Administrative and organizational capacities of civil society in EU Cohesion Policy

ABSTRACT Administrative capacities are among the crucial factors influencing success in European Union (EU) Cohesion Policy absorption. The current research concentrates on the public sector, while administrative capacities in other stakeholders are omitted. Our research focuses on whether local stakeholders from civil society have sufficient capacities to contribute effectively and efficiently to EU Cohesion Policy implementation. We performed our research on 57 integrated urban development plans in Czechia and Portugal and conducted 33 interviews with local entities. The results indicate a different level of capacity not only between the public and civil society organizations but also within civil society.


INTRODUCTION
Absorption capacity is a highly debated theme in European Union (EU) Cohesion Policy.It concerns not only the financial aspects (amount spent and successfully certified by the European Commission -EC), but also effectiveness and efficiency (i.e., whether the objectives were achieved and the associated costs) (Bachtler et al., 2014;Milio, 2007;Tosun, 2014).The lack of administrative capacities among public sector managing authorities is responsible for the low financial absorption of EU funding, especially when combined with changes in political representation (Hagemann, 2019;Surubaru, 2017).
Understandably, researchers concentrate on the public sector.This sector is the bearer of political and managerial responsibilities, and the largest beneficiary simultaneously.Nevertheless, this sector is not the sole contributor to the successful absorption of EU funding.Private firms and civil society organizations (CSOs) 1 are the stakeholders that help increase the relevance of the policy by helping in defining societal needs.Moreover, these stakeholders increase their absorption by implementing their investments and development projects (González et al., 2015).
Attempts to improve policy implementation by involving stakeholders generally concern public policies and programmes (Organisation for Economic Co-operation and Development (OECD), 2003).The current discussion on individual consumers participating in producing public goods (co-production), designing policies (co-creation) and on CSOs participating in designing policies (partnership) explains this development very well (Brandsen & Honingh, 2018;Potluka, 2021).Combined with the current crisis of trust in institutions, it underlines the importance of local social capital and local stakeholders.Stakeholders are crucial for improving the relevance of political decisions concerning local needs through co-creation and co-production processes.EU Cohesion Policy is not an exception (Potluka, 2021;Potluka et al., 2017b).
Co-creation and co-production also contribute to EU Cohesion Policy.Tools such as integrated urban development plans (IUDPs) and community-led local development (CLLD) represent place-based interventions in EU Cohesion Policy.These policy tools are tightly linked to functional territories in which local stakeholders participate in projects (Medeiros & van der Zwet, 2020a).The importance of co-creation and co-production in EU Cohesion Policy increased gradually with the debate on the territorialization of Cohesion Policy, particularly regarding the Barca Report's (Barca, 2009) place-based policy implementation approach.
Here, we study the case of the EU financially supported IUDPs in Czechia and Portugal.These two countries were shortlisted because of their transition history, which can reveal whether a longer EU membership makes a difference in how co-creation and co-production are implemented in EU Cohesion Policy.In the Portuguese case, the adhesion to the EU in 1986 had a direct positive impact in forcing the approval of municipal planning instruments with a territorial diagnosis and a strategic vision, as a sine qua non condition for the municipalities to access EU funding (Ferrão & Campos, 2015).This approach resulted in the acquisition of spatial planning know-how, which paved the way for designing sound and robust IUDPs from a strategic vision design and implementation process (Medeiros & Van der Zwet, 2020b).Meanwhile, Czechia became an EU member state by 2004, and consequently had less time to absorb EU policy implementation-related practices and procedures; this is reflected in the design and implementation process of their IUDPs.
In both countries, there is a positive opinion that CSOs can influence decision-making at the local and regional levels (74% for Czechia and 77% for Portugal; EU average ¼ 75%).Contrary to the Portuguese, the Czechs are more pessimistic about the general influence of civil society at the national or EU level (55% and 37%, respectively, compared with 70% and 53% in Portugal, respectively) (TNS Political & Social, 2013).People tend to feel connected to their nation and locality, and are far less connected to being Europeans (Capello, 2018), which amplifies the importance of tools such as IUDPs and CLLDs.
This study pays particular attention to civil society because individuals participate in their activities voluntarily and are highly motivated in their actions.Conversely, the main activities of CSOs are not primarily related to the development of local strategies and active participation in political decision-making.The incongruity between demands imposed on the CSOs in EU Cohesion Policy's implementation and CSOs' expectations raises questions about whether local CSOs have sufficient capacities to contribute effectively and efficiently to the EU Cohesion Policy's implementation.Notably, what capacities do they need and how can these capacities be built?While answering these questions, we intend to contribute to the information on improving the implementation of EU Cohesion Policy by involving local stakeholders.

Policymaking and policy implementation
The recent discussion on public service provision has reintroduced co-production and co-creation (Brandsen & Honingh, 2018;Nabatchi et al., 2017).In co-production, individuals or groups of individuals participate in producing public goods and services together with the public sector.Moreover, they are positioned as consumers at the same time.Thus, co-production can address the local needs more precisely as the consumers are directly involved in service production and can directly control the quantity and quality delivered (Cepiku et al., 2016;Loeffler & Bovaird, 2016).On a more strategic level, co-creation enables individuals to participate in policymaking and political decision-making (Brandsen & Honingh, 2018, p. 13;Lindenmeier et al., 2021), even if the politicians make the final political decisions.
The processes of co-production and co-creation are not straightforward.All stakeholders must be willing, capable and can provide some type of knowledge, resources, compliance, ideas and creativity, and legitimacy to co-production (Loeffler & Bovaird, 2016;Hager & Brudney, 2011;Haski-Leventhal et al., 2018).However, these processes are far from ideal.Individuals and CSOs can suffer from a lack of capacities, such as expertise, know-how and skills (Vamstad, 2012;Williams et al., 2016).Thus, further communication is required to achieve co-production (Crompton, 2019).Moreover, politicians may feel threatened and reject co-production (Loeffler & Bovaird, 2016).Thus, finding solutions accepted by all stakeholders is difficult and time-consuming.Our research investigates how CSOs are equipped with capacities to successfully implement co-creation in the case of IUDPs implemented in EU Cohesion Policy.
A lack of capacity leads to low effectiveness and efficiency of implemented policy interventions (Bachtler et al., 2014).Moreover, changes in the stakeholders involved further undermine the absorption capacities.This concerns not only unstable governments and changes in political representation (Hagemann, 2019;Milio, 2007Milio, , 2008;;Surubaru, 2017), but also partners participating in the design of programmes (Potluka et al., 2017b).Sometimes, the lack of capacity is filled in by consulting firms, notably when there is no previous experience in implementing EU-financed programmes, especially at the local level (van der Zwet et al., 2017, p. 57).However, the main burden related to the lack of these capacities is borne by the public sector, although other stakeholders are crucial for the preparation and implementation of projects.
For Milio (2007), administration capacity is based on the four phases of the programming cycle: (1) management (project preparation and selection, and financial management according to Bachtler et al., 2014), (2) programming, (3) monitoring and (4) evaluation (Milio, 2007).Milio's (2007) concept relates primarily to the public sector, but the framework is also applicable to the CSOs, though the term 'organizational capacities' is more appropriate rather than 'administrative capacities'.We use the term 'capacities' throughout this study.We always mean organizational type capacities, although we no longer emphasize the differences between administrative and organizational capacities.
The need for capacities also depends on the type of CSO, phase of the policy cycle, theme and intensity of co-creation and co-production required, and complexity of relationships among all stakeholders (Bovaird, 2007;Brandsen & Pestoff, 2006;Mazzei et al., 2020).The variance of requirements on capacities makes it difficult for CSOs to provide sufficient expertise in all four phases.CSOs can manage projects and help with programming, although with some difficulties relating to the fluctuation of their representatives (Potluka et al., 2017b).Nevertheless, similar issues concern the public sector (Tödtling-Schönhofer et al., 2014, p. 72).Evaluation is probably the most challenging part among the capacities for CSOs (Milio, 2007).According to recent research, evaluation capacities are deficient in CSOs in all European countries, even in those with a high evaluation culture in the public sector (Stockmann et al., 2020).Even in other parts, CSOs do not play a decisive role compared with the public sector.For example, CSOs were responsible for preparing strategies in a minimal number of cases of IUDPs (van der Zwet et al., 2017).
To support the role of CSOs in the design and implementation of EU Cohesion Policy, the EU has introduced several tools.The IUDPs in urban areas and CLLDs in rural areas are EU Cohesion Policy tools used to develop places.They represent a unique opportunity to study the functioning of co-production and co-creation in various circumstances at the local level when the general approach is the same for all EU member states.The place-based approach represented by these tools supports the opportunity for EU citizens to gain the advantages of EU Cohesion Policy (Barca, 2009).

CSO capacities relating to EU Cohesion Policy
To qualify for participating in EU Cohesion Policy, CSOs need sufficient capacities.The literature on CSOs has already investigated their capacities in several interrelated categories.Financial capacities, human resources, operational knowledge, expert knowledge and political networks are the most important (Bowman, 2011;Carmin, 2010).All are also relevant for the CSOs' participation in EU Cohesion Policy and IUDPs' implementation.EU Cohesion Policy makes it a specific case, as CSOs need to have capacities in both their specialization and the specifics of EU Cohesion Policy's implementation.This is for both the national and local levels.For example, the CSOs representatives' willingness to participate decreased due to an imbalance between the requirements of the superficial procedures in monitoring committees, CSOs' capacities and CSOs' chances of success.Thus, their willingness to participate decreased.This was regarding almost all CSOs in the programming periods 2004-06, 2007-13 and the beginning of the period 2014-20 in Central European countries (Batory & Cartwright, 2011;Potluka et al., 2017b).Another practical issue concerns the differences in the understanding of project management approaches.In EU-funded programmes, it is usually the project cycle management approach used by the EU (European Commission, 2004), while the private sector uses other approaches, such as those defined by the International Project Management Association (IPMA) (Fuster, 2006;Pantouvakis, 2017).These differences become an additional administrative burden for CSOs.
EU Cohesion Policy is also a specific case for CSOs as it provides direct funding.Sufficient funding can help CSOs hire or train staff to improve services, satisfy their clients' needs and participate in public affairs.Studies reveal only the effect of EU Cohesion Policy on shortterm financial capacities (improvement of budgets) in CSOs, but not on the long-term strengthening of CSOs' assets (Potluka et al., 2017a;Potluka & Svecova, 2019).Without sufficient financial resources, CSOs will not be able to participate effectively in public affairs (Lane, 2010).
We distinguish three defining factors for effective and efficient civil society participation in the implementation of EU Cohesion Policy.To engage in IUDPs' design and implementation, all these three requirements must be met simultaneously (e.g., for volunteering, see the same approach in Hager and Brudney, 2011;and Haski-Leventhal et al., 2018).The first factor concerns capabilities, including knowledge and skills among CSOs (von Schnurbein, 2014).These organizations need to know how to participate in the policy implementation process and how to manage their participation.This section covers the administrative capacity.Sundeen et al. (2007) also add the need to provide information about existing opportunities to involve CSOs.
The degrees of CSOs' capacities vary depending on different factors.Local stakeholders possess lower capacities than their national counterparts (Dąbrowski, 2013).This relates to opportunities that CSOs take to participate in programming.National CSOs can participate in various programmes and share expertise across their branches, while it is impossible for local CSOs.
Considering this research framework and the abovediscussed importance of capacities, we analyse two hypotheses covering the entire civil society sector, especially local CSOs and the participation of engaged individuals: Hypothesis 1a: CSOs have sufficient capacities to contribute to the IUDPs' implementation based on their knowledge of local needs.Hypothesis 1b: Individual people have sufficient capacities to contribute to the IUDPs' implementation based on their knowledge of local needs.
Hypothesis 2: Local CSOs have lower administrative capacity than national organizations.
The second factor relates to the possibility of CSOs participating in these processes.IUDPs count on the participation of CSOs (European Commission, 2014).Dependency on national rules and practices, legal norms, or approaches given by the long-term relationship between the public sector and CSOs defines the actual possibility of engaging in designing policies and their implementation.CSOs as a sector are established partners in designing EU Cohesion Policy programmes (Tödtling-Schönhofer et al., 2014).CSOs' inclusion in programme design and implementation is especially successful in 'mature' programmes (Polverari & Michie, 2009) as they get more time to experience and learn what they need.
The differences in national practices in implementing collaborative practice enable the comparison between Czechia and Portugal, as the perception in these countries differs from how influential CSOs are in political decision-making (TNS Political & Social, 2013).Thus, we test the following hypothesis: Hypothesis 3: Longer EU membership makes a difference to practices in the implementation of co-creation and coproduction in EU Cohesion Policy.
Finally, the third factor is the willingness of CSOs to participate.That is, whether the CSOs are motivated to find ways to participate actively.Although willingness is an intrinsic factor, changing capabilities or possibilities can increase motivation and the actual willingness to participate.We summarize the research framework and its dimensions in Table 1.

Analysis framework
To frame our analysis, we combine two approaches to capacities.The first dimension concerns the four phases of the programming cycle: management, programming, monitoring and evaluation (Milio, 2007).The second dimension is defined by capability, possibility and CSOs' willingness to participate in the EU Cohesion Policy's implementation.We follow the aforementioned requirements that must be met simultaneously (Hager & Brudney, 2011;Haski-Leventhal et al., 2018).The first factor concerns capabilities, including knowledge and skills among CSOs.CSOs need to know how to take part in policy implementation and how to manage their participation.This factor covers administrative capacity.The second factor relates to the possibility of CSOs participating in the policy implementation processes.The willingness of CSOs to participate is the third factor.

Data
In-depth analysis of the IUDPs Methodologically, we base our research on data from an in-depth analysis of the IUDPs, accompanied by interviews with public servants and civil society representatives in Czechia and Portugal.In the IUDP strategic documents, we analysed three aspects of their preparation: (1) the role of CSOs; (2) the role of individuals when designing the strategy; and (3) the value-added by local individuals and the CSOs to the strategy.
We used a five-point scale to evaluate the fulfilment of the three criteria in our data (for a full explanation of the coding, see Appendix A, file 1, in the supplemental data online).For example, rank 5active participation from both CSOs and the public sectoris when the CSOs actively entered the process with their ideas and proposals, and the public sector welcomed them.This best fulfils the criterion of the contributing role of the CSOs in designing the strategy.With the decreasing activity of the actors, there was a need for the public sector to invite them.In such a case, the public sector suggested some solutions and asked the stakeholders to comment on them.The stakeholders responded to this call (rank 3partly inactive CSOs, but an active public sector trying to establish cooperation or vice versa).The least interactive is the situation (rank 1) where the public sector is neither actively inviting CSOs nor are the CSOs interested in getting involved in the preparation of IUDPs.

Interviews
The documents cover the design processes of the IUDPs, but not the actual implementation.Thus, to understand the actual implementation, we use a secondary source of data via interviews with stakeholders (for the interview questions, see Appendix A, file 2, in the supplemental data online).With knowledge about the strategic plans, we contacted potential interviewees through purposive sampling.All interviewees participated actively in designing and implementing IUDPs.This reveals how participation processes look beyond EU Cohesion Policy.We organized the interviews in April, May, October and November 2019.Among the 46 cities implementing IUDPs in Czechia and 103 in Portugal which were analysed, we received responses from 12 public servants from cities and two regions in Czechia, and 13 in Portugal (Table 2).We also conducted three interviews with CSOs from both countries.The lower number of CSOs among the interviewees was due to the CSOs being absent as official partners in the IUDPs.Several interviewees participated in more than one IUDP, thus providing information for more IUDPs in one interview.
The 13 Portuguese case studies are all located in the Lisbon Metropolitan Area, which encompasses 18 municipalities.Hence, they are fully representative of the IUDPs in this Portuguese region (NUTS-II).The focus on this region stems from the fact that the implementation process of the IUPDs in Portugal is not as advanced in the remaining Portuguese NUTS-II (Norte, Centro and Alentejo).Moreover, due to the similar strategic visions of all 103 Portuguese IUPDs in focusing on solving problems faced by more deprived urban neighbourhoods and in similar policy areas (support for physical rehabilitation or decaying urban areas, and supporting neighbourhoods in extreme socio-economic deprivation), the selected Portuguese case studies can provide generalized findings for the whole of Portugal.The problems that the IUDPs target are similar in all intervention areas.Fundamentally, the Portuguese IUDPs can be compared with the Czechia IUDPs for the simple reason that the same EU policy framework rationale supports them to sustainable urban development via an integrated policy approach.
Twelve Czech interviewees were directly involved in designing IUDPs, although their roles varied during the implementation process.They represented departments in municipalities responsible for local development and the preparation of development strategies.Moreover, we interviewed two Czech members of the regional development councils who were responsible for allocating funds and implementing projects at regional levels.The interviews concerned local strategies implemented in the selected localities.We have information on the processes involved in the design and implementation of the strategies and practices used to involve local stakeholders.A similar scenario occurred in the Portuguese case studies.In detail, all interviewees of local public entities were related to the urban planning departments of the respective municipalities.Not all of them participated in the design of IUDPs.However, all were part of the current implementation process.The CSOs were not particularly involved in the design of IUDPs.However, they assisted the current implementation phase in the social inclusion domain in socioeconomically deprived urban neighbourhoods.

Methods
To test Hypotheses 1a + 1b, 2 and 3, we constructed 11 models (see Table B1 in Appendix B in the supplemental data online).Regarding Hypothesis 1, we expect that if CSOs have high capacities, they (1) prove their capacities by achieving high added value where they are welcomed as partners, and (2) do even better to achieve high added value in situations where the CSOs are not welcomed as partners.if the CSOs are welcomed but fail to achieve a high added value, their capacities are low.For this, see models 1-5.Another way is to take CSOs' added value from problematic territories, where the public sector could not solve the social problems in the long-term but where CSOs can add high value (model 6).Regarding Hypothesis 2, we test whether the achieved high added value by CSOs related to their regional and/or national networks (see models 7-9).To test Hypothesis 3, we used models 10 and 11 and comparisons of the results of models 4 and 5, where we distinguish between the Portuguese and Czech cases.
These models provide answers to the hypotheses.We added an analysis of the responses from the interviews to discuss the results in more detail.Thus, our approach is a mixed methods approach that uses both quantitative and qualitative analyses.

RESULTS AND DISCUSSION
Are CSOs capable of contributing to IUDPs effectively?
The policy collaboration process depends not only on national approaches, but also on local habits and stakeholders' capacities.Practices in the management of IUDPs vary according to both local political representation and CSOs.The intensity of CSOs' involvement varies from local officials, almost ignoring them to full involvement.According to the analysis of the IUDPs and interviews in Czechia, CSOs in cities dealing with excluded social groups tend to contribute more to societal problem solutions than CSOs in cities that are poles of growth (model (6), +0.526, p ¼ 0.128).Such CSOs possess capacities to successfully help in designing and implementing the IUDPs.These CSOs were regularly invited to participate in local policy discussions.
Compared with the involvement of individuals, these organizations can provide in-depth knowledge on the needs of the target groups and a pool of experts.Thus, CSOs, compared with engaged individuals, are more structured and more intensely engaged in collaborating on the IUDPs in both surveyed countries.However, the tests are insignificant (a difference of +0.101, p ¼ 0.698).
Moreover, in our sample from Czech cities, we identified 83 CSOs in the steering groups of the IUDPs.More than half were also active at the regional (11) and national (32) levels.These CSOs can provide expertise in various themes.This finding complies with the long-term situation in the civil society sector.There are approximately 100 fully professionalized CSOs with strong management  2007).To answer Hypotheses 1a + 1b and 2, we elaborated primarily on the relationship between management, capacities, monitoring and evaluation, and capacities.To answer Hypothesis 3, information was taken from the relationship management and programming, on the one side, and possibility and capability, on the other.
Administrative and organizational capacities of civil society in EU Cohesion Policy 749 REGIONAL STUDIES (GCNNO, 2008).These CSOs have an advantage because they share their experiences within their internal networks.Thus, they can increase their knowledge capacities to participate in local development initiatives due to economies of scale.In Portugal, local CSOs were primarily selected for partnerships in the IUDPs.Concerning long-term participation, a Czech public servant pointed out the following: 'The cooperation [between CSOs and one Czech city] lasts for over the years with a group of engaged people who are still the same.However, a small generation change is already evident' (INT-C10).
In almost all cases in our sample, cities viewed IUDPs as an opportunity to raise money.In essence, the projects would either not be implemented at all or only to a minimal extent (e.g., Vsetín, Liberec, Třebíč, Chomutov, Jablonec, Most, Ústí nad Labem, Alcochete, Barreiro, Loures, Mafra, Montijo, Odivelas, Palmela, Sesimbra, and V. F. Xira).This approach also corresponded to the limited range of strategies that cities prepared.The usual way of preparing the IUDPs and their content was the presentation of already prepared investment projects and approval of their form by other stakeholders (including engaged individuals and CSOs) (e.g., Vsetín, Liberec, Třebíč, and Ústí nad Labem).CSOs' capacities are limited compared with the public sector.Thus, CSOs' contribution was limited.In the Portuguese case, the preparation of IUDPs was a sine qua non condition for accessing EU funding for the 2014-20 period.Due to the short deadlines to prepare the strategic documents, most municipalities consulted external private firms to prepare the IUDPs.This largely explains why CSOs and the general population had a relatively small contribution to the elaboration of the IUDPs.In certain cases, however, they were consulted in public municipality forums regarding specific parts of the IUDPs, particularly on the intervention strategies for socioeconomically deprived neighbourhoods.This situation points to some existing capacities among the CSOs which were not large enough to react quickly to the developing situation.
Cities primarily used standard communication channels when preparing IUDPs (Vsetín, Liberec, Jablonec, Třebíč, Most, Lisboa, Almada, Oeiras and Setúbal).Participation by individuals was consistent with standard local policy tools commonly used by local politicians without empowering stakeholders, improving stakeholder ownership of the IUDPs, or building capacities among them.In most cities, communication usually took place only with people who were potentially directly affected by the investment.
Local inhabitants were invited less intensively to participate in designing the IUDPs for three reasons.The first issue was about the expertise provided by the participants.CSOs, compared with engaged individuals, are capable of sharing the workload within the pool of their experts.Usually, civic movements and initiatives develop into CSOs because of this institutional advantage.The second issue relates to management capacities, including financial capacities, according to Milio (2007).CSOs as legal persons are subject to state aid regulations, and usually the investment is not related to their property or other assets.Third, when city representatives ignored the CSOs and only invited the citizens of a city, they did that purely for political reasons as they regard people as an electorate that participates in the process but is not considered as a partner (Třebíč and Příbram).This situation shows that the primary objective was to meet the goals of local politicians and not the more general principle of partnership.
In terms of capacity assessment, Czech CSOs played a more critical role in preparing IUDPs than individuals, while in Portugal, the situation was the opposite (Table 3).City dwellers' interest increased in how and where to invest and develop the city, especially when it concerns investment in residential houses.Thus, there was an apparent observational interest in designing the IUDPs and actual activity in the implementation.However, in the preparation of IUDPs, it was evident that if there was a real interest by residents in the formation of the IUDPs, then these engaged people gathered in the CSOs and shared operational and expert knowledge in networks (Potluka, 2021).
In terms of capacities, the high time-consuming demand in the preparation of the IUDPs was problematic.

REGIONAL STUDIES
This was mainly about the amount of information to be studied and the subsequent work in working groups.This situation has prevailed since the previous programming periods (Polverari & Michie, 2009).In the CSOs, activities were performed in a complicated manner and without a claim to reimburse this effort.In that sense, one of the social services CSOs' representatives pointed out the following: 'The most significant obstacle is that I cannot adjust my working hours to go to the meeting' (INT-C18).
The presence of CSOs in steering committees provides supporting evidence on their potential success.The estimates exhibited +1.651 higher points in evaluating taking CSOs as partners compared with the IUDPs without the CSOs in steering committees (model 2, p ¼ 0.000).Moreover, the added value of these CSOs to the IUDPs is higher compared with the IUDPs without the direct presence of the CSOs in the steering committees: t-test (1) reveals a difference of +0.809, with the score of the added value of CSOs being 3.592, while without CSOs in steering committees being 2.783 (p ¼ 0.005).The results support Hypothesis 1: CSOs have the potential capacities to contribute to the EU Cohesion Policy's implementation.A public servant from one Czech city confirmed the following: 'CSOs provide an original view on problems and naming new social phenomena' (INT-C20).We also tested whether there is a relationship between the added value provided by the CSOs and whether they are considered as partners.High capacities will be a combination of the low role of CSOs as partners and the high added value provided by CSOs when contributing to the IUDPs.In contrast, low capacities will be a combination of the high role of CSOs as partners and the low added value by CSOs to IUDPs.The chi-square tests show that none of the two combinations are represented in the data (model 3: Chi 2 ¼ 141.648, Pr ¼ 0.010; model 4: Chi 2 ¼ 12.450, Pr ¼ 0.013; and model 5: Chi 2 ¼ 12.750, Pr ¼ 0.013) as combinations of low added value with low partnership with CSOs and high added value with high partnership with CSOs were the most common in our data.
Moreover, the problems to be solved by the IUDPs also inform us about the capacities of the stakeholders.In Czechia, 32 of the 57 surveyed IUDPs are primarily oriented toward territories with major socio-economic problems and 25 territories that are poles of growth.Since there is a higher share of the territories with problems, they have limited capacities to solve the obstacles by themselves.The cooperation helped to find solutions.For example, one public servant supports the continuation of the partnership: 'collaboration with CSOs continues at a much better level than before due to the broader scope of the Integrated Territorial Plans' (INT-C23).
In the Czech IUDPs, the estimates of the CSOs' success in participating in the IUDPs in the growth poles in model 6 do not reveal significant results.The estimate of the CSOs' added value in problematic territories is higher than in poles of growth; however, the difference is insignificant (+0.526, p ¼ 0.128).In the Portuguese case, all 103 IUDPs cover socioeconomically problematic urban neighbourhoods.
Concerning the differences in capacities between the local and national CSOs, we can neither confirm nor reject Hypothesis 2. Our t-tests revealed that local CSOs add +0.623 points to the achieved added value (see model 7; on average 3.611 compared with 2.988 for IUDPs without local CSOs, p ¼ 0.033).The regional and national CSOs achieved even better results (see model 8: estimate +0.786, with an average score of 3.715 for IUDPs with regional and national CSOs, p ¼ 0.009).When testing the combination of participation of local CSOs with simultaneous participation of national and regional CSOs, we found an average estimate of 3.831 (model 9: higher by 0.405 than IUDPs without such a combination, p ¼ 0.278).We would expect a positive impact of sharing knowledge concerning the local needs and professional capacities of national CSOs.Although the direction of the estimates points to the same, the result is not statistically significant.
Is it possible for CSOs to contribute to the preparation of IUDPs?CSOs were perceived as the providers of information.They possess knowledge about the local needs of particular target groups.Nevertheless, in some cases, their role was no longer considered important when the IUDPs were accepted for funding (Liberec and Ústí nad Labem).After the CSOs provided the public sector with information, they no longer participated in the IUDP implementation.One of the interviewees from a CSO stated: We provided information about the actual needs of the inhabitants of the given locality.Though there was no intense mutual communication … the local CSO has only a few active members and there is no community with intense relationships like in the suburbs or in neighbouring villages. (INT-C11) However, there were cases where the municipality tried to communicate more intensively and sought new ways of cooperation with the CSOs, although they may not have worked out successfully.In some Czech cities, CSOs were directly invited to participate in the preparation of the IUDP (Most).They participated in the preparation of the IUDP but did not submit any project within the IUDP's implementation.In Ústí nad Labem, politicians represented CSOs in the advisory bodies; therefore, it is unclear whether they represented political parties or the CSOs.This may indicate that CSOs have been accepted only as formal partners.
Conversely, it may also point to the growing capacity of CSOs, with members who actively behave like politicians but are connected with the CSOs.In the Portuguese case, the CSOs were consulted by a few municipalities, mostly in local forums and municipal advisory bodies, particularly during the implementation phase.In general terms, their active contribution to this phase, however, was not especially effective in moulding the IUDP strategies.
Instead, it pinpointed responses to concrete social inclusion interventions.
CSOs benefit from the fact that they identified projects for which funds were subsequently drawn.Therefore, it was an ex-ante verification of the financial absorption capacity, although it did not work everywhere (Most and Třebíč).
The estimates of differences between CSOs' capacities in Portugal and Czechia (Hypothesis 3) reveal that in Portugal, the role of CSOs in constituting IUDPs is higher than in Czechia; however, the difference is not significant (model 10: difference +0.605, p ¼ 0.261).Thus, the length of EU membership does not have any effect on the success of CSOs in shaping local strategies.Our estimates are statistically insignificant and point to the higher success of the Portuguese CSOs in adding positive value to IUDPs compared with those from Czechia (model 11: +0.112,average 3.333 in Portugal,3.221 in Czechia,p ¼ 0.770).
Do CSOs want to contribute to IUDPs?CSOs' willingness to participate in the preparation of IUDPs varied greatly between cities.The willingness to collaborate was highly dependent on engaged individuals, regardless of whether they were organized in CSOs.People generally have a weak interest in public affairs in Czechia and Portugal.This willingness is generally related to low trust and social capital in the society in both countries (Falanga, 2018;Potluka et al., 2019).The interviews showed similar tendencies, since occasionally, participation was very low (e.g., Most and Liberec).For example, one public servant from Czechia stated: The response of the inhabitants of the city to the IUDP was minimal.The public hearing was attended by approximately 20 citizens from the professional and lay public.During the public hearing, there was only one contribution concerning materials used in the renovation of the public spaces. (INT-C14) At the local level, CSOs are willing to help with providing information to design IUDPs.Moreover, CSOs are more transparent to local politicians than engaged individuals.If the CSOs do not see the results of their efforts, they will similarly give up on the national level (Potluka et al., 2017b).In certain instances, CSOs find the alignment of their agendas with the public entities' agendas problematic in terms of their time and resources.CSOs find project deadlines challenging, as they tend to take excessive time to be accomplished (Fundação Calouste Gulbenkian, 2015).One of the typical Portuguese response documents stated that: The external participation process was affected by the fact that the deadline to finalise the IUPD strategy was due in the summer vacation time. (INT-P3) The CSOs and the citizens found it difficult to follow the implementation of the IUDPs, due to a lack of appropriate governance mechanisms. (INT-P5)

DISCUSSION
Czechia has a centralized public sector system which frames partnership implementation at all levels of the public sector.It is a long-term issue of fierce political debates that started during the transition period in the 1990s (e.g., Potuček, 1999; for a more recent analysis, see Salamon et al., 2017).This tradition of reduced collaboration between the public sector and other stakeholders in Czechia supports the conclusion of Polverari and Michie (2009, p. 49).They point out that, 'Effective socio-economic and other non-governmental organizations' participation is also greater where structural Funds are being implemented in countries with a broader tradition of socio-economic and other non-governmental organizations' involvement in domestic public policy more generally.'The situation in the surveyed countries has not changed since then.Indeed, Portugal is also marked by a long tradition of a centralized public sector.However, the local (municipal) level has democratic legitimacy and an active role in urban planning, collective equipment, and basic infrastructure (OECD, 2008), similar to Czechia.
The variety in the quality of the design and implementation of the IUDPs reveals that CSOs' capacities are highly individual.The vast majority of local CSOs take EU funding procedures as an additional burden and have capacities to only contribute on a limited scale beyond their usual activities.This high variability in CSOs' capacities causes these organizations to play a secondary role in preparing strategies for the IUDPs (van der Zwet et al., 2017).Only a few of CSOs will be capable of designing and implementing programmes of such scale.Moreover, investment in public properties pre-defines the public sector as the primary stakeholder responsible for designing and implementing IUDPs.Even municipalities can suffer from a lack of capacity.In such cases, consultants play a crucial role in designing IUDPs, which is not unusual (van der Zwet et al., 2017, p. 57).CSOs do not generally have the financial capacity to afford consultancy services and typically rely on their capacities.This lack of expertise, know-how and skills from CSOs in certain domains is clearly highlighted in the current literature.
The limited administrative capacities of the CSOs concern all four phases of the programming cycle, according to Milio (2007).An improved situation relates to the capacities in management and programming (see Table B2 in Appendix B in the supplemental data online).Monitoring and evaluation are fields in which local CSOs suffer the most from a lack of administrative capacities, especially expert knowledge.These results are aligned with the conclusions from recent research on the challenges related to CSOs' evaluation capacities.
For CSOs, monitoring and evaluation capacities are less important than management and programming.Even watch dog CSOs are skilled in monitoring policy implementation, but are usually less adept at evaluating them.At the local level, such skills are even less necessary.The situation in the surveyed IUDPs thus provides us with 752 Oto Potluka and Eduardo Medeiros

REGIONAL STUDIES
low capacities in CSOs in these fields (see Table B3 in Appendix B in the supplemental data online).

CONCLUSIONS
Our contribution relates to whether CSOs have sufficient capacities to contribute to the EU Cohesion Policy's implementation.In particular, this article contributed to the extant literature by unveiling how CSOs have contributed to the implementation of IUDPs in two EU member states with quite different territorial governance idiosyncrasies, but which share reduced administrative capacities among public sector managing authorities.Hence, the research is also linked to a growing body of existing literature on administrative capacity.We find that capacities in CSOs are not a general barrier to effective participation.The primary role of CSOs is not in political decision-making.Instead, they provide information to the public sector on local needs.The participation of CSOs helps to increase the absorption capacity.CSOs do that by introducing the needs of their target groups to the programme designers, and thus, increase the relevance of the programmes.Then, CSOs apply for EU funding.
Meanwhile, we point out that stakeholders behave pragmatically.If they do not see an advantage in participation (e.g., limited compliance with their organizations' objectives), their willingness to participate decreases.However, the concern is the accessibility of the information on the possibility of participating in and accepting the CSOs by the public sector to be a part of design programmes funded by the EU.
Our research contributes to the validation of the theoretical concepts of capacity approaches in co-production and co-creation.In particular, it confirms that long-term learning and practice increases absorption capacity not only in the public sector, but also among CSOs.For CSOs, programming is a secondary activity in which they do not invest in besides their main focus.Therefore, we complement this conclusion by the involvement of national CSOs, which are stronger in terms of capacities than local CSOs.However, local CSOs compensate for this disadvantage with their vast knowledge of local needs.Therefore, national and local CSOs should work more closely together to fulfil the ideal vision of co-creation and co-production.
Our results do not provide us with significant insights into whether local CSOs have lower capacities than regional and national organizations.Nationwide active organizations are usually more professional than local CSOs, and thus, have better capacities.
Concerning our third hypothesis, we find a positive but insignificant relation to whether a longer EU membership makes a difference to the practices in implementing cocreation and co-production in EU Cohesion Policy.Specifically, in both countries, there is quite a positive opinion that CSOs have the capacities to influence policymaking at the local level.However, in Portugal, there is also a perception of the high possibility of CSOs influencing policies at the national level.This perception is not observed in Czechia.Moreover, Portuguese CSOs have a longer experience with this policy.Thus, they had the time to build expertise and the capacity to contribute to it more effectively and efficiently.
Based on the results, our recommendations relate to a combination of the local CSOs' knowledge and the professional capacities of nationwide CSOs.When implemented via a place-based approach, EU Cohesion Policy should support the collaboration of these two types of CSOs.
The limitation of our study is the size of our data sample.We have concentrated our efforts on the IUDPs publicly available.Thus, further research should test our results with a larger sample size as well as in other countries.

Table 1 .
Dimensions included in the analytical framework.

Table 2 .
Overview of the number of interviews and documents analysed.