Beyond Usual Suspects? Inclusion and Influence of Non-State Actors in Online Public Consultations in Croatia

Abstract

Despite the increasing use of various e-democracy tools in shaping new policies, there is still a general lack of empirical studies on the influence of non-state actors in online public consultations. This article addresses this gap in the academic literature by focusing on the case of Croatia, which may have relevant broader practical and theoretical implications due to the legally binding rules of institutional responsiveness to individual policy inputs received during e-consultations and the growing interest of citizens and various interest groups to get engaged in this form of policy dialogue. Drawing on the novel data set that includes the responses of 39 government bodies to 51,250 policy inputs of interest groups and individual citizens to online consultations during the first three years since the launch of the government consultation platform, the paper seeks to analyse the influence of different types of non-state actors on the outcomes of government-led online public consultations. Contrary to general expectations about the predominance of more resourceful interest groups, it is argued that individual citizens exert a noticeable influence on the results of online policy consultations of Croatian government bodies. It is also claimed that the specific design and patterns of online public consultations, especially improved responsiveness of government bodies, contribute to the pluralisation of interests, equalizing political representation, and empowering individual citizens and other new actors, beyond traditional interest groups and “usual suspects” in national decision-making processes.

1. Introduction

Over the last two decades, various forms of e-consultations have started emerging as important and cost-effective mechanisms of involvement of citizens and interest groups in policy-making processes. Compared to traditional modes of offline participation, e-consultation platforms may bring a number of benefits such as enhancing more interactive communication, improving efficiency and effectiveness, getting beyond traditional power players by expanding perspectives, diversifying the sources of input and preventing routine dominance of any interest group, increasing trust and establishing a new relationship with citizens and other stakeholders, improving civic literacy and addressing some causes of democratic deficit (Macintosh 2004; Tomkova 2009; Peters and Manon 2008; Coleman and Shane 2012). Unlike different forms of closed-access consultation instruments such as advisory committees, one-stop online consultation portals have the potential for ensuring higher inclusiveness and alleviating bias in the policy engagement of citizens, interest groups and various stakeholders (e.g., Quittkat 2011; Rasmussen and Carroll 2014; Bunea 2017; Roed and Hansen 2018). Some research on the use of digital tools for public engagement in policymaking point out that the low entry costs and non-hierarchical culture of digital tools may empower marginalised and resource-poor organisations and groups (van der Graaf et al. 2016; Vesa et al. 2022). On the other hand, one of the recurrent concerns in studies of public consultations at national, EU or international levels is the general dominance of business interests which are often better equipped to provide specific technical policy inputs expected by public institutions (Binderkrantz et al. 2015; Yackee 2015; Hanegraaff and Berkhout 2019; Fraussen et al. 2020).

Despite growing positive expectations from the use of various e-democracy tools there is still limited empirical evidence on the influence of non-state actors1 on the outcomes of government-led e-consultations, as well as on the potential of online consultations to reinvigorate democratic processes by changing inequalities in the landscape of political representation among non-state actors in national policy-making processes. It is the general lack of institutional responsiveness and the absence of data on what happened with policy inputs of non-state actors during online public consultations that prevent a more convincing measurement of the influence of various stakeholders on the consultation outcomes (West 2004).

The article addresses this gap in the academic literature by focusing on the single country study and using the empirical data from Croatian online public consultations. The findings from the Croatian case may have broader practical and theoretical implications, especially in view of the legally binding rule of institutional responsiveness to individual policy inputs received during public consultations and the growing interest of individual citizens and various interest groups to get engaged in this form of policy dialogue.

The following research question will be explored: What is the influence of different types of non-state actors on the outcomes of government-led online public consultations in Croatia?

Contrary to general expectations about the routine dominance of more resourceful interest groups, the paper argues that individual citizens exert a noticeable influence on the outcomes of open online public consultations of Croatian government bodies. It is also claimed that the specific design and patterns of online public consultations, especially improved responsiveness of government bodies, contribute to equalizing political representation, pluralisation of interests and empowering individual citizens and other new actors, beyond traditional interest groups and “usual suspects” in national decision-making processes.

In the first part, the conceptual framework for discussing the measurement of the influence of non-state actors in online consultations will be presented, along with some findings from previous research in this area. The next section brings a description of the research design and methodology. This will be followed by an overview of the evolution and main features of the Croatian e-consultations platform. The central part of the paper focused on the analysis of different types of contributors to e-consultations as well as responses of Croatian government bodies to received policy inputs, followed by the discussion on the influence of various categories of non-state actors on the consultation outcomes. Finally, concluding remarks will be presented along with a summary of the main findings and reflections on broader implications of the analysed case in both theoretical and practical terms.

2. Literature Review and Diverse Approaches to Analysing the Inclusion and Influence of Non-State Actors in Online Public Consultations

There are different approaches to e-consultations as an increasingly popular instrument of advancing the engagement of citizens and interest groups in public policy-making. E-consultations may substantially vary in goals, target groups, issue areas, use of technical tools, design of the procedures, initiators, owners, or administrators of the platforms as well as the administrative level at which they are launched. Whether they were initiated by public institutions, civil society organisations or jointly, a common purpose of online consultation platforms is to gather ordinary citizens, civil society actors, experts and other stakeholders to provide input, deliberate, inform and influence policy and decision-making (Coleman and Gøtze 2001; Macintosh 2004; Tomkova 2009; Medaglia 2012; Susha and Grönlund 2012). When it comes to the place of e-consultations in the cycle of public policy-making, they are primarily carried out in the policy formulation phase, before actual decisions are taken, to strengthen the quality and legitimacy of policies (Albrecht 2012). The European Commission defines consultations as “those processes through which [it] wishes to trigger input on policy from interested parties prior to issuing its decisions” (European Commission 2002). According to various international models of classification of participation levels (OECD 2001; Council of Europe 2009, 2017), e-consultations are one of the digital tools of consultations that are considered the second level of civic participation allowing public authorities to collect the views of individuals, NGOs and civil society at large on a specific policy or topic as part of an official procedure.

The focus of this paper is on the government-administered online consultation platforms enabling two-way interaction between democratic public institutions and citizens and various interest groups with the intention to elicit inputs that contribute to more sustainable policies and provide opportunities to influence decision-making processes (Peters and Manon 2008; Panopoulou et al. 2014). It is the opportunity to influence formal, institutional decision-making processes that distinguish e-consultations from other virtual, informal spaces for horizontal interaction among citizens (Tomkova 2009). The increasing presence of government-led central consultation platforms in most countries around the world (United Nations 2020; OECD 2019) confirms the relevance of studying this topic.

While some authors assume that the government-owned and managed e-consultations platforms are more likely to be used by relevant government bodies and provide an adequate response to citizens’ inputs and are also more likely to motivate public and interest groups to participate (Jensen 2003; Peixoto and Fox 2016), others draw attention to various concerns and weaknesses of these formal e-consultation instruments, often related to complex societal, administrative and organisational factors which may lead to failures in mobilising citizens’ engagement (Ostling 2010; Zheng et al. 2014; Toots 2019). Some scholars also point out the asymmetry of influence of actors in e-consultations, since the government bodies mostly initiate and manage e-consultations, and the actors involved substantially differ in the level of their authority, expertise, and potential for access to decision-making processes (Tomkova 2009).

The lack of adequate institutional response to receive inputs is the most frequent source of criticism of e-consultations tools. The responsiveness of government to the preferences of citizens is generally considered to be an essential feature of democratic governance and a central indicator of the quality of policy dialogue. Without proper feedback on the received inputs and adequate integration of evidence into the policy process, public participation cannot be considered truly meaningful (Dahl 1989; Pateman 1970; Coleman and Gøtze 2001; Buckwalter 2014) and can only result in an ever-growing public distrust towards the government. The increasing focus on low administrative capacities and human resources as reasons for weak government responsiveness to policy inputs through online consultations has also opened the space for debates on the use of new technologies for facilitating and improving the institutional response. Natural language processing is one of these recently tested technologies enabling automatic summarising of discussion threads in large-scale debates and potentially facilitating the preparation of an overview of online policy consultations. However, the evaluation of these technologies indicates that despite possible assistance provided by these technologies, they still cannot substitute human effort and role in presenting the outcomes of consultations and shaping policy recommendations (Albrecht 2012; Tigelaar et al. 2010).

Government or institutional response can be defined as a clearly identifiable action taken by the government, following individual or collective input by citizens or interest groups. Direct government response to an act of e-participation (submission of comments on draft laws through online platform e-consultations) may be considered an example of objective government efficacy in dealing with citizens’ and interest groups’ inputs (Peixoto and Fox 2016). Some literature on participatory democracy (including participatory budgeting), as well as research on the efficacy, indicate that levels of participation are intrinsically linked to institutional responsiveness. The higher the government’s ability to respond is, the higher citizens’ and interest groups’ willingness to participate in the future will be (Buckwalter 2014; Peixoto and Fox 2016; Sjobert et al. 2017). Besides, a positive impact of online consultations on broadening the scope of participants in public debates has been confirmed by some empirical research which highlighted that groups who are not participating by traditional means are indeed represented in the online debate (Albrecht 2006).

The limited number of studies on the effectiveness of citizens’ and interest groups’ engagement in e-consultations and the level of their influence on the consultation outcomes is mostly due to the low institutional responsiveness and the general lack of empirical evidence on accepting or rejecting policy inputs received public consultations (West 2004; Naughton et al. 2009).

In general, assessing the influence of stakeholders’ policy engagement has traditionally been one of the most complex challenges for researchers in this area, mostly due to different dilemmas related to the definition and methods of measurement of influence which have generally prevented more systematic empirical investigations of the topic (Baumgartner and Leech 1998; Dur and De Bievre 2007; Beyers et al. 2008; Dur 2008; Kluver 2009; Pedersen 2013).

The concept of influence is usually defined as the ability of non-state actors to shape a political decision in line with their preferences or a “causal relationship between the preferences of an actor regarding an outcome and the outcome itself” (Nagel 1975, p. 29, as cited in Dur 2008, p. 561). In the context of this paper, the focus is on the ability of non-state actors to generate the public institutional acceptance of their policy preferences openly stated during online public consultations on draft legislative acts before they are submitted to the government for adoption.

The limited empirical research on the influence of actors in public consultations and measuring preference attainment has predominantly focused on online consultations, which generally offer more documented evidence and allows easier mapping of political preferences of various actors, often using quantitative text analysis (Rasmussen and Carroll 2014; Bunea and Ibenskas 2015; Kluver 2015). It is often assumed that the complex and technical nature of draft legislative acts can be demotivating for ordinary citizens to get engaged and provide input to public policy dialogue (Culpepper 2011), while there is a general concern that consultations are biased towards business (Yackee 2006; Ulbig 2008).

To the best knowledge of the author, the argument proposed in this paper about the noticeable influence of individual citizens on the outcomes of online public consultations has not yet been empirically supported in other studies. This argument needs to be considered in the very specific context of the design and specific patterns of online public consultations in Croatia, especially regarding the mandatory institutional responsiveness, which is claimed to contribute to equalizing political representation, pluralisation of interests, and greater inclusion of individual citizens and other new actors, beyond traditional interest groups and “usual suspects” in national decision-making processes. The findings from the case study of Croatia can have broader theoretical and practical implications, particularly for revisiting approaches to designing and managing the online public consultation process, the impact of individual responses vs. collective, summary responses of institutions, as well as analysing the role, behaviour, and influence of individual citizens vs. traditional interest groups in that process.

3. Methodology and Research Design

Since there are no previous studies on the influence of non-state actors in online public consultations in Croatia, this paper will have an exploratory nature with elements of generating new perspectives on the topic, developing ideas for future research, and also providing some food for thought for practitioners in this area.

The research methodology is based on the quantitative analysis of the novel data set that includes the responses of 39 government bodies to 51,250 policy inputs of interest groups and individual citizens to 1968 public consultations held via the central Croatian government e-consultations platform esavjetovanja.gov.hr, during the period of more than three years since its launch—from April 2015 until July 2018. The responses of the government bodies for all analysed consultations and each input are publicly available on the platform in the official reports of responsible institutions which include individual responses of government bodies to each received input. The data were gathered for the exploratory research on government responsiveness conducted during 2018 and were received from the responsible government office based on the request for access to information. This will be mostly a descriptive study, relying on the use of descriptive statistics.

In this paper, the focus is on the observable effects of the influence of non-state actors (individual citizens and interest groups) on the outcomes of online public consultations launched by government bodies during the process of formulation of legal and policy initiatives, before these are submitted to the government for adoption. The units of observation are the responses of government bodies to individual inputs of citizens and various interest groups to open online consultations on draft acts. The preferences of various categories of non-state actors are stated in their submitted inputs/comments on different articles/parts of draft acts and publicly available on the central Croatian government portal for public consultations, along with responses of competent government bodies in reports on the results of public consultations, next to each submitted individual policy input.

The term “input” is used for the comments of non-state actors on specific articles or parts of legal/policy initiatives submitted to online consultations, where the identity of each non-state actor is visible and publicly available next to the provided comment. The input submitted indicates the attempt of the non-state actor to change the draft rule and influence the policy formulation process.

The government response is defined as a manner in which governmental bodies publicly respond to inputs of non-state actors. The response to all received inputs is a legal obligation, while the system/software of the central consultation platform enables the standardised use of four possible types of responses to suggested changes: “accepted”, “partially accepted”, or “not accepted”. The “partially accepted” comments are used by government bodies usually for more complex comments with several proposals of changes related to one article/section of a draft legal act/policy, where only some parts of inputs/proposed changes are accepted. When an input/comment does not require making changes to the rule, the government body responds with “acknowledge” (or “noted”). This is the case, for example, when the non-state actor simply endorses the proposed rule or when the input is not related or goes beyond the scope of the consultation.

Since the government responses to “not accepted”, “partially accepted” or “noted/acknowledged” inputs need to be accompanied by an explanation that is permanently publicly available on the central e-consultation platform, there is an expectation that the accepted or partially accepted inputs of non-state actors will be integrated into the policy proposals sent to the government for adoption. In other words, the official government acceptance of the received input of non-state actors is considered a recognition of the change made to the draft proposal which will be sent to the government, and evidence of the influence of the non-state actor on the outcome of online consultation.

The structure of the reports on results of consultations contains the full name of the non-state actor (individual citizen, association, company, etc.,), the content of the input/comment on the selected article of the draft legal act or part of the draft policy, type of government response (accepted/partially accepted/not accepted/acknowledged) and explanation of the reasons behind the government response. This standardised content of the consultation report of responsible government bodies allowed an automated search algorithm to retrieve all observations required for the analysis in this paper.

There are however several methodological limitations of the focus on official government responses as key indicators of the influence of different non-state actors on the outcomes of public consultations. The first limitation is that this level of analysis does not consider the content of the submitted inputs, where there may be a difference between “technical” inputs which aim at the technical aspects of the proposal, and “substantial” inputs which aim at substantially changing the policy direction or instruments. The next limitation is related to the lack of statistical data on the individual responses to inputs of citizens and interest groups prior to the launch of the central e-consultation portal, which prevents the comparison of the degree of inclusion and influence of non-state actors before 2015. Finally, the methodology applied in the paper is focused on the stage of policy formulation during which online public consultations take place after the first draft has already been produced, which does not cover the previous stage of legislative/policy drafting where inputs of certain non-state actors might have already been integrated into the draft that is submitted to online consultations (Lowery 2007; Naughton et al. 2009). Given the number of the inputs included in the analysed data set (51,250), the bias is expected to be approximately equally distributed between the non-state actors and should not threaten the validity of the conclusions.

4. Evolution of Croatian E-Consultations

Central Government e-consultation portal esavjetovanja.gov.hr was launched in April 2015, as a single web access point to all open (and closed) public consultations launched by state bodies in Croatia.

The launch of the e-consultations portal proved to be an important landmark in the development of e-government initiatives and the introduction of more inclusive policymaking in Croatia. This was acknowledged by the substantial increase in the country’s ranking in the United Nations e-Participation Index—from 97th place in 2014 to 25th place in 2016 (United Nations 2020). The functioning of the e-consultation portal has been also praised in the OECD review of regulatory performance in Croatia for 2019 (OECD 2019). where Croatia has been scored among the top three EU member states regarding stakeholders’ engagement in developing both primary and secondary legislation.

The launch of the e-consultations platform was preceded by six years-long efforts to implement The Code of Practice of Public Consultations in Procedures of Adopting Laws, Other Regulations and Acts (Official Journal, 140/2009). These efforts included appointing and training consultation coordinators in all government bodies, introducing regular and systematic training on public consultations open for all civil servants at the State School for Public Administration, standardizing templates and practices of reporting on the results of public consultations, and publishing annual reports on public consultations (Vidačak 2013). According to the official government data, before and after the launch of the central e-consultations platform, more than eighty training sessions were held for all state bodies, to make sure they are all acquainted with the modalities of use of the new central online consultations tool2. The most important innovations introduced by the platform are open and permanent access to all received policy inputs of registered users, as well as to responses of government bodies to these contributions, allowing open monitoring of the quality of institutional responsiveness; email notifications of newly opened consultations; a simple way of commenting on draft laws, regulations and acts, user-friendly overview of the legal proposal under consultation, article-by-article commenting and a real-name identification of contributors.

Since the adoption of the above-mentioned Code on Practice of Consultations in 2009, and despite its legally non-binding nature, expectations of better-quality government response to policy inputs of stakeholders started growing (Vidačak and Škrabalo 2014). Nevertheless, the low number of conducted public consultations in 2010 and 2011, as well as the limited institutional response to received inputs, was due to the very accelerated legislative dynamics in the last years of the EU accession negotiations of Croatia. This was reflected in more than 86% of laws adopted through an urgent parliamentary procedure, which has significantly narrowed the space for the meaningful participation of non-state actors in the process of policy formulation during the EU accession process (Vidačak 2013). In 2012, after the conclusion of the EU accession negotiations, the new centre-left government brought much-needed political will for more advanced implementation of standards of public consultations. This was manifested in establishing a dynamic online communication team that launched a proactive campaign on open consultations via social networks, as well as in the adoption of necessary legal changes needed to prescribe the obligation of submitting consultation reports for all acts of public interest passed by the Government, as well for draft laws sent to the Parliament. In addition to the amendments to Government Rules of Procedures in 2012 and Parliament Rules of Procedures in 2013, the new Law on the Right of Access to Information was adopted in 2013 which, among others, prescribed the obligation for all government bodies to publish reports on results of consultations with an overview of accepted and rejected inputs. In 2013, the Constitutional Court abolished two implementing acts of the Ministry of Education due to non-compliance with public consultation standards, stressing in its Decision that: “… the democratic nature of the procedure under which takes place public dialogue on issues of common interest is what makes an act, as a result of these procedures, constitutionally acceptable or unacceptable”3. This Decision of the Constitutional Court had a far-reaching impact on political decision-makers and contributed to raising awareness of the importance and values of public consultations (Vidačak 2020).

The analysis of data from annual government reports on the rise of contributors to online public consultations via websites of ministries during the period 2011–2014 (before the launch of central e-consultations) offers interesting findings. As shown in

Table 1. Increase in the number of contributors to public consultations conducted via individual websites of government bodies from 2011 to 2014.

Year	2011	2012	2013	2014
Number of contributors to open consultations	171	4786	8299	7482

Source: Annual Croatian Government Report on conducted public consultations in 2014.

, in only two years the number of stakeholders contributing to public consultations increased by more than 4000%.

The more advanced analysis would enable determining various factors that have influenced the rise of interest of citizens in online public consultations. One of the possible incentives could be the introduction of the legal obligation for government bodies to publish the individualized, elaborated responses to all received written inputs of contributors to online consultations, along with an elaboration of reasons for not accepting certain comments.

In parallel with the launch of the central e-consultations platform in 2015, the Law on the Right of Access to Information was amended (Official Journal 85/2015) to lay down the use of the central e-consultation platform as a legal obligation for all government bodies when launching new public policy consultations. From 2015 until 2019, the Platform was coordinated and managed by the Croatian Government Office for Cooperation with NGOs, one of the government policy coordination units responsible to the Prime Minister, while from 2019 the platform management has been taken over by the Government Office for Legislation (responsible also for regulatory impact assessments) and Central Government Office for Digital Society. In addition, the independent Commissioner for Access to Information, appointed by the Parliament in 2013, has been overseeing the compliance of public administration bodies with standards of public consultations.

All this contributed to a substantial increase in the number of public consultations conducted by public administration bodies—from only 144 in 2012 to 1033 in 2018. This was also followed by a significant increase in the number of users of the central e-consultation platform—from 4000 in 2015 to 23,214 in 2018 (savjetovanja.gov.hr). In addition to the growing number of conducted public consultations and mobilisation of new users, the launch of the central e-consultations portal has contributed to the easier preparation of consultation reports but also ensured permanent access to the evidence on responses of institutions to individual contributions.

One of the negative consequences of the increased use of e-consultations is a continuous decrease in the use of other methods of public consultations, especially open, face-to-face meetings, as can be seen from

Table 2. Trends in use of e-consultations by Croatian government bodies, compared to other public consultation methods (2015–2018).

Method of Public Consultation	2015	2016	2017	2018
E-consultations	63%	73%	85%	86%
Open public meetings	25%	14%	3%	3%
Closed consultative sessions	12%	10%	10%	7%
Other methods	4%	3%	2%	4%

Source: author, based on annual government reports on conducted public consultations 2015–2018, available at: savjetovanja.gov.hr.

.

The diversification of public consultation methods remains one of the main challenges and further efforts are needed to ensure better accessibility of open consultations to various types of vulnerable or marginalised groups, especially in remote and underdeveloped regions.

5. Types of Contributors and Responses of Government Bodies to Policy Inputs Received through the E-Consultation Platform

There are 19 categories of users registered at the central e-consultations platform: individual citizens, companies, NGOs4, chambers5, crafts, employers’ associations, trade unions, cooperatives, institutes, artist organisations, religious communities, family farms, foundations, the academic community, political parties, state bodies, local and regional government bodies, international organisations, others. Individual citizens account for more than 80% of all registered users6.

The statistical data gathered for this research from the Croatian Government e-consultations portal for the period from April 2015 until July 2018 cover 51.250 inputs of 12,535 individual contributors (citizens and interest groups) sent to 39 government bodies (ministries, central state/government offices, agencies, etc.,) during online consultations on 1968 draft acts published at e-consultation platform during that period. The overview of the types and number of contributors to online public consultations in

Table 3. Overview of types of contributors to online public consultations via the central government online consultation portal esavjetovanja.gov.hr (April 2015–July 2018).

Type of Contributors to E-Consultations	Number of Contributors to E-Consultation by Years
	(from April 2015)	2016	2017	2018 (until End July)
Citizens	571	2983	4327	2173
NGOs	110	220	307	201
Companies	88	160	218	164
Chambers	12	17	25	18
Trade unions	5	5	8	5
Employers’ associations	0	1	4	1
Crafts	3	13	58	12
Institutes	85	121	146	79
Religious communities	2	5	2	1
Academic community	12	20	22	11
Artist organisations	1	2	2	1
Family farms	0	3	8	8
Cooperatives	1	1	3	2
Foundations	3	0	5	1
Political parties	1	1	4	1
Local/regional government	56	38	52	33
State bodies	13	11	14	14
International organisations	0	1	1	2
Others	3	13	10	10

Source: Author, based on data from savjetovanja.gov.hr, 2018.

includes also (for statistical purposes) some categories of state bodies such as local regional bodies or state agencies that are not in the focus of this analysis. As can be seen from the table, individual citizens are by far the most numerous contributors to public consultations, followed by NGOs and companies. Moreover, the visibility of real names of all individuals and organisations and their comments (in chronological order of submission to the e-consultation portal) contributes to the perception of increased pluralisation and equalisation of interests in the policy development process.

Citizens appear as by far the most numerous contributors in online public consultations launched by the Ministry of Science and Education (32% of all contributors from 2015 until 2018) where, traditionally, the most comments are coming from individual teachers in primary and secondary schools who seem to be the most active contributors at e-consultation platform (esavjetovanja.gov.hr).

Table 4. Types of government responses (N = 51,250) to citizens and interest groups’ policy inputs via central e-consultations platform from April 2015 to July 2018.

Type of Government Response	Accepted Fully	Accepted Partially	Not Accepted (Rejected)	Acknowledged	No Response
Number of inputs	7188	4195	14832	18453	6582
% of all received inputs	14.1%	8.2%	28.9%	36%	12.8%

Source: Author, based on official data from e-savjetovanja.gov.hr, 2018.

provides an overview of the frequency of the four types of responses of each government body to individual inputs of citizens and interest groups to specific articles/parts of draft legal or policy initiatives (1) accepted, (2) partially accepted, (3) acknowledged, and (4) not accepted. As already stressed, the response to each individual input needs to be publicly accessible and accompanied by an explanation providing the contributor with an elaboration of reasons for not taking the proposed input into account. The data in the table below cover also inputs that received no response, mostly in cases when draft laws or regulations were withdrawn from the procedure, or where the consultation reports have not yet been published by July 2018.

If both fully and partially accepted inputs are considered indicators of success or influence of citizens and interest groups on the outcomes of public consultations during the process of formulation of new public policies, then a 22.3% success rate may still send a positive message that engaging in this form of e-participation is meaningful. Some preliminary findings from previous research on the quality of response of Croatian government bodies to policy inputs in the field of environmental policy pointed out the difference between accepting substantial (two-thirds of all inputs) and technical inputs (one-third of all inputs) of interest groups and citizens, where technical inputs often refer to harmonizing terminology and do not address the substance of policy proposals (Vidačak and Đurman 2017, p. 96). Here is it is important to mention that the Croatian e-consultations platform enables more structured consultation on specific parts of proposed texts of laws, regulations, strategies or other acts, and encourages citizens and interest groups to focus on proposed text and formulate precise suggestions, which may be an additional factor encouraging a more productive and meaningful policy dialogue.

When it comes to one-third of overall policy inputs (36%) that received the response “acknowledged”, previous research indicates that this type of response is used predominantly for inputs that are formulated as general concerns, questions, or inputs that do not refer precisely to the text of policy proposal or provisions of the law, while part of these inputs contained, among others, concrete suggestions that deserved more precise reply (ibid.). Even though the permanent visibility of received inputs and institutional response puts additional pressure on government bodies to prepare a reasonable reply, the “over-use” of the response option “acknowledged” indicate that there is substantial room for improvement in the quality of government response to citizens and interest groups’ inputs.

6. Assessing the Influence of Various Types of Non-State Actors on Outcomes of Online Consultations

The analysis of responses of 39 government bodies to 51,250 policy inputs of all mentioned categories of registered users to 1968 draft legal and policy initiatives in the first three years of the use of the central e-consultations platform (from April 2015 until July 2018) has highlighted the high interest of individual citizens to participate in the government-led online consultations, as well as their potential to influence the consultation outcomes (please see

Table 5. Overview of responses of government bodies to policy inputs of various types of non-state actors via central e-consultations platform, from April 2015 until July 2018.

Type of Contributors to E-Consultations	No of Fully Accepted Inputs	No of Partially Accepted Inputs	No of Rejected Inputs	No of “Acknow-ledged” Inputs	No of Inputs with No Response 1	Total No of Inputs	No of Fully or Partially Accepted Inputs	% of All Fully or Partially Accepted Inputs	% of Accepted Inputs in All Sent Inputs
Citizens	3726	2074	7523	12,114	3792	29,229	5800	50.95%	19.84%
NGOs	1085	593	2665	2210	890	7443	1678	14.74%	22.54%
Companies	1117	590	1963	1170	660	5500	1707	15.00%	31.03%
Chambers	341	286	800	687	165	2279	627	5.51%	27.5%
Trade unions	36	21	95	30	3	185	57	0.50%	30.81%
Employers’ associations	6	7	23	12	1	49	13	0.11%	26.53%
Crafts	76	134	215	139	6	570	210	1.84%	36.84%
Institutes	311	201	649	937	628	2726	512	4.50%	18.78%
Religious communities	7	3	4	17	0	31	10	0.09%	32.25%
Academic community	28	29	69	222	98	446	57	0.50%	12.78%
Artist organisations	1	0	1	11	3	16	1	0.01%	12.5%
Family farms	10	5	19	15	6	55	15	0.13%	27.27%
Cooperatives	2	3	1	4	2	12	5	0.04%	41.66%
Foundations	5	2	12	8	0	27	7	0.06%	25.92%
Political parties	3	7	14	5	0	29	10	0.09%	34.48%
Local and regional government bodies	226	144	484	633	207	1694	370	3.25%	21.84%
Central government bodies	147	87	251	204	106	795	234	2.06%	29.43%
International organisations	0	0	2	3	1	6	0	0.00%	0%
Others	61	9	42	32	14	158	70	0.61%	44.30%

Source: author, based on official data from esavjetovanja.gov.hr, 2018. ¹ This also includes data for 2018 where for a number of consultations final reports on consultations outcomes have not yet been published. Therefore a higher number of “no response“ to received inputs.

).

As shown in Table 5, individual citizens are by far the most active users of the e-consultation platform with more policy inputs than all categories of various interest groups together. That is, in itself, an encouraging finding as this form of e-participation can certainly contribute to improved civic literacy and strengthening active citizenship. When it comes to the overall number of inputs, individual citizens are followed by NGOs, companies, institutes, chambers, local and central government bodies, crafts, the academic community, and trade unions. The number of policy inputs by NGOs is relatively high compared to other interest groups, but, on the other hand surprisingly inadequate, given the fact that the number of NGOs registered on the e-consultations platform is only around 1.5% of all registered NGOs in Croatia.

A very active direct involvement of individual companies in attempting to influence policy formulation through e-consultations is an interesting finding, as it shows the tendency of by-passing of national business associations (Croatian Employer’s Association and Chamber of Commerce) as “usual suspects” and traditional power players. The growing interest of professional chambers, institutes and crafts in contributing to this form of policy dialogue is uncovering new advocacy potential of professional communities that are usually not that visible in policy development processes in Croatia.

An increasing number of policy inputs of local and regional government bodies is not surprising given the overall number of Croatian counties, cities, and municipalities, but also given a high number of legal acts where local and regional government units figure as decentralized implementors of policies adopted at the State level. On the other hand, it is rather unexpected to see a relatively high level of contributions of central government bodies via e-consultations, as it may indicate a still inadequate level of policy coordination, poor inter-ministerial consultations, and low involvement of formal institutional actors in the earliest phases of agenda-setting as well as drafting policy proposals in working groups. A still very low interest of trade unions and employers’ associations in contributing to e-consultations may be a consequence of traditionally privileged access to top decision-makers these actors enjoy as one of the key social partners in tripartite social dialogue with the government. The low involvement of the academic community in e-consultations is a discouraging finding, considering the expected role and potential of universities and scientific, research institutes in raising the quality of public debates on important policy reforms in the country.

If the number of policy inputs that are fully or partially accepted by government bodies is considered as the main criteria of influence through e-consultations, then individual citizens are emerging as the most influential actors in the formulation of policies, with the share of almost 51% in all policy inputs accepted by competent State institutions. A more thorough qualitative analysis of accepted contributions and the scope of resulting changes in policy proposals would give a broader understanding of the impact and effectiveness of individual citizens in this form of e-participation. Nevertheless, this finding may shed new light on the potential of the direct influence of citizens on policymaking processes, beyond usual suspects and traditional power players that have been identified in previous research as dominant in policymaking processes (Petak et al. 2019). Companies, with a share of 15%, and NGOs with 14.7% share are the next most influential actors in policy formulation via e-consultations, followed by chambers (5.5%) and institutes (4.5%), local and regional government bodies (3.25%), central government bodies (2.06%) and crafts (1.84%), while other actors have a share below 1% in all accepted policy inputs.

A somewhat different picture of the involvement and influence of citizens and interest groups in policy formulation via e-consultations can be observed if the overall number of fully and partially accepted policy inputs of different categories of e-consultation users are compared to the overall number of inputs sent by them. Among the non-state actors with the higher rates of participation in e-consultations (more than 1000 inputs in three years), companies stand out with 31% of successful policy inputs, followed by chambers (27.5%), NGOs (22.5%), and individual citizens (19.8%).

The results of a recent qualitative survey on the attitudes on the e-consultations platform conducted among civil servants as well as various categories of non-state actors -contributors to e-consultations (NGOs, companies, chambers, academic community, etc.) indicate that more than 80% of contributors consider their influence to be low or very low (Đurman 2021). The same survey findings show that civil servants consider the inputs submitted via e-consultation from the following stakeholders to be the most useful: other state/public administration bodies (53.2%), NGOs and the academic community (49.4%), business sector (38%) and individual citizens (32.5%).

The success of advocacy through e-consultations as an open access consultation instrument should be seen only as one of the indicators of influence in policy-making. For a more comprehensive analysis, advanced qualitative research on the engagement of interest groups in different formal, institutionalized bodies (expert working groups, committees, etc.), in the earlier phases of drafting legislative/policy initiatives, as well as informal lobbying should be conducted.

The observed emergence of individual citizens as increasingly engaged and influential actors in Croatian online consultations can also be further explored in the context of different factors which affect the patterns of the use of Croatian e-consultations portal, greater institutional responsiveness in online consultations, and the behaviour of citizens and interest groups in that form of structured policy dialogue.

Additional qualitative research should be conducted to further examine the factors determining the inclusion and influence of non-state actors in online public consultations, as well as the degree and quality of institutional responsiveness to received policy inputs.

Based on the available data from official annual government reports on conducted public consultations (savjetovanja.gov.hr) as well as findings from semi-structured interviews conducted from May to July 2019 with an expert sample of ten representatives of government bodies, NGOs, and businesses using the e-consultation portal, several factors seem to be particularly relevant for the inclusion and influence of citizens and various interest groups, as well as for ensuring more open and transparent response of government bodies to received policy inputs during online public consultations. These factors include, among others, mandatory use of central government online consultation platform by responsible government bodies; legal obligation of institutional response to received inputs; capacities of institution managing the platform; real name policy instead of anonymity of users/contributors; clarity of rules of mutual vertical interaction of users and institutions; full disclosure of received policy inputs and institutional responses, as well as individualized response with the elaboration of reasons for not accepting certain policy inputs (Vidačak 2020). The mentioned factors which affect the patterns of behaviour of both institutions and non-state actors in online consultations are not exhaustive and further research and surveys among a wider spectrum of actors would be needed to evaluate the relative weight of each factor.

7. Concluding Remarks

The paper aimed at exploring the influence of different types of non-state actors on the outcomes of online public consultations conducted through the Croatian government’s central e-consultations platform.

Based on the analysis of a unique dataset of responses of government bodies to received policy inputs of citizens and interest groups to open consultations on draft acts during the first three years of functioning of the e-consultations portal, individual citizens are confirmed as the most numerous participants in e-consultations. The policy area which attracts by far the biggest number of inputs from citizens is education and science.

Depending on the criteria used for the assessment, different perspectives on the influence of various non-state actors in Croatian online public consultations can be observed When considering the share in the overall number of policy inputs that are fully or partially accepted by government bodies as the main indicator of influence through e-consultations, then individual citizens emerge as stakeholders with the relatively biggest influence on the outcomes of e-consultations, followed by companies and NGOs. On the other hand, if the overall number of fully and partially accepted policy inputs of different categories of e-consultation users is compared to the overall number of inputs sent by them, the “usual suspects”—business actors (companies and chambers of commerce) and NGOs stand out as the most successful participants, followed by individual citizens. A more thorough qualitative analysis of accepted contributions and the scope of resulting changes in policy proposals would give a broader understanding of the influence of individual citizens compared to other stakeholders in this form of e-participation.

The ample evidence provided through mandatory responses of government bodies to each received individual input can enable more advanced research focused on several explanatory variables that could be used to highlight the influence-seeking behaviour of various non-state actors and different types of institutional responses to their contributions via the e-consultations platform. The findings from the Croatian case could potentially have broader practical and theoretical implications, especially in terms of possible causal links between improved institutional responsiveness and the increased interest of citizens and various interest groups to participate in e-consultations as open-access instruments of policy engagement. Further studies would be helpful to highlight the factors that may generate higher participation of citizens in e-consultations, particularly the impact of the individualized versus collective (summary) government responses to received policy inputs, the real name policy versus anonymity of users/contributors, etc.

Overall, the design of the e-consultations platform seems to have contributed to the increased pluralisation of interests represented in Croatian policymaking processes and opened new opportunities for strengthening the voice of individual citizens and less resourceful organisations in decision-making arenas that are usually dominated by traditional power groups. On the other hand, the increased use of e-consultations has been followed by a continuous decrease in the use of other methods of public consultations, especially open, face-to-face meetings. Therefore, the diversification of public consultation methods should be considered one of the main priorities in order to ensure the inclusion of various types of vulnerable and marginalised groups.