One of the main motivations for measuring the weakness or strength of civil society is to obtain a reliable basis for understanding the dynamics of its development as well as its social and political potential. In this article, we argue that the paradox of the “weakness of civil society” in Central and Eastern Europe can be explained by insufficient methodologies involved in what we call the static approach to the strength (or weakness) of civil society. We present a more appropriate alternative, called the dynamic approach, in which weakness or strength is not an inherent property of a civil society and consequently cannot be measured by a set of indicators collected for a single point in time. Moreover, in a dynamic approach, the weakness or strength of civil society is a derivative of the dynamics of its development over time along multiple axes of indicators. In other words, we propose that the weakness or strength of civil society ought to be conceived of as the ratio of its development over time and that it must be evaluated inside a data-rich environment where comparison over time is possible.
Introduction: Approaches to Strength and Weakness of Civil Society in Central and Eastern Europe
For political scientists, Central and Eastern Europe (CEE) was the region that in the early 1990s was undergoing a rapid transition from communism to liberal democracy although it became evident very quickly that there were many different paths to democratization, some less and some more successful. This also applied to civil society and its role in democratization. In the case of CEE countries in the immediate post-communist period, some scholars assumed that civil society had a strong democratizing potential or at least was a necessary arena for democratic consolidation (Linz & Stepan, 1996). When this expectation proved to be problematic, more and more attention was given to the phenomenon of a “weak post-communist civil society” (Howard, 2003, pp. 14–15) and diverging paths of CEE countries during transition (Carothers, 1997; Zielonka & Pravda, 2001).
One of the main motivations for measuring the weakness or strength of civil society (CS) is to obtain a reliable basis for understanding the dynamics of its development as well as its social and political potential. The claim that the post-communist civil society is characterized by a certain type of weakness and that this weakness threatens further democratization of the region was made by Ekiert (1991) and Arato (1991) and has been subsequently reiterated in several other analyses (cf. Ekiert & Foa, 2011; Grabowska 1995; Howard, 2003; Szklarski, 1993; Rychard, 1993; Wojtaszczyk, 1991). Bernhard (1996, p. 310) described this apparent weakness of civil society as paradoxical. After all, the fall of communism in CEE has been attributed largely to the strength of civil society in the region (Kopecky & Mudde, 2003, pp. 1–2). How could a civil society capable of successfully opposing the communist regime become so weak almost immediately after its collapse?
In this article we argue that this paradox can be explained by insufficient methodologies involved in what we call the static approach to the strength (or weakness) of civil society, and present a more appropriate alternative, called, for contrast, the dynamic approach. We examine implicit preconceptions present in these early debates, which in our opinion embrace a type of a static approach. It is based on three assumptions: (i) that there exists an inherent property of civil society, either qualitative or quantitative, which translates into its position on the weak–strong spectrum; (ii) that this property can be measured through an analysis of a set of parameters relative to an approximately singular point in time; and finally (iii) that the position on the weak–strong spectrum is fundamental to the dynamics of a civil society and, conversely, the dynamics of civil society understood as its decline or strengthening is a derivative of the property characterized in terms of relative strength or weakness.
Our alternative proposal is a dynamic approach to CS in which (i.*) weakness or strength is not an inherent property of a civil society, and consequently (ii.*) should not be measured by a set of indicators collected for a single point in time. Moreover, in a dynamic approach the opposite to (iii.) is adopted, that is, (iii.*) the weakness or strength of CS is a derivative of the dynamics of its development over time along multiple axes of indicators. In other words, we propose that the weakness or strength of CS ought to be conceived of as the ratio of CS development over time. This implies that we obtain better results when strength or weakness of CS is evaluated within a data-rich environment where comparison over time is possible. However, we do not specify which parameters or indicators are definitive for the weakness or strength of a CS.
We argue that the dynamic approach to CS has several advantages over the static approach. One of its major benefits is that the weakness or strength of CS is expressed in operational terms as opposed to the static approach, which involves postulating a property of CS as a theoretical prerequisite to empirical research, not directly available for empirical testing. While cross-sectional analysis of CS and comparisons across countries at a given point in time present useful data and should not be abandoned, the dynamic approach allows for a plurality of testable, quantifiable yet meaningful understandings of CS as well as an unprecedented method of comparison of the results.
This article is structured as follows. In the first section, we focus on theory and methodology of civil society research, and explore what we see as the static approach to CS weakness that developed in the discussion on post-communist civil society after 1989. We examine multiple problems in methodologies of measuring CS that are amplified by this type of approach. We also discuss the attempts of salvaging the static approach to CS, as well as the latest alternative methods of exploring its dynamics in the post-communist context. In the second section, we show how empirically justified country groupings may be used for augmenting the dynamic comparison between CEE countries.
Theory and Methodology of Civil Society Research
We understand civil society as a sphere of voluntary associations, organizations, and social movements as well as various other types of civic activism that create networks of trust and solidarity and serve numerous communal goals. It is a civic space that provides room for cooperation with others who share the same values, attitudes, norms, or interests and are willing to engage in activities that require responsibility for common or public matters. The major problem in civil society research so far has been the lack of operational definitions that would allow for measuring it. As scholars have already claimed, the static liberal approach to civil society has limited explanatory potential in the CEE context as it is based on highly normative assumption on civil society and its role in democratic consolidation (e.g., Gellner, 1994; Linz & Stepan, 1996). The dynamic approach, which we advocate, is guided less by normative assumptions and more by the actual experience of societies practicing various forms of social self-organization (Pietrzyk-Reeves, 2022). It requires a broad understanding of civic activism as the key dimension of civil society, and a clear focus on its potential that is changing over time.
In the historical perspective, the static approach obfuscated the debate on the weakness of CS in CEE. Bernhard (1996) initially attempted to connect the weakness of civil society to several independent factors, such as the demobilization of insurgent civil societies, loss of leadership in civil society organizations (CSOs) due to migration into mainstream political channels, and post-totalitarian residual effects and social transformations introducing antagonism between various occupational and social groups. Bogdanor (1995) and Pearson (1995) argued that the weakness of civil society in CEE was caused by the propensity to populism and nationalist attitudes. Szelenyi et al. (1996, p. 466) described a “strong social democratic vacuum” in Hungarian politics following the fall of communism. Similar arguments were formulated with respect to other CEE countries (Brzezinski, 1988; Szporluk, 1998). However, Kopecky and Mudde (2003, pp. 1–2) observed that these weakness arguments fell short of explaining the dynamics of civil societies in post-communist countries, pointing to several methodological missteps that, they argue, strongly influenced the overall evaluation of civil society in CEE.
Ultimately, a closer examination of methodologies utilized to justify the weakness of civil society in CEE brought about a suspicion that the initial hypothesis might have been flawed or even entirely unwarranted. Bernhard inspected more closely the relationship between some CEE civil societies and democratization to understand their dynamic and observed that for each of the analyzed countries, there existed a unique relationship between the reconstitution of civil society and democratic breakthrough (Bernhard, 1993, p. 317) and that this unique relationship was likely to extend onto the relationship between the post-communist civil societies and democratic consolidation (p. 325). Coppedge et al. (2011) developed new theoretical guidelines for measuring democracies and democratization, and soon after Bernhard et al. (2015) proposed an implementation of some of these guidelines in the form of the Varieties of Democracy (V-Dem) Project. V-Dem methodology allowed for generating a set of civil society indicators for 173 countries from 1900 to the present by making explicit the expert knowledge from an extensive base of sources. The program introduced a clear and explicit methodology for creating individual indicators and indices pertaining to the condition of the civil society: the Core Civil Society Index (CCSI) and the Civil Society Participation Index (CSPI). Each published dataset is accompanied by a comprehensive codebook, which allows for unparalleled traceability of the obtained results and relatively easy identification of potential errors.
Bernhard et al. (2017) used the V-Dem dataset to reexamine the claim about CS weakness in CEE based on a time-series cross-sectional analysis of 2,999 country-year observations in the years 1989–2012. Their results indicated that there exist no substantial differences between civil society in the post-communist region and other regions. However, Bernhard’s original hypothesis about major intraregional divergence among the post-communist countries was verified. His study found significant differences between the post-Soviet subsample and other post-communist countries in relation to both other regions and each other. Hence, the emergence of V-Dem conceptualization of civil society began a gradual diversion from the claim about weakness of CS in CEE. For example, Foa and Ekiert (2017, p. 419) admitted that CS in CEE is “not as feeble as commonly assumed,” and others (e.g., Jacobsson & Korolczuk, 2017, pp. 1–18) have acknowledged that a better operational conception of CS was required as well as more rigorous research on its condition, actors, and development.
Some authors reject the claim that CS in CEE was particularly weak as based on insufficient methodology and the comparison that conflates civil societies that developed in entirely different contexts (Korkut, 2005). We explore an analogous solution together with a proposal to construct a new, improved understanding of weakness or strength of CS. In proposing this modified approach, we reject the claim that the weakness of CS impacts democratization and treat CS and democratization as separate processes until their relationship can be explored within a data-rich environment.
Our approach aligns with that of Meyer et al. (2020), who attempted to provide a partial explanation of the perceived weakening of CS in post-communist countries by investigating the institutional context of the process of European integration from the perspective of historical institutionalism. More specifically, Meyer et al. (2020) analyze the perceived influence of institutional actors in CEE countries and discuss how these actors relate to the role and activities of CSOs. Their analysis contradicts another prominent assumption concerning CEE countries, namely that the development of their CS proceeds relatively uniformly. The results of Meyer et al.’s study suggest that CEE countries fall into three divergent groups with respect to their CS development incentivized by the integration with the European Union and its institutions. The placement in each group is decided by the stage of progress toward EU accession, but, perhaps relatedly, the countries within each group also display a significant level of economic similarity. The three groups within CEE are characterized by very different processes underlying the potential for CS weakening or strengthening and suggest different causes related to the EU institutional influence.
Problems of the Static Approach
The main problem associated with the static approach concerns insufficient coverage of CS scope and development due to inappropriate or biased definitions of CS. Variants of this problem appear both in studies focusing on CEE and in broader contexts. The various definitions of CS in CEE have a clear liberal underpinning and orientation because activism in the civic sphere is closely linked to the popular mobilizations against communist rule and the democratic transitions in the late 1980s (Foley & Edwards, 1996; Ekiert & Kubik, 2014). In Poland, Solidarity’s decade-long struggle against the communist government contributed to the expectation that post-communist CS would be vibrant and liberal, and that it would make use of specific forms of organization such as nongovernmental organizations (NGOs) or civil society organizations (CSOs). When this was not observed to be the case in the early 1990s (cf. Sokolowski, 2001, pp. 210–218), scholars declared that CS in post-communist countries was in decline, if not dead, and that people were passive and uninterested in civic activism (Kolarska-Bobinska, 1990; Tarkowska & Tarkowski, 1991). Although several studies documented the low levels of trust in political institutions and membership in voluntary associations in post-communist Europe (cf. Rose, 1994; Newton, 2001; Magner, 2005), other observers criticized the “narrow borders of the concept” challenging scholars to rethink and broaden their definition of CS (Kopecky & Mudde, 2003).
Bernhard et al. (2017) pointed out that broader studies likewise pay too much attention to structural factors and not enough to cultural and other context-related factors. In the literature of recent decades, the term civil society remained predominantly reserved for those forces which were perceived as pro-democratic and progressive against those considered national, populist, or backward-looking (cf. Tismaneanu, 1998; Pietrzyk-Reeves & McMahon, 2022). The static approach adds to the problem of undercoverage of CS scope and development by (1) neglecting the informal types of mobilization (often via restrictions on indicators or favoring structural factors) and (2) relying on one-size-fits-all assessment methods.
Kopecky and Mudde (2005) argued that we should “reject the assumption that the vibrancy of associational life can be measured by the numerical strength and/or organizational density of CSOs alone.” Additionally, we suggest that no measure of diversity among CS actors provides the capacity to evaluate weakness or strength in static terms, that is, based on a single round of measurement, although it may serve as a reliable indicator of multiple other properties of CS. Multiple indices contributing to CS (with a notable exception of a discontinued CIVICUS Civil Society Index) focus exclusively on formal organization and legal and regulatory structures that shape the environment for civic self-organization, and thus perform poorly at measuring informal types of civic participation that are growing today, and at evaluating their impact on political processes and citizens’ awareness of public concerns (Magner, 2005; Pietrzyk-Reeves & McMahon, 2022). Moreover, none of the indices satisfactorily measures the diversity of CS actors, the strategies individuals or groups use to achieve their goals, or the content of their mobilization.
Related to this is a more general issue of relying on one-size-fits-all assessment methods, which include civil societies in an unfavorable political, institutional, economic, or geopolitical position. This issue has been acknowledged and avoided in a recent study by Meyer et al. (2020). Earlier, in a series of case studies conducted between 1995 and 1999, Klingemann et al. (2006) combined comparative approaches (using the USA, Norway, and West Germany as reference countries) with a multimethod country-by-country analysis for East Germany, Poland, the Czech Republic, Slovakia, Hungary, Romania, Bulgaria, Estonia, Latvia, Lithuania, Belarus, Ukraine, and Russia. Using such combination of methods was motivated by the fact that none of the available comparative approaches yielded satisfactorily informative results concerning the homogeneity thesis in the context of CEE. More specifically, Klingemann et al. (2006) claim that
While it was easy to identify some general patterns within the region and even beyond it, each country proved to have its own peculiarities. Individual case studies explain the historical and political context of democratic orientations expressed by the respondents. This shows that similar patterns are often generated by different factors, and they help us to explain the many unexpected results emerging from the statistical evidence. (p. 7)
Moreover, the static approach to CS makes it particularly difficult to establish the limits of applicability of global surveys in analyzing and explaining the development of CS, either in CEE or in other regions. As mentioned before, in the studies preceding V-Dem, the question of whether CS in CEE was weak or strong was preliminarily decided before any data were collected. Only subsequently, either data or available expert knowledge was used to justify and motivate the thesis already agreed upon based on the often-implicit theoretical background of the research framework. This testifies to the deep isolation of the past hypotheses concerning the weakness or strength of CS from the relevant phenomena in the context of a specific country or region.
Another problem typical of the static approach is insufficient conceptual background. CS is often treated as a ready-made social good that contributes to democratization and strengthens social capital (Putnam, 1993). This in turn encourages using coarse-grained evaluations, which often limit our understanding of the underlying factors relevant to CS condition in a given case (cf. Civil Society Monitor, Freedom House, etc.). The reduction to radically simplified categories and labels leads to gradual information loss over time and requires that we do not rely on the score yielded by a particular assessment method in making prognoses about CS development. Instead of observing the shifts in relevant parameters on a micro scale, we are forced to withhold the analysis until a country in question is moved from one category to another, while also keeping in mind that countries in the same coarse-grained category do not necessarily share any of the underlying causes for category membership. Such coarse-grained evaluations are used widely even in the context of data-rich environments such as the CIVICUS Civil Society Monitor (Anheier, 2000). Although we do not contest the action-oriented use of such evaluations, the question of how to stipulate the weakness or strength of a CS based on these evaluations remains open, which suggests that coarse-grained evaluations fail at explaining adequately the dynamics of change of CSs.
Lastly, the static approach to CS is highly tolerant to data-poor environments, which corresponds to the heavy theoretical charge of the conjectures formulated in the context of this approach. Kopecky and Mudde (2005) noted that little empirical evidence for expert opinions is given, which allows for arbitrary and noncomparable scores for a group of countries even if the study is aimed at the same type of index. More importantly, these deficits in empirical support enforce arbitrary decisions when comparing inconsistent results of studies produced using different methodologies. Bernhard et al. (2017, p. 334) indicated that data-poor environments typically suffer from far-reaching decidability limitations due to an insufficient amount of data.
Beyond the Static Understanding of CS
A number of recent studies have made progress in producing an improved understanding of CS’s weakness or strength, as well as in making up for the insufficiencies of the static approach to CS. We now briefly introduce these findings that contributed to the development of the dynamic approach to CS. In particular, we acknowledge the need for a multithreaded analysis of CS presented in the recent literature, including a partial analysis by Meyer et al. (2020), and the possibility to create and explore the data-rich environments that have been out of reach for scholars involved in early debates on the weakness of CS in CEE.
Determining the boundaries of CS and its scope and strength is a complex task, but various CS indices provide some indication of CS’s size and impact and the direction of its development in specific cases. However, one must be vigilant about what different indices measure and do not measure. Currently, there are three dominant perspectives on measuring CS. The first and most straightforward is the organizational perspective, which considers voluntary organizations and associations (NGOs and CSOs) and membership of them as the core of engagement in CS. The measurement then focuses on, first, the number of formal registered organizations and, second, declared membership of such organizations by the citizens, also looking at parameters such as the income generated, the financial and human resources employed, and the funding received by civil society organizations. Historically, civic associations of various types have been at the core of social self-organizations. Today, however, this relatively static way of civic participation is no longer the only one (and may not be the dominant one in some contexts), and thus measuring CSOs and their sustainability does not provide an adequate picture of citizens’ public engagement. Although CSOs are still relevant actors of CS and play a vital role as mediators, facilitators, advocates, or service providers, their professionalization often diminishes the need to mobilize wide participation. Furthermore, organizational approaches are predominantly interested in the forms of collective citizen action and pay less attention to the content of their actions.
The second approach is the functional perspective, which measures structural, cultural as well as legal-institutional characteristics of CS. A good example of this approach is the CIVICUS Civil Society Diamond (Anheier, 2000), which distinguishes four major dimensions of CS including (1) the structure or makeup of civil society: its size, actors and their main characteristics, breadth and depth of citizen participation, level of organization, diversity within CS, interrelations, and resources; (2) the environment, which includes as sub-dimensions the political context, basic freedoms and rights, socioeconomic context, cultural context, legal environment as well as state-CS relations and private sector–CS relations; (3) the principles and values promoted and adhered to by a given CS, including transparency, tolerance, nonviolence, gender equality, poverty eradication, and environmental sustainability; and (4) the impact dimension that measures how CS influences individual and communal lives and society as a whole. The CIVICUS Civil Society Index (Anheier, 2013) was a study that used this comprehensive perspective across 72 indicators within the four dimensions. The CSI’s operational definition of CS explicitly described it as involving “individual citizen participation, demonstrations, social movements and other unorganized forms of civic engagement” (Heinrich, 2005, p. 217). The index relied on a significant amount of secondary and primary data, which allowed experts to provide final scores on the basis of multiple data sources. Unfortunately, after two rounds of measurement conducted between 2000 and 2011, with the most recent round that covered 35 countries for the period 2008–11 (Pisarev, 2011), the CSI study was suspended in favor of less resource-costly methods of assessment.
The third perspective is a mixture of the institutional and functional approaches, and was adopted by, for example, the USAID Civil Society Organization Sustainability Index. It takes into consideration the formal conditions for CS development and sustainability, including six dimensions, ranked on a scale from one to five: legal environment, organizational capacity, financial viability, advocacy, service provision, sectorial infrastructure, and public image (USAID, 2020). Like the organizational perspective, this approach also focuses on formally organized types of civic participation. A similar perspective has been adopted by Freedom House Nations in Transit reports, which include CS in their evaluation of the state of democracy and provide a CS score on the basis of expert assessments. The reports evaluate “the organizational capacity and financial sustainability of the civic sector; the legal and political environment in which it operates; the functioning of trade unions; interest group participation in the policy process; and the threat posed by antidemocratic extremist groups” (Freedom House, 2020). These last two approaches, similarly to the CIVICUS Civil Society Monitor, evaluate the conditions for CS with the potential to worsen or improve, and they vary in terms of which conditions they measure.
An entirely new horizon for CS evaluation presents itself in the context of data-rich environments, such as V-Dem, which allows researchers to conduct a time-series cross-sectional analysis, allowing for comparison of development of specific indicators over time, with a set of CS indicators for 173 countries from 1900 to the near present (Bernhard et al., 2015). Two original indices, the Core Civil Society Index and the Civil Society Participation Index offered by V-Dem, gauge CS development and participation, and measure such specific parameters as the degree to which decisionmakers consult with CS actors, the presence and nature of anti-system movements, state harassment and regulation of CSOs, the predominant organizational pattern of CSOs, and the degree to which civil society excludes both female and religious actors (Bernhard et al., 2017, p. 342). The overarching ambition of the V-Dem Project is to use expert knowledge to measure not only the organizational strength of CS in a given country, but also its impact on political life and, in historical perspective that goes back to 1900, its effect on sustainability of democracy. The key to understanding V-Dem’s approach comes from its definition of CS actors (Bernhard et al., 2017): “CS is populated by groups of citizens organized to act in pursuit of their interests, broadly conceived (both material and ideal). We refer to these groups of self-organized interested citizens as CSOs” (p. 346). Nevertheless, V-Dem still focuses on formally organized actors (CSOs, which are clearly defined) and the impact of their actions as well as formal conditions for their functioning.
Augmenting Dynamic Comparisons with Empirically Grounded Categories
The dynamic approach differs from the static approach to the condition of CS in that it focuses not on whether CS in a given country simply is weak or strong at a given point in time, but rather on whether CS is becoming weaker or stronger in the country, based on a series of observations over time. In other words, the dynamic approach requires that the strength or weakness of CS is decided based on the ratio of change in relevant indicators, rather than intuitively, based on untestable assumptions such as described earlier. Instead of asking whether, for instance, Polish CS is stronger than Czech CS, we ask whether we have observed relative strengthening (or weakening) of CS in recent years in both countries or which of these CSs grows weaker or stronger at a greater pace. This means that the ratio of change may be more favorable to CS in a country that is considered intuitively less developed or democratized than another, similarly as the ratio of economic growth may be higher in countries that are less developed economically or have a lower GDP than others. But it also might not be the case, and the same level of democratization might produce different scores for CS strength or weakness. The dynamic approach naturally favors time-series analyses rather than regression models and related techniques. Bernhard et al. (2017) standardized the use of time-series cross-sectional models in representing democratization and CS development but did not offer a way to redefine the measure of CS weakness or strength in terms of this new type of model.
In this context, dependencies in indicators may be used to obtain weights for specific indicators, based on their category within relevant country groupings. Assuming that a correction is needed in how weights are attributed to each indicator’s value, what basis do we have to actually assign a specific number to each case? In the case of V-Dem indices, certain indicators contribute to the final score more than others. However, there is little empirical justification behind the weight assignation. Thanks to the high transparency in the V-Dem codebook and methodology documentation, the weight assignation may be inspected and adjusted in each case. For certain other civil society and democracy indices and scores, such as CIVICUS’s CSI or CS Monitor, the methodology is not as easily tractable, which makes in-depth consideration of the details of the estimations presented there almost impossible. But is there a way to produce the weights assignations in a more justified, fact-based way? We propose one method of weight assignation, based on country groupings.
A country grouping is a classification of countries, either in a specific region or globally, based on measurable and empirically verifiable criteria. In the case of the study by Meyer et al. (2020), a grouping is produced for CEE countries based on a solid institutional approach, which disproves the hypothesis of a uniform development among CEE countries. As a result, three groups of CEE countries emerge, based on their relationship with major institutional factor in local CS development, that is, the EU. Moreover, economic and structural analysis provides further empirical evidence for using historical institutionalism as a “theoretical lens” for thinking about CEE countries. Of course, other types of grouping, equally justified, may be found for the same collection of countries. It is also perfectly possible for a country to be included in a number of groupings, depending on what kind of empirical evidence for various types of “lenses” emerges from data collection.
In the country grouping suggested by Meyer et al. (2020), the Czech Republic, Hungary, Poland, Slovakia, and Slovenia fall into group 1, where after a period of Europeanization and stabilization, we observe attempts by the central governments to regain control over the CSOs and their funding. This is especially evident in Poland and Hungary, where in recent years governments expressed increasingly hostile attitudes toward the CSOs that either are perceived as or self-identify as liberal. Consequently, CSOs and CS activity have been subjected to limited funding, since the EU tended to de-prioritize funding in these countries post accession, and the central governments and local communities failed to step up in providing appropriate funding in the years that followed.
In group 2, including Bulgaria, Croatia, and Romania, both domestic governments and foreign institutions, especially the EU, are perceived as exerting a high degree of influence in the local communities. The EU influences policymaking and CS policies, but the relationship between the state and the CSOs remains rather poor, with a high level of bureaucracy, overregulation, and poor potential for building social capital reported by Bežovan et al. (2017). Differently in group 3, the EU is considered the most influential contributor to policy development in Albania, Bosnia and Herzegovina, Kosovo, North Macedonia, Montenegro, and Serbia, together with the foreign donor organizations and other foreign actors. The attitudes toward liberalization and Europeanization are largely positive, and the policymaking and legal frameworks of group 3 countries are often adjusted to the EU requirements.
This type of grouping suggests that the robustness-oriented indicators may favor countries toward group 3 (meaning group 2 over 1 and group 3 over 2). Conversely, society-oriented indicators may tend to give extra advantage to countries where the development of CS has a longer tradition, providing time for the growing complexity of CSOs and creating favorable conditions for, for example, inclusion of women.
Our second step in this research that needs a separate publication will be a fully developed model to test the proposed method of weight assignment based on empirical examples. The model characterizes the strength of CS as the ratio of change defined over a data-rich environment, using the relevant V-Dem Core Dataset indicators for 1991–2019 (cf. Bernhard et al., 2017) for selected countries in CEE and based on the classification proposed by Meyer et al. (2020) toward diversifying the comparative score algorithms according to the economic, social, and political conditions indicated in the institutional-historical analysis. Applying the same general approach may be conducted for other indicators, measured by other methodologies, but always adjusting to the role of the time variable in evaluating the pace of development of civil societies.
Conclusion
The dynamic approach to civil society allows for obtaining an operational notion of strength and weakness of civil society, both relative to a given indicator within a data-rich environment and in terms of an aggregated (and weighted) civil society score. According to the dynamic view, we say that a civil society has weakened or strengthened compared to a previous year as opposed to the static approach, which requires that we judge the momentary weakness or strength of a civil society based on the value of an indicator and the available scale. Along with the trade-offs stemming from technical difficulties (such as arbitrary judgments in annual score evaluations), the static approaches enforce the reliance on a theoretical prerequisite to empirical research, such as deciding in advance which societies are model civil societies for the sake of comparison. Instead, the dynamic approach offers a measure of CS strength and weakness that is both meaningful and quantifiable, allows for measurement of its own reliability in quantifiable terms based on empirical testing, and allows for unprecedented comparison between various indicators, while leaving a possibility to build and test groupings of case countries based on independent studies, such as that by Meyer et al. (2020). To their advantage, static approaches typically require only a single round of measurement for each of the indicators. However, their value in comparative research on CS development is not to be accepted without scrutiny. Dynamic approaches, whereby each specific approach is relative to a particular data-rich environment, offer a much more reliable and testable way to describe CS.
Perhaps one of the most useful tools available to us in the case of dynamic approaches to CS development is that we have empirical bases for creating distinct geopolitical groupings and evaluation periods for countries that share political, historical, economic, or institutional background. Although it was noted before—for example, by Klingemann et al. (2006)—that a comparison between the democratic development of Russia and the USA is somewhat unfair or imbalanced, no way of equalizing the base level of evaluation was available. In our approach, results by Meyer et al. (2020) offer an empirical basis for creating geopolitical groupings within the countries of CEE. Similar groupings may be created and compared for countries in other regions, based on either historical institutionalism or other methods. Thus, we not only argue that comparing Russia and the USA is different than comparing Russia and Belarus, but also offer a way to state precisely how two such comparisons are different.
Similarly for measurement periods, we propose to base reliability criteria of the average CS evaluation on how well-justified a given measurement period is from the empirical perspective. In the case of Poland, for instance, one could argue for an evaluation in terms of a single score CS between 1980 and 1989 or 1991, or CS between 1991 and 2004 (or perhaps even 1991 and 2020), but a score for CS in Poland for 1980 and 2020 is bound to be unreliable. If this rule of periodization is rejected, the measure of how well CS has developed in a given time period would obviously favor countries that undergo significant institutional improvement or regime change, both of which may have little to do with the strength or weakness of CS.
We introduced and relied on a notion of country grouping, a classification of case countries, either regional or global, which provides a theoretical backing to the discount and markup in weight assignations for each subclass of countries. We assume that an analogous approach may be applied to all data-rich methodologies currently available in literature, using not only country groupings, but potentially also other data-grounded, empirically justified arguments for specific weight assignations. Note also that the country grouping obtained from the study by Meyer et al. (2020) is not considered exclusive, as other types of country grouping could be produced and included in the weight assignment. An interesting open problem emerges in the context of grouping unifications, where potentially conflicting weight assignations will be ordered according to a meta-analysis of dependencies between country groupings themselves.
Financial Support
This research was supported by a research grant of the National Science Center in Poland (project no. 2018/30/M/HS5/00437).
Published online: January 13, 2023