CSR in Non-Large Public Interest Entities: Corporate Talk vs. Actions

: Smaller companies’ understanding of and attitude toward corporate social responsibility (CSR), both in terms of actions and disclosure, is distinct from that of other organizations, including large public interest entities (PIEs) that dominate the existing literature in the ﬁeld. In this study, we examine the interdependencies between non-large PIEs’ CSR practices and disclosures with the use of the organizational hypocrisy concept as a theoretical lens. Our sample consists of 111 companies operating in Poland and pursuing 646 CSR-related practices in 2017. We perform content analysis of their websites using the disclosure index to assess the extent of their CSR disclosures. The total number of observations equals 1227. Both practices and disclosures are analyzed according to ISO 26000 standards. The relationship between the sample companies’ CSR talk and actions is analyzed by means of the Pearson coe ﬃ cient. Our ﬁndings suggest that CSR practices and disclosures of non-large PIEs are loosely coupled. For the whole sample and for the non-SMEs (small and medium sized entities) subsample, the strongest association between the two was observed as regards the organizational governance area. As far as the SMEs are concerned, the statistically signiﬁcant association between their CSR reporting and actions was identiﬁed for the environmental area. Our study contributes to the CSR literature, as it provides new insights into the relation between voluntary CSR talk and actions of non-large PIEs operating in a relatively unexplored setting.


Introduction
The Directive 2014/95/EU has changed the landscape of non-financial reporting in Europe. According to the new accounting requirements, large public interest entities (PIEs) with an average of over 500 employees during the financial year are obliged to disclose a non-financial statement in the management commentary or provide non-financial information in a separate report. PIEs are entities governed by the law of an EU Member State whose transferable securities are admitted to trade on a regulated market of any Member State, credit institutions, insurance undertakings, or entities designated by Member States as public interest entities. They include public companies, banks, or insurers [1]. Large PIEs' non-financial statements (or separate reports) should contain information on their business models, policies, principal risks, and key performance indicators (KPIs) relating to several matters that fall under the umbrella of corporate social responsibility (CSR), such as environment, society, or employees. Non-large PIEs (small and medium sized entities (SMEs) and other companies which do not come under the definition of large PIEs) are exempt from meeting this requirement, as it would make them incur excessive administrative burdens and costs. and (7) community involvement and development [15][16][17]. We follow the standard classification to evaluate the CSR disclosures provided by the sample companies on their websites and in their CSR reports with the use of a binary disclosure index. SMEs are a special group among non-large PIEs, and therefore we have divided the sample (non-large PIEs) accordingly into two subsamples: SMEs and non-SMEs. We formulate the following research questions: RQ1: Which of the ISO 26000 areas are most often referred to in the CSR practices of non-large PIEs operating in Poland? Are there any differences between SMEs and non-SMEs in this regard?
RQ2: Which of the ISO 26000 areas are most often referred to in the CSR disclosures of non-large PIEs operating in Poland? Are there any differences between SMEs and non-SMEs in this regard?
RQ3: Are there any interdependencies between the CSR practices and disclosures of non-large PIEs operating in Poland? Are there any differences in these interdependencies as regards the ISO 26000 areas and both subsamples (SMEs and non-SMEs)?
According to the study findings, the practices reported most often by the whole sample as well as both subsamples (SMEs and non-SMEs) fell into the category of community involvement and development. As far as the corporate disclosures are concerned, 1227 observations were made, including 205 for the SMEs subsample and 1022 for the non-SMEs subsample. This difference clearly indicates that the investigated companies represent various levels of engagement in providing CSR disclosures. The three areas most often reported on were community involvement and development, labor practices, and environment. The same reporting pattern was observed as regards the whole sample, as well as both subsamples.
The analysis of the corporate CSR practices allows us to state that practices concerning such ISO 26000 areas as labor and environment differ significantly between SMEs and non-SMEs. The significant differences are also observed as regards the CSR disclosures related to all CSR areas except for human rights. Our main findings suggest that CSR talk and actions of non-large PIEs are loosely coupled. We have found a statistically significant association between the extent of overall disclosures and total practices for the whole sample, and for the non-SMEs subsample. The strength of this association was the highest for the organizational governance area. However, for the SMEs subsample, the statistically significant association exists only for one out of seven areas mentioned by the ISO 26000 standard, namely the environment.
We believe that our findings contribute to the CSR literature for three reasons. Firstly, non-large PIEs' engagement in CSR, in contrast to large PIEs, remains relatively under-researched. Many studies that were published after the introduction of the Directive 2014/95/EU focus on large PIEs' disclosure practices [18]. Secondly, despite the researchers' increased interest in the CSR practices and disclosures of non-large companies, most of the studies in this area refer to SMEs. To the best of the authors' knowledge, the companies that do not fall into the SME or PIE categories have not been covered by the up-to-date studies. Thirdly, we investigate an unexplored setting-a Central Eastern European (CEE) country with an ambiguous attitude towards CSR. According to e.g., [19][20][21][22][23], present literature is based specifically on Anglo-American and West-European countries. In this way, we provide evidence about a new institutional context as regards both the country and regulatory framework. Fourthly, our study refers both to corporate practices and reporting, and the relations between them, not to each of these aspects separately. Since the financial crisis of 2008, organizations have been under pressure to find innovative ways of connecting business performance with the impact on society and the environment, and to effectively communicate with stakeholders. We believe that we provide evidence that this trend has also been spreading among non-large companies, including SMEs.
The remainder of the paper unfolds as follows. The next section presents a summary of the relevant academic literature which underpins this paper. Next, the theoretical background is presented. Then the hypothesis is developed based on the conducted literature review. In the following section the methodology employed for empirical research is briefly explained. The paper then proceeds to present the results of the study. The paper concludes by summarizing the findings, presenting the overall contribution, addressing the limitations of the research, and suggesting recommendations for future studies.

Background Literature
The CSR-related research is usually undertaken in the literature in the context of larger companies such as public companies, often multinational enterprises (MNCs) [4,24,25]. It is quite understandable given the fact that size has been identified as a factor which influences both corporate practices (e.g., [26]) and their non-financial disclosures (e.g., [19,27]). The reason for this is the greater significance of larger firms for the economy and society, and related greater political, regulatory, and media pressure ( [28]). The more the company is visible, the more it is exposed to claims made by the media and the public, and the more likely it is to engage in CSR actions ( [29]). Furthermore, disclosure cost is lower for larger firms ( [27,30]). Larger companies' non-financial practices and disclosures were found to be significantly influenced by their corporate governance mechanisms. In general, companies with good governance have better performance ( [31]). Board composition influences both CSR actions and talk ( [32]). The presence of CSR committee was also found to be an important factor determining CSR disclosure ( [33]). When it comes to CSR decisions and related disclosures in smaller firms, owner-managers' judgment is crucial as they are the key decision-makers ( [34]). In general, the companies' leaders' behavior, and the values they share, are considered critical to these organizations' success (e.g., [35]).
Given the fact that the bigger the company, the more advanced it is in CSR practice and disclosure, it is understandable that the researchers' attention is focused on larger firms. After the introduction of the Directive 2014/95/EU, large PIEs (public interest entities) were often investigated (e.g., [36][37][38][39][40]). Researchers relatively rarely approached non-large PIEs. However, environmental and societal impacts of SMEs and other non-PIEs, which do not meet the Directive requirements, are often significant and, as mentioned before, the nature of their CSR practices is likely to be influenced by their specific characteristics.
For example, as far as SMEs are concerned, ref [41] argues that although they are not a uniform group, they have significantly different behavioral characteristics in respect of CSR management compared to larger firms. SMEs have relatively informal organizational structures and are often managed by their owners. As a result, personal choices and attitudes can significantly affect their socially responsible behavior [42][43][44][45]. There is no doubt that many small companies are naturally adjusted to the personal attitudes and values of their founder or owner (or the head of a family if the company is also a family business). They are also more powerfully impacted by their local embeddedness than PIEs or especially MNCs. SMEs are particularly tightly integrated into the local communities where they operate, and there is a sort of 'closeness' between the firm and the environment. According to [41], this 'closeness' is defined as a fundamental feature of SMEs because their founders and managers are embedded in the company and involved in community matters. Therefore, they are often perceived as local benefactors or activists [43]. The owner-managers of SMEs operating in developing countries were found to consider community development and employee welfare to be the two most important aspects of CSR [17]. Environmental matters are not central CSR issues for them. What is more, some companies do not even consider that their operations have environmental impacts and do not pay much attention to this problem.
Due to these close relationships with the environment they operate in, non-large PIEs are more likely to engage voluntarily in CSR in their local community by supporting local events and residents, creating jobs, and providing innovation. This engagement in CSR activities is generally more difficult and challenging for them than for PIEs, particularly due to lack of the owner-manager's understanding of CSR, lack of time, funds, and other resources [9]. Therefore, according to [4], smaller companies are more dependent on their partners and local public actors and governments to support their CSR actions. For all companies, regardless of the size, cooperation with society is a key driver in promoting and implementing CSR practices [7]. However, ref [46] indicates that small companies usually do not have such well-developed communication and public relations cells in their organizational structures, which makes it more difficult for them to interact efficiently with stakeholders when compared to bigger firms.
According to [41], more has been written on the adoption and interpretation of CSR practices by SMEs than on the disclosure of these actions through the publication of social and environmental (CSR, sustainability) reports. As [9] notes, smaller companies make less use of CSR tools than larger firms, have less formal CSR strategies, scarcer funds to invest in CSR actions, and are less expected to provide voluntary CSR disclosures. The barriers to engaging with CSR reporting that are encountered by SMEs may include lack of knowledge of CSR communication and related benefits, and the perceived low stakeholder interest in their nonfinancial disclosures. What is more, SMEs may also feel that talking on CSR achievements may be regarded as boasting [47] or may involve increased bureaucracy due to governmental requirements regarding measurement and reporting procedures. SMEs may consider CSR reporting as a superfluous and burdensome activity due to the high expenses of data collection and management, and the lack of resources [48]. According to [49] (p. 364), small companies which are involved in supply chains may also 'experience inconsistency between the idealistic CSR communication of some large companies and their actions'. Nevertheless, there is a growing pressure exerted on SMEs to make their implicit CSR communication more explicit. This applies especially to SMEs suppliers whose customers are MNCs [49,50].

Theoretical Underpinning and Hypothesis Development
Proponents and critics of non-financial reporting have their own arguments as regards its relation to corporate practice. Proponents see the potential of CSR disclosures for making firms more accountable and transparent about their social and environmental impacts. Management uses CSR reports to demonstrate that it cares about the non-financial aspects of the firm's operations and manages social and environmental risks. Critics question voluntary CSR reporting because it tends to be limited in scope and untruthful. Some argue that there exists divergence between companies' CSR-related policies, programs, and impacts, which leads to the situation where CSR is purely ceremonial [51,52].
They also point out that 'corporations do not walk the sustainability talk' [53] (p. 79), [54,55], which leads to the situation where CSR reports present false claims and commitments rather than rational plans and strategies about how to deal with certain social and environmental problems. Research findings suggest that companies engage in CSR reporting mainly to secure their own position and private interests. Corporate managers use it to muddy, glance, and complicate the real CSR performance. Nevertheless, corporations are expected to provide complete and transparent accounts of their CSR efforts. As [53] (p. 81) argues, 'discrepancies between corporate talk and action are problematic, since without trustworthy reporting neither is accountability fulfilled nor is society able to evaluate corporate activities and impacts appropriately'.
The theories usually applied to the problem of CSR disclosure and practice are typically stakeholder theory, institutional theory, and legitimacy theory [56]. Ref. [53] refer in their study to two other concepts, namely organized hypocrisy and organizational façade. In this paper, we focus on the former. The organized hypocrisy model, discussed by [53] (p. 79) and presented by Brunsson for the first time in 1989 [57], 'attempts to explain the discrepancies between a corporation's talk, decisions, and actions, and how these discrepancies may allow corporations flexibility in their management of conflicting stakeholder demands'. Brunsson's model distinguishes between companies' talk, decisions, and actions [57]. Talk is understood as spoken or written words that an organization uses to communicate with stakeholders. Together with decisions and actions, they are three outputs of organizations that are considered as tools of legitimacy and provide different approaches to stakeholder management. Organized hypocrisy occurs when the organization shows inconsistency across these outputs, namely when talk and decisions are believed to be inconsistent with actions. All three outputs are still causally related ('coupled'), rather than 'decoupled' or 'loosely coupled', but the causality is the reverse. The organization's talk or decisions in one direction decrease the likelihood of corresponding actions, while actions in one direction decrease the likelihood of corresponding talk and decisions. This way, corporate talk and decisions may compensate for inconsistent actions, and vice versa-actions may compensate for inconsistent talk or decisions [53,58].
According to [53] (p. 84), legitimacy and signaling theories 'allow to conclude that organized hypocrisy is associated ultimately with negative outcomes for broader society. However, taken strictly on their own terms as ways in which organizations manage conflicting stakeholder demands, organized hypocrisy may indeed make room for potentially positive outcomes for broader society'. Hypocrisy can offer opportunities for positive consequences and should, therefore, be tolerated [59]. However, to make this happen, organizational talk must not be dishonest; instead, it should be aspirational. 'The tricky issue here, of course, is whether the motive behind hypocrisy is to conceal an unpleasant truth or to reduce the difference between current and aspirational reality' [53], [59] (p. 385).
In this paper, we use the organizational hypocrisy approach to explain the relations between the non-large PIEs' practice and reporting. The factors that might influence the associations between the CSR 'actions' and 'talk' of non-large PIEs include the management decision, the high costs of disclosure, and the lack of adequate reporting standards.
The loosely coupled relationship between smaller companies' reporting and disclosures might be caused by the fact that their approach toward CSR is often 'sunken' and not 'explicit' [12,41]. Managers of non-large PIEs may not implement CSR reporting, although the company has CSR practices in place, if they do not see that it might benefit them directly. They are less likely to report non-financial information because they are not aware of the benefits related to CSR disclosures or lack technical knowledge. Sometimes they may not even perceive certain CSR-related practices as such and, at the same time, worth sharing with stakeholders. As [60] points out, a survey conducted in Danish SMEs shows that communication of CSR is disorganized and handled on an ad hoc basis. CSR-related decision making, action, and communication are mostly informal and based on personal relations. According to the authors, this approach to corporate communication does not take into consideration that CSR reporting influences SMEs' reputation and stakeholder relations.
External pressure from stakeholders, including regulators, or, more precisely, the lack of such, might also be relevant. Ethical standards developed mainly for large PIEs might be inappropriate for small firms [61]. Besides, from the non-large PIEs' perspective, the possible ineffectiveness of voluntary standards might be explained by the requirement of greater investment in terms of time, funds, and knowledge. What is more, none of the many different types of non-financial reporting frameworks which have been developed seems to meet the increasing needs of SMEs. The Global Reporting Initiative (GRI) guidelines [41] are no exception.
Implementation of CSR initiatives is thought to make smaller companies incur a high cost with no short-term benefits. The managers who are obliged to focus on the economic situation of a company are primarily concerned with ensuring its survival. This is true especially in the CEE countries [62]. Therefore, the prevalent view is that SMEs are unlikely to report CSR information due to their financial constraints and the opinion that they do not have a performance on which they might report [41]. However, contrary to this view, ref [63] demonstrates that SMEs report non-financial information regardless of their financial restrictions. These findings allow the assumption that for smaller companies, good reputation is also important in attracting resource holders and meeting their expectations [64]. Therefore, the awareness of the importance of CSR disclosures for their reputation may force SMEs to disclose more than they do in practice.
This dissonance between what companies do and what they report on is a source of criticism aimed at CSR reporting. According to [65], prior literature argues that there exists a gap between actual organizational practices and their representation in voluntary disclosures. As [11] argues, smaller firms (such as non-large PIEs we focus on) are challenged by a lack of resources such as human or financial capital, and they face less public pressure to engage in CSR. However, it is not correct to assume that small firms 'do less' in terms of CSR. Many of them engage in CSR 'silently'. They do implement some sort of CSR strategies, but, at the same time, they typically do not publish CSR reports or provide CSR-related disclosures on their websites. According to [50] (p. 6), SMEs' CSR communication is characterized by informal and poorly articulated implementation of a wide range of socially responsible practices and procedures rather than explicit and formalized CSR communication (see also [66]). Thus, the visibility of their CSR practices tends to be low. As a result, smaller firms seem to be 'walking', but not 'talking' CSR.
Given the above, we are not able to predict if the non-large PIEs' CSR practices and disclosures are related. Therefore, we formulate our hypothesis as follows: Hypothesis 1. CSR practices (actions) and disclosures (talk) of non-large PIEs are not coupled.

Research Sample and Method
Our research focuses on Polish companies included in the 16th edition of the report 'Responsible Business in Poland. Good Practices' issued by the Responsible Business Forum (RBF) [14]. Being referenced in this report means that a company has proved its CSR engagement by voluntarily subjecting the CSR-related practices to evaluation by RBF experts. RBF is one of the most active and prominent non-governmental organizations which aim at supporting CSR initiatives in Poland. The report entitled 'Responsible Business in Poland. Good Practices' is the largest overview of socially responsible business activities undertaken in a particular year by companies as well as institutions and non-governmental organizations operating in Poland.
The 'Responsible Business in Poland. Good Practices' report has been issued annually since 2002. Both the number of companies and the number of their CSR-related practices included in the subsequent editions of the report have been growing from year-to-year. In the 16th edition of the report, published in 2018, the total number of submitted practices (new and long-term ones) equals 1190 and is by 30% higher as compared to the previous edition [67]. The number of entities reporting their engagement in CSR in 2018 edition is 177 and, compared to the first report issued in 2003, increased by 63 [14,68]. These positive changes suggest that the Polish companies' engagement in the CSR practices rises with time.
Our initial sample consisted of 177 companies subsumed in the RBF 2018 report. The selection of the final sample consisted in excluding institutions, non-governmental organizations, large PIEs, companies with websites in languages other than Polish, and companies with no websites. The final sample comprises 111 companies. As indicated in the literature review section, factors such as company size, related media visibility, and governance mechanisms were found to be especially relevant for the investigated topic [19,[26][27][28]. Therefore, the total sample was divided according to size. Two subsamples were distinguished: SMEs (32) and non-SMEs (non-large PIEs which do not meet the SMEs criteria; their number equals 79). The list of the analyzed companies is presented in Appendix A.
The proxy for the companies' engagement in CSR actions is the number of CSR practices included in the RBF report. The practices are analyzed in total and according to the seven CSR areas proposed by the ISO 26000 standard. The information about the practices was obtained from the report, which is available in a PDF file, and inserted into an observation sheet in an Excel file. The CSR practices included in the report are classified into the following seven areas according to the ISO 26000 standard: (1) Organizational governance, (2) human rights, (3) labor practices, (4) environment, (5) fair operating practices, (6) consumer issues, and (7) community involvement and development. The ISO 26000 standard was also used to evaluate sustainable business activities of SMEs by, e.g., [17].
The analysis of the corporate CSR talk is performed with the use of disclosures posted on companies' websites (including information presented in CSR reports). The disclosures are also divided into the above-mentioned seven CSR areas in line with the ISO 26000 standard. For each area, more detailed categories (disclosure items) are identified with the use of RBF report. Their number varies from 4 to 17 depending on the CSR area, with the largest number (17 items) for community involvement and development, and the lowest (4 items) for human rights.
The content analysis of the companies' websites and reports is performed with the use of a binary disclosure index (0-an item is not disclosed, 1-an item is disclosed). The proxy for the extent of the CSR disclosures is the number of scores assigned to the disclosure items reported by the company. The maximum number of scores is 61, and the minimum is 0. Both authors were responsible for the coding process in order to increase the reliability of coding. The assigned scores were compared, and any discrepancies were discussed in order to reach consensus. The total number of usable observations equals 1227.
The significance of the differences between the SMEs and non-SMEs subsamples concerning the CSR practices and reporting is analyzed by means of the Wilcoxon rank-sum test due to the fact that these subsamples are not equinumerous. The Pearson correlation coefficient is used to investigate the relationship between CSR disclosures and practices. The hypothesis is tested at the significance level of 0.05.

Research Results
As far as CSR actions are concerned, the analyzed companies submitted a total of 646 CSR practices to the 16th edition of the "Responsible Business in Poland. Good practices" report. Most of them (561) were submitted by larger companies (non-SMEs) ( Table 1). For the whole sample, the most often reported practices were those related to community involvement and development (35.8% of all reported to the RBF by the sample companies), followed by labor (29.4%) and environment (16.7%) areas.
Not surprisingly, in the SMEs subsample, almost half of the practices reported to the RBF concern the ISO 26000 community involvement and development area (49.4%), which supports the findings of [41,43]. The companies were involved, among others, in philanthropic and charitable actions, education of children and youth, and public service announcements. Actions connected with labor practices area account for 21.2% of all SMEs' CSR activities. Other areas, including environmental issues, were far less often mentioned in the RBF report.
Non-SMEs' practices related to community involvement, labor practices and environment are far more often indicated in the RBF report than the others (organizational governance, fair operating practices, consumer issues, and human rights) (Table 1). Altogether, they make up 82% of the non-SMEs' practices in total. Table 2 presents the number of CSR disclosure items for the total sample and for the two subsamples. One thousand, two hundred and twenty-seven observations were made, including 205 for the SMEs subsample and 1022 for the non-SMEs subsample. The number of disclosure items and their share in the total disclosures provided by the sample companies were the highest for the community involvement and development area (28.1%), followed by labor practices (24.8%) and environment (19.5%). When comparing the corporate 'actions' and 'talk' in both subsamples, we notice that in general they follow a similar trend, with the human rights, fair operating practices, and consumer issues being practiced and reported the most rarely. At the same time, companies' engagement in labor practices and community involvement and development is the highest as regards both corporate practice and disclosure.
Tables 3-5 present descriptive statistics regarding the number of CSR practices and the extent of CSR disclosure for the total sample and the SMEs and non-SMEs subsamples. On average, companies declare 5.8 CSR activities and disclose 11.1 CSR-related items ( Table 3). The maximum number of practices is 16, and the maximum number of disclosed items is 11. Both refer to the area of labor practices. The sample is diversified as regards both CSR practices and disclosures, which is testified by the values of the particular coefficients of variation. Twenty-nine out of 111 companies submitted only one CSR activity to the RBF for evaluation. Almost one-fifth of the companies (19.82%) do not disclose any CSR-related information on their websites. The average number of practices declared by SMEs equals 2.7 (Table 4). When it comes to CSR reporting, SMEs disclose on average 6.4 CSR-related items. The maximum number of practices, and the maximum number of disclosed items refer to the labor practices area and equal 4 and 11, respectively. The diversification of the disclosures in the SMEs subsample is higher and the diversification of practices is lower, as compared to the non-SMEs subsample. The value of the coefficient of variation is 126% in the case of total disclosures and 84% in the case of total practices.
Larger enterprises submitted on average 7.1 practices to the RBF for evaluation and inclusion in the report ( Table 5). The maximum number of practices is identified for the labor practices area and equals 16, and the maximum number of disclosed items, for the same area, is 11, which is the same result as for the total sample. As regards CSR disclosures, the average non-SME discloses information on twice as many items as its SME counterpart (12.9 and 6.4, respectively). The non-SMEs sample is diversified in terms of CSR activity and reporting (the coefficient of variation equals 102% for CSR practices and 76% for CSR disclosures).
To summarize, both the practices and disclosures are diversified as regards the total sample and the two subsamples. The diversification of practices is higher than that of disclosures for the total sample and for the non-SMEs subsample. As regards the SMEs subsample, disclosures are more diversified than practices. Table 6 presents the results of the Wilcoxon rank-sum test. The findings indicate that there are significant differences (at the level of 0.05) between the SMEs and non-SMEs subsamples regarding the total number of CSR-related practices included in the RBF report, and the number of practices related to ISO 26000 areas such as labor practices and environment. There is also a significant difference between both subsamples as regards the overall level of disclosures, as well as disclosures on all ISO 26000 areas, except for human rights. The differences between the total number of practices and total number of disclosed items are in line with the previous findings, which indicate that SMEs are less engaged in these areas than larger firms (e.g., [9,46]). It seems reasonable that larger companies which employ more workforce have more labor-related practices in place. As regards the environment, our findings are in line with the study by [17], who reported low interest in environmental matters expressed by SMEs. Table 6. Results of Wilcoxon rank-sum test for the SME and the non-SME subsamples. In order to verify our hypothesis, we have analyzed the correlation between CSR practices and disclosures for all non-large PIEs and the SMEs and non-SMEs subsamples. We use the data concerning the sample companies' actions and talk in general as well as for each of the ISO 26000 areas separately. Table 7 presents the values of the Pearson correlation coefficient. We find a statistically significant association between the extent of overall disclosures and practices as well as between disclosures and practices regarding five CSR-related ISO 26000 areas: Organizational governance, labor practices, environment, consumer issues, and community involvement and development (at the level of 0.05 significance). However, the strength of this association is moderate only for one CSR area, namely organizational governance (i.e., the correlation coefficient is above 0.4 and below 0.7 [69]). The strength of the rest of the significant relationships is weak (i.e., the correlation coefficient is above 0.1 and below 0.4 [69]). The organizational governance practices and disclosures refer to such issues as compliance with laws, ethical codes, or anti-corruption policy. The CSR-related disclosures concerning the above issues seem to be better coupled with actions than in the case of other ISO 26000 areas.

CSR practices
Similar results hold for the non-SMEs subsample. For the six identified significant relationships between CSR disclosures and practices, only that of organizational governance is moderate. As far as the SMEs subsample is concerned, there is no statistically significant association between their CSR reporting and actions except for the environment area (the association is moderate).
Based on these results, we can conclude that our hypothesis is not supported. There is a loosely coupled relationship between CSR practices and disclosures of non-large PIEs.

Discussion and Conclusions
As [70] notes, we can observe an enormous increase in the interest in CSR from different perspectives and interest groups. In response to various external pressures, including institutional, but also in line with personal beliefs, managers have introduced sustainability criteria into the corporate strategy. However, putting CSR into practice remains a challenge.
As far as CSR reporting is concerned, large PIEs operating in the EU are obliged to follow regulations imposed on them by the Directive 2014/95/EU. For most of the non-large PIEs, CSR reporting remains voluntary. Understandably, non-large PIEs do not have the same incentives and resources to implement sophisticated CSR strategies and communication plans as large firms, including MNCs.
In our study, we focus on 111 non-large PIEs operating in Poland. We investigate their CSR practices and disclosures to find out if they are coupled. Our analysis is based on the dataset encompassing a total of 646 CSR practices and 1277 CRS-related disclosure items identified on the sample companies' websites. Both the practices and disclosures are diversified as regards the total sample and the two subsamples. The research findings do not allow us to fully support our hypothesis, according to which CSR practices and disclosures of non-large PIEs are not coupled. We have found a statistically significant association between the extent of overall disclosures and total practices and between disclosures and practices regarding five CSR-related ISO 26000 areas for the whole sample and for the non-SMEs subsample. Therefore, we can conclude that CSR actions and talk of non-large PIEs are in a loosely coupled relationship. However, for the six identified significant relationships, the strength of only that of organizational governance is moderate, whereas the strength of the other significant relationships is weak. This finding suggests that in the case of the larger, and at the same time more visible, companies included in our sample (non-SMEs), more attention is given to the issues related with compliance, ethics, or anti-corruption policy. It might result from the fact that larger companies operating in Poland follow the western companies' standards in this regard or they perceive compliance with the law and codes of ethics as the fundamental part of CSR. Organizational governance seems to be important to the companies, and at the same time, practices and reporting strategies related to this area seem to be easier to implement in a coherent way. With this finding, we argue that strong governance mechanisms and law enforcement can create an environment which facilitates better alignment between corporate CSR performance and disclosures.
As regards SMEs subsample, a statistically significant association can be observed only in the case of one out of seven ISO 26000 areas-environment. This might be accounted for by the fact that local regulations and guidelines concerning the environment are more advanced as compared to community, labor, or customer-related areas. For example, in Poland, the information regarding the use of the environmental resources and the GHG emissions must be reported by the companies to the Marshal Office. On the EU level, the second set of European Commission's guidelines published in 2019, focuses on climate change. It provides more detailed guidance as regards material and comparable corporate environmental disclosures and includes a set of KPIs concerning: GHG emissions, energy, physical risk, products and services, and green finance [71]. Our study findings might suggest that better regulations can lead to better alignment between the companies' disclosure and practice. Therefore, practice and disclosure in other CSR areas, especially related to the employee and social matters, should be encouraged by further regulations and guidelines put forward on the national and EU levels.
Our study provides unique insights into the essential, yet understudied community of non-large PIEs, which are not SMEs. These companies are not covered by the research agendas, and therefore not enough research has been conducted so far as to broaden our understanding of the very nature and dynamics of their CSR activities and disclosures. The research findings contribute to the literature in several ways. They provide substantial insights into CSR and its importance to non-large PIEs in a developing country with a short CSR-related history. It also offers new information on a crucial relation between companies' CSR actions and talk. Our results support the findings presented in the CSR literature according to which SMEs' engagement in CSR disclosure is lower than that of larger companies [9,47]. Nevertheless, as [49] (p. 364) notes, 'the absence of social reporting does not imply that SMEs do not behave responsibly'. In accordance with the hypocrisy theory and in line with our findings, we believe that, as far as SMEs are concerned, the relationship between CSR talk and actions is even more loosely coupled than in the case of non-SMEs. These results are in line with [50] (p. 6), who observed that: 'Although some explicit CSR communication may occur ( . . . ), SME communication is more embedded in the CSR practice itself than in a post facto formal report of that practice.' On the one hand, this finding may indicate that in order to avoid transmitting inconsistent signals to their stakeholders, SMEs should increase their interest in the CSR concept and try to integrate CSR talk and actions. However, on the other hand, as [53] note, the dissonance between the two is not bad, provided corporate talk is not dishonest, but aspiring and motivating. For example, the disclosure on the implementation of the environmental protection strategy in one year should lead to the real environment-friendly actions in the following year.
We believe that our findings are also relevant for current regulation and practice. Our study clearly indicates that non-large PIEs operating in CEE EU Member Country practice CSR and report on CSR-related activities voluntarily. The interest in the CSR concept development in smaller companies is strongly supported by the EU. On February 20, 2020 the European Commission launched a public consultation to collect the views of stakeholders about possible revisions to the provisions of the Directive 2014/95/EU [72]. One of the issues concerned the possible extension of the Directive's scope to encompass the SMEs. According to 89% of respondents who expressed their opinions on this matter, the simplified standard and/or reporting format for SMEs should be introduced. The opinions on whether it should be voluntary or mandatory were mixed. Nevertheless, it seems that there is a general agreement that SMEs should be engaged to a larger extent in providing non-financial disclosures. Given the fact that non-financial disclosures are believed to place an administrative burden on small companies and therefore should not be required, these companies should get additional support for CSR practice and reporting, especially from governmental agencies.
Our findings can also be useful for accountants and managers involved in implementing sustainability strategies in non-large PIEs. The results derived from our analysis could make them aware of the fact that there are dissonances between corporate talk and actions, and some additional efforts might be needed to integrate them better. For instance, SMEs' owners and managers could put stronger emphasis on governance and enforcement mechanisms. Given the example of larger firms, coupling better their talks and actions in the organizational governance area might be easier than in other areas. What is more, if we consider organizational area as this part of corporate activity in which laws, codes, and policies better align corporate practices and disclosures, then we could argue that stricter external and internal regulation in other areas, such as environmental issues, human rights, or labor practices could be beneficial for better alignement of related non-large PIEs' CSR practices and disclosures.
This study is not free from limitations. Our sample is relatively small. However, it deliberately includes only these companies which have some experience with CSR practice and reporting. We believe that we have managed to cover most SMEs familiar with the CSR concept. However, we might have omitted some SMEs which operate in Poland and do provide CSR-related disclosures, but do not submit their CSR practices to RBF for evaluation, and therefore are not included in the RBF report.
We believe that non-large PIEs and their engagement in the CSR practice and reporting are worth further investigation. Studies could focus on the role of media visibility, and corporate governance mechanisms with a view to investigating more thoroughly their role in the context of non-large PIEs' CSR actions and disclosures. The problem of the dishonest or aspirational nature of non-large PIEs' voluntary talk also seems to be worth researching. Therefore, a more detailed analysis of the CSR disclosures provided by the non-large PIEs most engaged in CSR could be performed. The use of other research methods, such as semi-structured interviews, is also advised. The interviews could be conducted among CSR managers employed in the non-large PIEs to find out about their attitudes towards the discrepancies between their companies' CSR actions and talk.