Work systems as frames of reference for HR disclosure

Размещено на http://www.allbest.ru/

Graduate School of Management, St. Petersburg State University

Work systems as frames of reference for hr disclosure

A.K. Bordunos, S.V. Kosheleva

Russia

Abstract

The paper focuses on the challenge of coding and interpreting the human resource disclosures. Treating disclosed signals as positive or negative depends on the implicit contextual frames of reference. Such frames represent a combination of legitimate human resource practices, processes, initiatives, values and beliefs. The research explores applicability of human resource management systems, also known as work systems, within the confines of such frames. Additional attention is paid to the reasons of variety in the human resource disclosures and the rationale of preferring one work system over the other despite limited freedom of choice among available standards. Through analysis of the content, extent and quality of the employees-related disclosure by 18 top banks in Russia for the 2016 period, the research discusses opportunities and barriers of work systems as the frames of reference for HR disclosure. Research design relies on thematic and content analysis to compare existing standards with the ones proposed by the strategic human resource management literature.

Keywords: annual report, disclosure, discourse analysis, high-commitment work system, high-involvement work system, high-performance work system, human resource management.

Аннотация

Системы организации труда как референтные фреймы для раскрытия информации о человеческих ресурсах

А. К. Бордунос, С. В. Кошелева

Институт «Высшая школа менеджмента» Санкт-Петербургского государственного университета, Россия

Исследование посвящено проблеме кодирования и интерпретации раскрываемой информации о человеческих ресурсах в годовых отчетах компаний на примере российских банков. Смысловая интерпретация предоставляемых данных зависит от референтных (эталонных) стандартов, которые выступают в качестве идеализированной системы легитимных практик, процессов, инициатив, ценностей и убеждений в отношении персонала. Системы управления человеческими ресурсами, также известные как системы организации труда, рассматриваются в данной работе в роли таких референтных стандартов. Анализируется связь между типами систем организации труда и предпочтениями в выборе и реализации практик управления человеческими ресурсами на основе анализа содержания, количества и качества раскрываемых фактов о персонале в годовых отчетах 18 ведущих банков России в 2016 г. Кроме того, обосновываются причины, ограничения и возможности выбора конкретной системы организации труда и разнообразия в предоставлении информации о человеческих ресурсах. Дизайн исследования опирается на тематический и содержательный анализ при сравнении сложившихся систем организации труда с эталонными стандартами, предложенными в работах по стратегическому управлению человеческими ресурсами.

Ключевые слова: годовой отчет, дискурс-анализ, система организации труда, производительность труда, приверженность персонала, вовлеченность персонала, управление человеческими ресурсами.

Introduction

HR disclosures cover all revealed information about employees, HR and HR management. Interest in HR disclosures emerged in 1960s within academic debates regarding HR accounting. This research stream attempts to show value of human capital in the balance sheet [Brummet, Flamholtz, Pyle, 1968]. For instance, it raises the problem of how to treat human capital: as a source of costs, as assets or as liabilities. In the 1980s, HR disclosures gained the attention of researchers with interest in intellectual capital reporting. This research direction also attempted to quantify value of employees' education, qualification, relevant competencies, knowledge and entrepreneurial spirit [Guthrie, Petty, 2000].

In 1990s, such attempts were criticized, considering that not only human capital brings value, but certain actions that align employees' behavior with the corporate strategy [Coff, 1997]. This stream highlighted the problem of interpretation of disclosed facts. For instance, the idea to use shareholders' funds for social projects could be highly criticized [Gelb, Stawser, 2001] or highly welcomed as it leads to better financial performance [Pava, Krausz, 1996].

The novel stream focused on disclosures related purely to HR and HR management, due to acknowledgment of its high strategic role, especially for knowledge based industries. “HR disclosure plays a significant role in the determination of productivity, profitability and sustainability, while also forming a basis for decision making in the management of a company” [Mishra, Mishra, 2017, p. 9]. To date, the most widely applied method for data analysis in HR disclosure studies is regression modelling [Kumar, 2012; Mishra, Mishra, 2017]. However, it narrows the focus to the disclosure extent through construction of the disclosure index. Such tendency raises two general concerns. Firstly, similar to analysis of sustainability reports, it provokes exploration of proper frames of reference for HR disclosure. For example, such research direction compare mandatory and voluntary disclosure, effect of disclosure regulation on the credibility of the disclosed data and consequent limitation of the regulated disclosure to firms [Healy, Palepu, 2001]. Secondly, it provokes exploration of alternative methods of HR disclosure analysis, aiming to improve the interpretation of existing data.

Traditionally, HR disclosures are perceived as part of the corporate disclosures, and analyzed from the corporate governance perspective [Petera, Wagner, 2017]. However, current research selects an alternative view -- from the world disclosures perspective [Kompridis, 1994]. Nevertheless, the research accounts for a specificity of corporate disclosures for minimizing bias in interpreting the variance in HR disclosure.

Demand for corporate disclosures arises from an information asymmetry [Akerlof, 1978]. Corporate disclosures influence external perceptions about firm's activities and success, increasing its value, improving reputation and helping to achieve competitive advantages [Kent, Zunker, 2013; Spence, 2002]. Credibility of the disclosed information is enhanced by regulators, auditors, and existing standards [Healy, Palepu, 2001]. Significant regulations governing corporate reporting predetermine the firm's disclosure strategies, thereby, limiting variety in HR disclosures too. As corporate disclosures are especially critical for the capital market, public joint-stock companies usually face the most severe regulation. This effect might diminish value of the information asymmetry, despite higher credibility of the disclosed facts. At the same time, it motivates exploration of solutions on how to escape from such predetermination.

World disclosures focus on how actors (managers) recognize, interpret and structure relevant facts [Kompridis, 1994], framing them into a meaningful output and relatively stable meaning systems [Cornel is- sen, Werner, 2014]. Despite the seeming freedom, world disclosures also face significant latent regulations. They are predetermined by a legitimacy of the revealed facts: how close they are to the audience' self-interests, how well they pass their normative approval and how easily they are comprehensible [Suchman, 1995]. From this perspective, HR disclosures are challenging due to the discursiveness of the related concepts that often creates a barrier to interpretation of the reported signals as positive or negative. Discursiveness relates to discourse. Discourse is the interactive process of conveying ideas, while ideas -- are switches for interests, road maps, focal points; narratives that shape understanding of facts, intentions; frames of reference [Schmidt, 2008]. It means that HR disclosures require implicit contextual agreement on specific frames of reference.

The goal of the current research is to explore the work (HR management) systems as aspirational frames of reference for HR disclosure, which influence disclosure content, extent and quality. Work system is a systematic and integrating approach of HR management toward the alignment of the HR functions with the firm strategy [Wei, Lau, 2010], aiming to facilitate corporate competitive advantage [Becker, Hu- selid, 2006]. The current research relies on a widely accepted framework “HR Architecture” by [Lepak, Snell, 1999] as a starting point of analysis. Mixed-method design allows testing applicability of the rich quantitative-based heritage of the work systems- related research, accounting for contextual nuances.

The novelty of the current paper lies in a perception of the HR disclosures and the work systems from the discourse-analytical perspective, following the call to shift from the widely used positivist [Harley, 2015] or normative [Bratton, Gold, 2015] perspectives. Despite positivist trend to decontex- tualize HR management research, many scholars acknowledged ethnocentricity of the work systems due to the unique role of the context in understanding the meaning and intentions behind the revealed facts [Boxall, Macky, 2009; Child, Marinova, 2014]. The current research analyzes the rationale behind preferring one frame of reference over the other under conditions of high predetermination of HR disclosures.

The current paper analyzes annual reports of large commercial banks in Russia for the period 2016, shortly after a crisis, when banks had to search for univocal signals of success, differentiating them from competitors, at the same time, resisting volatility, uncertainty, complexity, ambiguity and strict rules for HR disclosures and for HR management in general. The context plays an illustrative role, showcasing the method of differentiation from competitors, despite limiting firms' freedom of choice among aspirational HR management standards.

The first section of the paper reviews discourse-analytical perspective on HR management. The second section describes the methodology of data collection and analysis. The third section provides key results of the analysis. The fourth section contains conclusions and discussion, including implications and future research directions.

Making Sense of Human Resource Disclosures

Voluntary disclosure in the discourse-analytical research The current paper reports findings of the discourse-analytical research, focused on frames of reference for HR disclosure. The goal of the research is to test applicability of the work systems in a role of such aspi- rational referent standards. The research pays attention to the opportunities and barriers from their adoption and to reason behind preference of one system over another, accounting for a limiting role of the context.

We follow epistemology of social constructionism. It assumes that any concrete text production and interpretation is based on the specific discourse structures -- frames for mental representation of realities and opinions towards them [Van Dijk, 1993]. Voluntarily disclosed facts represent a hyper-reality, which is a mixture of secret routines and an aspirational ideology [Al- vesson, 2013]. Aspirational ideology is a frame shaped by the dominant discourse that latently influences socially shared knowledge and opinions [Van Dijk, 1993] and thus, the dominant discourse defines legitimacy of the disclosed intentions. Following a discourse-analytical perspective, the current research explores variance in such aspirational ideologies on three levels: discursive -- language used; contextual -- level of social conditions; and ideational -- shared values, beliefs and meanings [Al- vesson, 2013].

Discursive level. Different aspirational HR referent standards represent specific source of the linguistic patterns both helping and limiting firms in sharing their success [Alvesson, Skцldberg, 2017]. Thus, analysis of the discursive level is a tool for recognizing preference of the particular HR referent standard.

Contextual level. The HR architecture framework [Lepak, Snell, 1999] grounds on the assumption that companies are self- determined in their choice. Conditions of emerging markets could greatly undermine this expectation due to idiosyncrasy of institutions for developing markets [Rottig,

. Thus, additional essential interest of the research analyses how context causes variability in HR disclosure, limiting firms' self-determination in selecting suitable referent aspirational HR standard.

Ideational level. The ideational level of a discourse analysis assumes that companies cannot manage employees directly, they rather manage “the insides”: workers' hopes, fears, aspirations [Deetz, 1995]. It affects HR management philosophy that shapes each work system and inspires specific individual and organizational identity, thus, reducing the range of decisions to the choices compatible with it [Alvesson, Willmont, 2002]. Referring to ideational level is more beneficial due to its higher stability over time and consentience across different stakeholders [Inayatullah, 1998]. For a researcher, it means an interest to exploring nuances, which is possible with a small number of accounts [Alvesson, Skцld- berg, 2017]. Most studies on HR disclosure strive to high sample, usually integrating HR facts in the intellectual capital or corporate social reporting [Petera, Wagner,while neglecting the variance in the content.

Methods for analysis of voluntary disclosure. The discourse-analytical research of HR management has received increasing attention in the literature [Heizmann, Fox, 2017; Keegan, Francis, 2010; Zanoni, Janssens, 2004]. These papers provide positive illustration in combining the strategic HR management and the institutional theory. However, these researchers mostly focused on oral discourse relying on the semi-structured interviews, and foregrounding the language of HR practitioners as a force for change. To avoid biases associated with the contextual level of disclosure, discourse- analytical researchers generally prefer rou- tinized sources of information rather than those formed as a result of interaction between the researchers and the participants [Alvesson, Skцldberg, 2017]. Annual report is an example of such routinized source.

Analysis of the discursive, contextual and ideation levels are suitable for the analysis of the content of HR disclosure. There are two additional, more widely spread points of interest related to voluntary HR disclosure -- its extent and quality.

The disclosure extent is the quantity of the disclosed facts. The most widely used method for it is a content analysis [Hackston, Milne, 1996; Guthrie et al., 2004]. This method analyses disclosed facts systematically, objectively and reliably by codifying revealed HR information into pre-defined categories to derive patterns [Guthrie et al., 2004]. However, studies vary in coding unit of analysis, selecting sentences, words, paragraphs or portion of pages. Words have little meaning without a context, while sentences, paragraphs or portions of pages might consist of several distinct facts [Milne, Adler, 1999].

Quality of disclosure is achieved by supporting facts with narration and monetary or actual physical quantities. It increases transparency and credibility of the disclosed facts. The most widely used method for analysis of the HR disclosure quality is disclosure index [Beattie, McInnes, Fearnley, 2004; Davey, Schneider, Davey, 2009]. It was designed to measure series of items (themes), giving a contextual surrogate score. Disclosure index usually refers to a six-point scale (from 0 to 5). Five scores are assigned if at least one disclosed fact in the predefined sub-category is supported by the monetary or actual physical quantities mentioned along with the narrative statements. The lowest, zero score, is used for non-disclosure. When banks mention only quantitative information, the score is four; when banks illustrate facts only narratively, the score is three; when the disclosure is obscure, the score is two; when the bank asserts that a disclosed fact is immaterial to organization, the score is one. Review of the previous research reveals several studies, which applied similar method [Schneider, Samkin, 2008; Yi, Davey, 2010]. However, most of them focused on the intellectual capital, including general HR-related facts. Researchers did not distinguish between various work systems, measuring only the willingness to disclose information voluntarily. Another perspective aimed at analyzing effects caused by the disclosure of different facts, including HR disclosure [Ma- riappanadar, Kairouz, 2017]. An alternative stream of research followed deductive logic, analyzing HR facts as part of the firms' sustainability programs, meaning having particular aspirational framework as a single referent standard for the whole sample [Ehnert et al., 2016]. This framework is known as “sustainable HR management”, guided by the Global Reporting Initiative (GRI) reporting guidelines. It means narrowing the focus mostly to labor practices and human rights, neglecting other HR- related elements and possible alternative aspirational work systems.

In summary, the current research refers to discursive, contextual and ideational levels of the written discourse in order to recognize latent frames of reference, based on the content of HR disclosure. Analysis of the quality and extent of HR disclosure aim to diminish bias of attribution. Reference to existing HR management systems, also known as work systems aim to ease the task of attribution of actual latent frames of reference.

Work systems as aspirational referent standards for HR disclosure Work system is a systematic and integrated approach of HR management toward the alignment of the related functions with the firm strategy [Wei, Lau, 2010] aiming to facilitate corporate competitive advantage [Becker, Huselid, 2006]. Each work system contains such elements as HR management philosophies, policies, practices and processes [Beer, Boselie, Brewster, 2015]. The interrelationship of these elements are usually analysed with other elements: organizational system, the multiple players who enact a work system, and the multiple stakeholders who evaluate the organization's effectiveness and determine its resilience [Jackson, Schuler, Jiang, 2014].

Most of the current knowledge regarding work systems was gained as a result of positivist quantitative approach. While analyzing work systems, researchers usually ask a manager to mark the presence of certain HR management practices in their organizations. They assess the extent to which these elements emphasize high productivity, high commitment, or high involvement, and thus define presence of a corresponding work system [Jackson, Schuler, Jiang, 2014]. Such attitude assumes that the more specific practices are involved, the better chance they have at success. Such universalistic reference to HR practices was criticized for ignoring interdependence of different elements in HR management [Demortier, Delobbe, El Akremi, 2014]. Shift of attention from the list of practices to the discourse structures suggests solutions for this critique. Thus, the current research applies more holistic approach by considering additional implicit components of the work systems. It helps to decrease bias of attribution.

Discursive level of analysis is the most diverse. For example, there is no consensus regarding exact amount and content of work systems, and their titles are often used interchangeably [Chang et al., 2014]. The current research relies on the differentiation suggested by [Lepak, Snell, 1999] as a starting point of analysis. These authors grouped various standards according to the strategic value and uniqueness of human resources possessed by employees. Dimensions “value” and “uniqueness” formed a matrix of four groups of employees associated with a particular HR standard, three of which are of higher interest for the disclosure. High level of strategic value and uniqueness suits high-commitment work system (HCWS). It relies on development of employees' competencies and employees' empowerment [Lepak, Snell, 2002]. High level of uniqueness, but low level of the strategic value suggest high-involvement work system (HIWS). It focuses on group incentives, cross-functional teams and retention of relationship rather than investment in human resources [Lepak, Snell, 2002]. High level of strategic value and low uniqueness imply high-performance work system (HPWS). It focuses on immediate performance and arranging standardized jobs [Tsui et al., 1995].

The current paper reports findings of the second step of the bigger research aiming to explore key aspirational referent standards and reasons for favoring one standard over the another. Prior theoretical review of the work systems in AngloSaxon countries [Bordunos, Kosheleva, allowed reconstructing patterns for differentiating known referent standards, based on the ideational level of analysis (Table 1).

Table 1 Review of the ideational level of the work systems

Source: adapted from [Bordunos, Kosheleva, 2016; 2018; Kosheleva, Bordunos, 2018]; structured according to [Excellence through people, 2017].

Cooke also warns that “existing research of HR management has been mainly conducted in developed countries or applying a western research topic to a nonwestern country with preconceived research question and theoretical framework that are insensitive to local context and history” [Cooke, 2018, p. 9]. The proposed study attempts to contribute to literature on the strategic HR management by clarifying contextual specificity of key aspirational referent standards.

Reasons for variety of the HR disclosure

Variety in HR disclosure means difference in the content, extent and quality of the disclosure. [Petera, Wagner, 2017] suggested three contextual reasons for variety: 1) company size, due to higher pressure of stakeholders and regulating authorities on large companies, and because large companies face lower costs of production for such report; 2) governmental ownership, due to reputational concerns related to higher agency costs and weaker governance; or servi ng a benchmark; 3) placing securities on the stock exchange, due to additional regulation, higher interest of investors and media.

According to strategic HR management perspective, the content of HR disclosure depends on the firm's strategy. Institutional theory, particularly signaling theory [Spence, 2002], added that managers have superior information than the outside stakeholders on companies' future performance, so managers can improve the quality of their reporting and voluntarily providing additional information [Connelly et al., 2011; Healy, Palepu, 2001]. This view suggests that the dominant reason for HR disclosure is to diminish information asymmetry influencing external perceptions about the firm's activities and success, increasing firms' value, as well as improving its reputation [Kent, Zunker, 2013; Spence, 2002].

However, voluntary disclosure is always a potential threat to the company's competitive advantage, as competitors might utilize gained information for their own perfection [Williams, 2001]. This threat is not valid for firms that follow a sustainability-driven approach [Prahalad, Hamel, 2000] suitable for high-commitment work system. Firms which adopted this work system as frame of reference usually increase density and quality of the disclosure, without being afraid that competitors might imitate the proposed actions. Their human resources are not only valuable, but also rare, imperfectly imitable and non-substitutable [Barney, 1991]. Thus, companies adopting high- commitment work system might show high-er quality and extent of the HR disclosure compared to companies favoring the other work systems.

An additional reason for a better level of disclosure is referring to less legitimate referent frames. Disclosed facts should have certain meaning to stakeholders, being an unequivocal signal of success or failure in a particular context [Suchman, 1995]. In case firms follow less legitimate options, they have to explain additionally the nature and the attractiveness of the disclosed initiatives, in order to provide information about the intentions behind them [Stiglitz, 2000]. In Russian context, the dominant discourse suggests legitimacy of high-performance work system, restricting adoption of the high-commitment work system and high-involvement work system [Bordunos, Kosheleva, 2018]. Thus, companies that adopt high-involvement work system might show higher quality and extent of the HR disclosure than companies favoring high-performance work system.

Legitimacy theory [Suchman, 1995] assumes that companies disclose information as a reaction to expectations of the institutional environment: certain political, economic, social, and environmental factors. If companies disclose facts to legitimize their actions [Bitektine, Haack, 2015], they mainly exhibit implementation of certain institutional rules and recommendations [Kotonen, 2009]. Such facts usually do not differ from the competitors within the industry. In this way legitimacy decreases the value of HR disclosure due to the similarity in the quality, extent and content of the disclosure.

Asymmetry of information, on the contrary, increases value of HR disclosure, due to the difference in the quality, extent and content of the disclosure. Managers voluntarily can share certain facts about companies' current and future performance [Hea- ly, Palepu, 2001] to improve the perception of their reputation, prestige and ability to earn positive cash flows in the future [Connelly et al., 2011]. In this case, their HR- disclosure should significantly differ from their competitors' [Deegan, 2002; Elitzur, Gavious, 2003; Stiglots, 2000].

In summary, not only content, but also quality and extent of HR disclosure help differentiating among frames of reference. Analysis of these apsects also reveals barriers and opportunities for preference one work system over another and explain the contextual reasons for variety in HR disclosure. The research intentionally refers to several categories and sub-categories of HR patterns in order to explore which of them are more helpful in a differentiation from competitors.

Preference of work systems by banks in Russia

In 2016, banks in Russia had potential interest in all three work systems.

Most existing requirements and recommendations of the Central Bank of Russia, the Financial Stability Board and the Federal Law motivated the disclosure of applying high-performance work system patterns. Here are some examples of such requirements: interest in short-term outcomes, the productivity-based assessment and the remuneration system, limited autonomy and control of information distribution. In 2016, authorized representatives of the Central Bank of Russia carried out the total of 585 inspections of commercial banks and their branches, using risk-based approaches to organize inspections. Requirements of the Central Bank of Russia were affected by the Regulatory Consistency Assessment Program maintaining conformity of banking regulation in the Russian Federation to the Basel II, Basel 2.5, and Basel III standards “The Basel Committee on Banking Supervision (BCBS) is the primary global standard setter for the prudential regulation of banks and provides a forum for regular cooperation on banking supervisory matters”; as well as the Financial Sector Assessment Program of the International Monetary Fund and the World Bank. These regulations, along with the Federal Law, affected both actual and disclosed preferences of work systems by banks and their performance, legitimizing high-performance work system.

On the other hand, 1999 banks worldwide were inspired by the Global Reporting Initiative (GRI) -- the most known framework for the voluntary disclosure of social firms' performance. [Brown, De Jong and Lessi- drenska, 2009] noticed that firms seriously relying on this framework, had to invest considerable resources in report production and external verification, motivated by its transforming experience. Thus, they willingly routinized it as a standardized internal practice. Institualization of the GRI reporting inspired shift to the sustainable development, associated with the high-commitment work system [Bordunos, Kosheleva, 2016].

However, emerging markets usually face weak institutions of corporate governance that create barriers to proper monitoring [Okhmatovskiy, 2005]. In such situations, the role of banks is reflected in their position in inter-organizational networks, intensified by affiliation with particular business groups, invitation to boards of directors or on management positions representatives of important business-partners, governments, and shareholders [Okhmatovskiy, 2005]. These facts highlight the importance of social capital, suitable for high-involvement work system [Bordunos, Kosheleva, 2016]. Furthermore, the growing role of technology, adjustments in government regulations, industry consolidation, financial innovation and securitization [Wilson et al., 2010] intensify favoring dynamic capabilities concept that requires high-involvement work system. However, such standard requires specific management style: high level of autonomy, coaching, and certain team climate [Xanthopoulou et al., 2009]. These practices might be criticized in Russia, requiring additional disclosure of intentions behind them.

Research Methodology

In this research we analyzed HR disclosure using three methods. We started with thematic analysis on a latent level for a qualitative interpretation of the content. Thereafter, the content analysis for defining quantitative difference in extent of the HR disclosure was conducted. Finally, we assessed the quality of HR disclosure and compared different work systems based on this indicator. The aim of this first step was to reveal actual latent referent frames for HR disclosure. Thereafter, all observations were distributed into several groups according to the revealed preferences. The second step included visual analysis of the difference in patterns along with the role of the context. Both steps included comparing of the revealed patterns with the existing theoretical expectations, associated with the widely spread work systems (Table 1). They helped to explore the HR management (work) systems as aspirational frames of reference for HR disclosure, which influence content, extent and quality of HR disclosure.

Data source

The main source of data in this research are annual voluntary management reports for the 2016 year, published by the selected 18 large commercial banks operating in Russia. Banks tend to accelerate in voluntary management reports key information from other sources for disclosure, including financial reports and reports of the sustainable development, considering it as an important medium, through which managers commonly signal what is important [Guthrie, Petty, 2000]. Thus, the research omitted the observation financial reports and reports of the sustainable development. Another reason for their exclusion is higher specificity and predetermination of their content. The research refers to the reports published by the Interfax Center of Disclosure and banks' corporate web sites. Findings supplement observations from annual reports of the Central Bank of Russia on banking development and supervision, and from rankings published by banki.ru. The research focuses on large commercial banks, because of their better transparency, higher variety within the industry and diversity of stakeholders dealing with them. Crucial role of banks in the economic success of a country, highly turbulent environment that intensifies a need to enhance competitive advantage and demonstrate organizational health raised an additional interest to this sector.

Role of the context

An essential motivation of the current research is to observe the role of the context in the work system preferences influencing extent, quality and content of the disclosure. To explain the variance, the research focuses on size (position in a rank by assets gross), government ownership, placing securities on the stock exchange (public joing stock company, PJSC) and legitimacy of referent frame. Additionally, the research relies on such performance measurements as ROA and ROE (by 01.01.2017), as the most widely disclosed metric in voluntary management reports. We also considered other institutional factors associated with banks' head office location. Table 2 summarizes information related to the sample.

Table 2 Characteristics of the sample

Company sample

The initial research sample comprised of 100 largest banks, according to assets-based rank: the lowest level of assets was above P150 billion Here and later the currency exchange rate is accurate for 01.01.2017, according to the Central Bank of Russia. (above €780 thousand). There are several reasons for this selection. Firstly, more stakeholders are interested in their success. Thus, these banks are more likely to be proactive in disclosing their HR facts voluntarily, owing to their visibility [Guthrie, Petty, Ricceri, 2006]. Additional reason is an assumption that the bigger the bank, the higher is reliability of its annual report due to more attention from the regulating authorities and due to availability of resources required to collect corre- spondig data [Brown, De Jong, Lessidren- ska, 2009].

According to the Central Bank of Russia, by the end of 2016, there were 975 registered banks, 623 of which were in operation.

Despite registration of new 110 banks, there was a shrink of banks amount by 15% due to ongoing process of “purging the system of unscrupulous and unstable banks” for an increase of banks' transparency; reducing the cost; ensuring overall stability of the banking system [Banking Supervision Report..., 2016, p. 13]. In 2016, 14 banks terminated their activity as a result of reorganization through merger or transformation, 351 banks' licenses were cancelled, and banks started process of liquidation, mostly due to insolvency (288 banks), and also due to forced liquidation (37 banks). In understanding banking share concentration, the Central Bank of Russia refers to the share of the largest 200 commercial banks by assets -- 98.0%; and to the relative share of the top five banks -- 55.3%. 13 banks had equity capital of over --100 billion (€1560 millions), 96 banks -- over P10 billion (above €155 millions), 312 banks -- over P1 billion (€16 millions), accounting for 98.7% of Russian banks' capital.

Thus, initial sample covers big share of the market. However, in the year 2016, only 18 banks out of 100 added to annual reports voluntarily disclosed information. This differs significantly from the previous years, when almost all banks published the annual management reports. The first reason of such change is the processes of mergers and acquisitions, when a group of banks submits one annual report for the whole group, while assets-based ranking accounts for each bank separately. The second reason is that some banks use title “annual report” for a joint presentation of financial and accounting results, including only those required for disclosure facts. Such reports were not taken into consideration. The third reason is a necessity to invest considerable resources in a report production and an external verification, so only the most motivated by its transforming experience banks usually provided voluntary reports in 2016 contrary to more stable 2012, when each bank was interested in such disclosure [Bordunos, Kosheleva, 2015]. The final sample size is more beneficial for a discourse-analytical research, as it allows better exploration of variance, instead of commonalities, on three levels -- language used, social conditions and shared beliefs [Alvesson, Skцldberg, 2017]. However, findings could not be applicable for the whole industry -- they serve illustrative purposes, providing best cases to follow in the particular context. For some ethical reasons, we do not disclose banks' titles, but refer to them by the number (Table 2).

Data analysis

Thematic analysis on a latent level was a key method for the qualitative analysis of the content in the current research [Braun, Clarke, 2006]. The research referred to a predefined list of themes, grouped into five categories (Table 1). The categories were too broad, thus, each category was additionally divided into two sub-categories (Table 3).

Table 3 Framework for content analysis of HR disclosure

Source: based on [Excellence through people, 2017].

The current study referred to content analysis for exploring extent of HR disclosure, adverting to the groups of patterns presented in the Table 2. However, the coding unit of analysis was each single fact, rather than sentence, words, paragraphs or portion of pages. The main priority was coding for meaning, rather than looking for exact phrases [Milne, Adler, 1999].

Disclosure index was adopted to analyze the quality of HR disclosure. Similarly to the content analysis, the current research instead of scoring each fact or sentence, scored the whole sub-category, due to different length of the sentences and dispersing similar facts across the whole report. It helped in avoiding mentioned bias common for the content analysis, while counting similar facts as different units.

Here is an example of statements' scores, using annual report for 2016 of bank 1: human resource labor

“(Bank 1) actively encourages employees' innovativeness (fact 1): in 2016 more than 100 thousand employees (fact 2) provided above 30 thousand ideas (fact 3), 13 thousand of which were implemented (fact 4). The economic effect was above 4 billion rubles (fact 5)”.

From the perspective of HR disclosure extent, the statement contained five facts for the category “2. Effective Communication & People Engagement”. In terms of HR disclosure quality, one statement was narrative, three statements were physical quantities, one statement was monetary. Thus, five scores were assigned to the sub-category “2.2. Support for employee engagement”, as monetary, actual physical quantities and narrative statements were made. For the thematic analysis, this statement was not informative.

Results

Grouping banks based on disclosed work systems

In the first step of analysis we revealed four groups of banks distinct from each other in terms of HR disclosure. All banks showed evidence of high-performance work system. Thus, the research focused only on the difference between the groups, separating signals of one work system from the others.

The first group consisted of eight observations (banks 11-18), for which this work system was the only revealed aspirational standard. The second group formed five observations (banks 2-6), which additionally to high-performance work system favored high-commitment work system. The third group consisted of four observations (banks 7-10), which in addition to highperformance work system showed evidence of the high-involvement work system. Only one bank (bank 1) showed evidence of all three work systems.

Figure 1. Distribution of banks according to the extent and quality of HR disclosure

Figure 1 represents distribution of banks according to the content of HR disclosure. Additional analyses of the disclosure extent and quality revealed additional patterns that help in differentiating work systems. All banks favoring high-performance work system showed lower level of HR disclosure quality and extent. Bank 11 was the only exception. However, it is national bank, similarly to Bank 1, it shows much higher level of quality and quantity of the disclosure than other banks of the same group. Banks favoring high-involvement work system showed mixed patterns.

Content analysis also helped to explore contextual predetermination of the choice of work systems. For example, it revealed that group of banks that favor high-commitment work system was formed only by Russian banks.

Variance in content of HR disclosure General overview. The discourse analysis showed full match of the disclosed facts with the expected patterns illustrated in Table 1. Thematic analysis showed that “2. Effective Communication & People Engagement” was the least sophisticated by the institutional environment category, providing freedom for differentiation. However, this category lacked legitimacy [Suchman, 1995], leading to the difference in understanding and measuring key aspects of the related HR management initiatives and a need to explain corresponding intentions. Facts in the categories “3. Leadership & People Management” and “4. Learning & Development” were the most similar, due to strict regulations of these areas. Figure 1 summarized distribution of observations into three groups after this step.

Discursive level. Gained observations enabled better insight into commonly shared meaning of HR-related concepts within each group. For example, high-performance work system-oriented companies referred to employees as “staff”, considering that employees perform individually in accordance with existing standards, bearing different kinds of threats. high-commitment work system- oriented companies referred to employees as “personnel” calling them one of the main resources and bearers of the competitive advantages, differentiating the most valuable of them. High-involvement work system-oriented companies referred to employees as part of the whole system, “team” or “partners”, whose performance depended on general corporate strategy. Nevertheless, banks favoring high-performance work system strive to mimic high-commitment work system (banks 11, 17) by intensifying extent, quality or using different words than expected by the actual work system. Banks favoring high-involvement work system strive to mimic high-performance work system (bank 9).

Contextual level. Banks, which favored only high-performance work system were very similar in the HR content. Explaining the disclosed intentions, they often referred to existing requirements. Banks, which favored high-commitment work system in addition to high-performance work system, added more visuals to the reports: figures, graphs, tables and pictures of people. Contrary to the previous group, they explained their intentions referring to the own code of corporate conduct, code of ethics or corporate social responsibility policy. Banks favoring high-involvement work system explicitly expressed opinion that despite own interest in optimizing operational performance, more valuable results could be gained only through business development, exploring new possibilities and negotiating new deals (bank 9). While banks from the previous two groups mentioned standardization, these banks valued universalization of employees (banks 8, 10).

Bank 1 differed in the extent, quality and content of disclosure. It stated that its report corresponded to the requests, principles and recommendations of the Central Bank of Russia, and the Corporate Governance Code; to the requirements of the Listing Rules of Moscow, London and Frankfurt Stock Exchanges; to principles of corporate governance for G20/OECD; to international standards of conduct and principles enshrined in international standards in the field of corporate social responsibility and sustainable development: ISO 26000, AA1000, as well as disclosure standards developed by the GRI group. Nevertheless, its report provided better examples of high- involvement work system than other banks in the third group. Thus, the mentioned requirements seem not to be the reason for similarity across reports, raising the question, if legitimacy is a cause or excuse for similarity in HR disclosure.

The analysis revealed four main aspects specific for the context. There was a limited freedom of choice due to high level of institutional saturation for high-performance work system, institutional gaps for high-involvement work system, and limiting accessibility of resources required for high- commitment work system. Banks had to combine work systems if they were not satisfied with high-performance work system, providing evidence of hybridization. There was a high interest in corporate volunteering and employees' wellbeing, which was perceived by different banks as part of employees' engagement, internal communication, part of motivation system or part of responsibility in HR managers' job description. According to the disclosed facts, national HR managers were heavily loaded with tasks related to managing corporate holidays for employees, their families and communities. Measuring success in HR management was mostly oriented on the process and was not generalizable or comparable across the sample, but only to previous years. This was a new trend, as in the earlier reports, banks applied widely spread methodology for measuring employees' involvement and in most cases, referred to national HR awards, easing comparison of banks' success in HR management across the sample.

Ideational level. Preference of a particular referent system caused significant difference in the content of HR disclosure. For example, banks favoring high-commitment work system focused more attention on employees' career plans and leadership management system than to training and development, more common for high-performance work system. Banks favoring high-involvement work systems often treated employees as part of their corporate strategy, while banks favoring high-performance work system mostly categorized employees under human resource policies.