[1] It is necessary to clarify the term of “developing country”. Different countries are labeled as developing. In this paper, we use data from Kazakhstan and discuss intensively civil society in post-socialist countries. Therefore, though some results can be relevant for a broad context, the main part of discussion focuses on the post-socialist case.
[2] To some extent an exception is the article written by Nezhina and Ibrayeva (2013). However, despite the use of mass survey, qualitative methods (mixed methods research) and a scope allows us to categorize this research project as close to a classical case study.
[3] It is worth to note that, although social network analysis is very popular approach for civil society studies (Diani, 2015; Baldassarri & Diani, 2007; Diani & McAdam, 2003), with some notable exceptions (Taylor & Doerfel, 2011; von Bülow, 2010) few scholars have paid close attention to applying this approach to studies of civil society studies in post-socialist and/or developing countries.
[4] It is worth to note that in the case of domestic integration, Stark and colleagues offer three types of domestic integration: participation, embeddedness and associativeness (Stark, Vedres, & Bruszt, 2006: 328-329). The first type is participation or involvement of different social groups and/or local communities in the activity of NGOs. The second type is called embeddedness, a degree of interrelations between NGOs. The third type is associativeness, a degree of interrelations with actors from diverse institutional fields (government, business, mass media etc.).
[5] First of all, high level of sensitiveness to details.
[6] Critical remarks by Nick Crossley (2011) that demonstrate visibly limitations of individualism and holism seem very useful for our critique of previous qualitative and quantitative research of non-governmental organizations in the developing countries. Holism that overwhelmingly exaggerates the role of “whole” can be problematic because social actors – for example, NGOs – are presented as ‘cultural dopes’, who internalize external constraints, such as authoritarian rule, and/or donors’ power (Luong & Weinthal, 1999). Individualism, conversely, states society is no more than the sum of the individual actors, and, as consequence, neglects the structure of relations.
[7] Social bonds are “strong identity ties [that] embed associations into dense clusters of interaction” (Baldassarri and Diani, 2007: 771).
[8] It is worth to note that current research of homophily pay little attention to the investigation how homophilous processes flow in different network topologies. In general, homophily is associated with polycentric network structure, in which homophilious groups form densely connected clusters.
[9] Starting from this point, we observe the symbiosis between the regime, which can be defined as semi-democratic or soft authoritarian, and civil society. In our opinion, Anthony Spires (2011a) described in detail a very similar situation in the case of China.
[10] According to experts of CIVICUS, public associations and foundations are the most popular form of NGOs in Kazakhstan (Makhmutova & Akhmetova, 2011: 21).
[11] Questionnaire, dataset and supplementary materials that were used in the research project can be found here: http://e-valuation.kz/social_capital_en.html.
[12] Trying to cope with the problem of non-respondents, we were moving forward based on ideas of Harry Robins and his colleagues (see Robins, Pattison & Woolcock, 2004), who have intensively studied the question of the missing data in networks.
[13] It is worth to note that Stark, Vedres and Bruszt (2006) used this as a control variable.
[14] Of course, we were able to transform this an attributive variable – spheres of activity of NGO – in relational one and to measure the number of joint spheres of activity (denoted by Sij) by the following expression: where xij is the number of spheres of activity by NGOs i and j, xi is the total number of spheres of activity of NGO i, and xj is the total number of spheres of activity of NGO j. This measure ranges from zero (no joint spheres) to one (only joint spheres). The same technique of measure was used by Val Burris (2005: 257) in his study of interlocking directorates. However, we have decided not to do this computation because our observations show that Kazakhstani NGOs are easily to change the sphere of activity (in other words, this category is very changeable).where xij is the number of spheres of activity by NGOs i and j, xi is the total number of spheres of activity of NGO i, and xj is the total number of spheres of activity of NGO j. This measure ranges from zero (no joint spheres) to one (only joint spheres). The same technique of measure was used by Val Burris (2005: 257) in his study of interlocking directorates. However, we have decided not to do this computation because our observations show that Kazakhstani NGOs are easily to change the sphere of activity (in other words, this category is very changeable).