any direct effect on the extent to which technocratic networks impact the ENP countries, the agencies nonetheless play an important role in strengthening transgovernmental co-operation. Their involvement stimulates more intense and broader transboundary co-operation, especially in policy areas that are not marked by high safety priority. In doing so, agencies serve as promoters of tighter links between ENP countries and the EU, increasing regulatory outreach of the EU (especially in policy areas that do not compose a core security interest of the ENP, such as environment and food safety). In doing so, they broaden the EU's regulatory extension and cover policy areas which otherwise might not have been covered.
These findings have important implications for policy making. They suggest that EU agency involvement indirectly simulates EU external regulatory outreach. Furthermore, the results inform us about wider implications of how the EU regulatory state extends beyond its borders. The findings suggest that transgovernmental networks make a difference in how the acquis spreads to the neighbourhood: in the policy areas where technocratic co-operation is dense, the EU is able to induce regulatory change. However, we also demonstrate that transgovernmental co-operation opportunities are unequally distributed following policy-specific interdependence patterns, leaving other policy areas behind.