Human Resource Disclosures in UK Corporate Annual Reports: To What Extent Do These Reflect Organisational Priorities Towards Labour?

There have been several exhortations to improve the practice of accounting for HR since its formal inception in the UK in the 1970s (AAA 1973; Day and Woodward 2004; McCracken et al. 2018). Some of these initiatives influenced regulatory frameworks and legislation (i.e. Companies Act 1985; Employment Act 1982; Employment Protection Act 1975; Companies Act 2006). Following the election of a Labour government in 1997, policy initiatives relating to HC management practices resurfaced in the early and mid-2000s (DTI 2003a, b) and fostered a lively debate. Among these, the Accounting for People initiative introduced by the UK government in January 2003 was seen to be a concerted effort to formalise a corporate practice of accounting for HR, thus recognising labour as a key stakeholder.

In January 2003, the UK government announced a taskforce on HC management, which completed the Accounting for People report in November of the same year. In particular, the report highlighted the possibility of accounting for people by UK firms, particularly for large quoted public firms (DTI 2003a, b). Initiatives to account for people took place in parallel with efforts to mandate the Operational and Financial Review (OFR) for UK listed firms; it was proposed that the OFR would be taken as the platform to account for HR as well. In January 2005, draft legislation on reforming company reporting of employees went to Parliament, and was legally enacted in March 2005. According to the legislation, the provision of HR information as a part of the OFR became a statutory requirement for large UK-quoted companies with the financial year commencing April 1, 2005. Surprisingly, by November 2005, the requirement to compile an OFR (inclusive of the legal obligation for UK firms to disclose information relating to their employees) was abandoned. The UK accounting profession was seen to have effectively lobbied against the proposed initiatives, leading to academic criticisms (Roslender and Stevenson 2009; Roslender et al. 2004, 2015), and impacting changes in corporate HRD practice over time. A corporate reluctance to engage with employee-related disclosures has been previously noted. For example, Day and Woodward (2004) revealed that firms disregarded even the statutory minimum disclosure requirements. Interestingly, the legal sanctions, which would have been applicable for such non-compliance, did not appear to be enforced, possibly indicating a lack of interest from firms, enforcement agencies, and/or limited market-led monitoring mechanisms. In conclusion, there has been little in the way of empirical analysis to study HRD following such key events. We now consider the theoretical framing of our study.

Firstly, stakeholder theory underlines the role of internal and external actors and the influence these actors may have on firms as a result of their financial or non-financial interest(s) (Donaldson and Preston 1995; Jensen 2002). In this regard, stakeholder theory has been relied upon to explain the reporting of social practices reflecting the informational needs of multiple stakeholder groups (Bouten et al. 2011; Pedrini 2007; Roberts 1992). Faced with demands for accountability in relation to these interests, the ethical variant of stakeholder theory suggests that organisations need to unilaterally address these demands, however multiple they might be, and are expected to adopt comprehensive, substantive HRD practices to reflect these accountability relationships. A different, and arguably more common, perspective is the instrumental variant of stakeholder theory, whereby a firm seeks to manage and placate stakeholder concerns by relying on a blend of substantive and symbolic information—in effect a more strategic approach in the mapping of the relationship between organisation and stakeholders (Soobaroyen and Ntim 2013). The power, legitimacy, and urgency characteristics of the target stakeholders guide disclosure priorities in order to ensure a strategic form of stakeholder management (Clement 2005; Mitchell et al. 1997). For example, Mitchell et al. (1997) argue that stakeholders that are more legitimate and powerful and have urgent concerns will require more corporate attention and disclosure. Hence, from an instrumental perspective, stakeholder theory underpins a need to prioritise communication to target the specific audiences that are more likely to impact a firm’s ability to maintain and sustain its activities. As previously argued by Soobaroyen and Ntim (2013), a strategic approach to organisational disclosure can involve a combination of substantive and symbolic information, since even symbolic displays of commitment may be sufficient to manage stakeholder interests in given contexts and circumstances. At the same time, a perceived or actual change in stakeholder salience may lead to a variation in the nature of the disclosures. Specifically, the decision by the UK government to endorse and foster the Accounting for People project can be seen as a change in the salience of a given stakeholder (labour/employees), and can bring out a corporate response.

Therefore, it may be argued that organisations may substantively engage in practices and initiatives to bring about genuine change in employee/labour conditions (sustainable HR) together with meeting HR procedures and processes (procedural HR), as is reflected in a comprehensive or substantive form of HRD. Alternatively, organisations may partially adopt practices and/or emphasise procedural matters without bringing about any meaningful change. A selective form of HRD may thus be observed, which would reflect a symbolic acknowledgement of employee/labour interests.

Secondly, in line with the legitimacy argument (Chen and Roberts 2010), firms voluntarily disclose or withhold information because they assume such communication is relevant and important to legitimise their actions and dispositions towards societal norms and values. Thus, firms seek to legitimise their continued survival by taking desired actions or by appearing to take desired actions (Abeysekera and Guthrie 2005; Ehnert et al. 2016). In so doing, the communication of an affinity with, and/or adherence to, the expectations of wider society in terms of how employees should be treated by their employers reflects an organisational aspiration to be doing the “right thing.” Suchman (1995, p. 578) refers to this as the pursuit of “moral legitimacy” whereby the emphasis is not on whether a given practice (HRD) will benefit a given external evaluator (employee/labour), but whether the activity is seen to be appropriate to promote societal welfare (Mahadeo et al. 2011).

In this respect, and informed by Suchman (1995), moral legitimacy can be evaluated from two perspectives: “consequential” (e.g. disclosures may refer to social outcomes) and “processual” (e.g. disclosures outline use of procedures). Within the moral legitimacy perspective, disclosures could be of any form or content (whether comprehensive or selective) since even symbolic displays of commitment may be sufficient to secure legitimacy. Soobaroyen and Ntim (2013) contend that more detailed forms of reporting may be required if a particular issue involving the organisation has become pressing from a societal perspective (e.g. rights of employees, tax evasion).

Considering the depth of disclosure in the annual reports, HRD may vary from the selection of information presented symbolically to convey an impression, to a very comprehensive analysis of information. In terms of the breadth of HRD items contained in our disclosure index, we therefore consider a continuum along which some items are inherently more related to long-term/sustainable HR investment, while others are naturally more procedural in orientation, focusing on HR practices, processes and systems. The breadth and depth of the categories are illustrated in Figs. 1 and 2.

Relying on the stakeholder and the legitimacy arguments, and on the context underlying the UK debate over the recognition of the role and contribution of labour/employees during the mid-2000s, we argue that UK firms will be expected to engage in more comprehensive and substantive forms of disclosure following the attempted reforms. This leads us to the first hypothesis:

Whilst the stakeholder and legitimacy framework outlined above can contribute to an explanation of trends in the different types of HRD over time in a given context, they are somewhat limited in terms of explaining variations between companies within the same context. In this respect, the literature already provides for a number of mainstream firm-level variables (e.g. size, industry and governance structure), which seek to reflect general managerial intentions or orientations towards a given social practice or behaviour (Barako et al. 2006; Abeysekera 2010). For example, the presence and/or prevalence of independent directors on the board has been associated to a greater extent of social responsibility disclosures, due to the influence they may have on the organisation’s strategy and approach towards social stakeholders (Barako et al. 2006), and, hence in terms of their ability to influence managerial intentions. In our case however, we extend this consideration of managerial orientations towards labour by drawing upon Brown’s (2000) work on the typology of employee relations ideologies: namely “unitarist,” “pluralist,” and “radical” forms. An understanding of using these ideologies to explain how/why certain information and disclosures are communicated to different stakeholders has only recently emerged in the literature (Mäkelä 2013).

Firstly, within a “unitarist” ideological environment, employees/labour and management/owners tends to have common and mutual objectives that are subsequently pursued in a friendly and supportive environment for HC management and disclosure (Brown 2000). In this scenario, HRD tends to take a more comprehensive or substantive form, and in recognition of the ethical variant of stakeholder theory, communication is viewed as a means to improve work relations by “…reducing conflict, overcoming suspicion and promoting greater harmony in the workplace” (Smith 1985, p. 1).

By contrast, in a “non-unitarist” (pluralist/radical) ideological environment, the relationships between labour, management, and to some extent, capital providers, with regards to the central objectives of a firm, tend to be antagonistic and confrontational as a result of divergences in their expectations relating to HR investments, management, and disclosure. According to Brown (2000, p. 51), such conflict is based on genuinely different interests. Disclosure becomes part of a conflictual accountability process between the firm and its staff, and underpins negotiations between the parties. In this case, and confronted with these pressures, we argue that HRDs are likely to be more selective in orientation, and symbolic at times (Ashforth and Gibbs 1990), reflecting the degree of conflict amongst the key corporate actors. At the other end of the spectrum, the radical perspective draws on the Marxist argument of labour exploitation and why/how companies extract the surplus value from labour to generate more profit; employers are more powerful and have more authority over employees. In this context, corporate disclosure is reflective of a capitalist logic and employee representatives are inherently suspicious of engaging with the assumptions, language and discourses underlying HRD.

In a corporate context where a unitarist ideology underpins a firm’s managerial intentions, we argue that the firm provides a more explicit position on industrial relation practices, including union activities. In contrast, an inherent antagonistic relationship in a non-unitarist environment implies that disclosures relating to industrial relations and union activities are less likely to be comprehensive and consistent in terms of their information content, reflecting apparent divergences from the expectations of the key stakeholders. Since pluralist/radicalist ideologies are associated with a state of disagreement between employee unions/labour and management/owners, we classify these concepts as a non-unitarist (i.e. pluralist/radicalist) ideology of employee relations. Therefore, we contend that a disclosure with a comprehensive coverage of HRD information would be consistent with a unitarist employee relation ideology. In contrast, a non-unitarist (pluralist/radicalist) employee relation ideology will be associated with a more selective and symbolic form of reporting. This leads to the second research hypothesis:

Content analysis has been a very common strategy in codifying and analysing voluntary disclosure practices in firms’ annual reports, such as accounting for IC (Li et al. 2008), corporate governance (Armstrong et al. 2010; Barako et al. 2006), HRD (Beattie and Smith 2010; Cormier et al. 2009), and corporate social responsibility (Bouten et al. 2011; Roberts 1992; Vuontisjärvi 2006). We adopt a similar strategy for our study of HRD practices by UK FTSE100 firms.

In order to identify if/how a firms’ disposition towards HRD changed following the UK government’s attempts to encourage more human resource-related information, a 5-year time span (from 2005 to 2009) was considered. Moreover, relying on FTSE100 firms’ annual reports from 2005 ensured that we would observe disclosures from the period when international financial reporting standards (IFRS) became mandatory in the UK and the European Union. The sample period also covers the period the Labour government was in power, given that it lost to a coalition of the Conservative and Liberal democrat parties in the 2010 elections, signalling a notable change in political priorities. Similarly, the period of analysis does not address the effects of the introduction of the UK corporate governance code (Financial Reporting Council 2010, 2012), since it would have had an impact on reporting practices involving employees, as one of the major corporate stakeholders. The sample of annual reports was limited to 210 annual reports that met the required data, covering a 5-year time period (2009—49; 2008—46; 2007—40; 2006—40 and 2005—35). One limitation of the sample selection technique is the possibility of self-selection bias. The sample, however, accounts for a minimum of 26% to a maximum of 36% of the total market capitalization of firms listed on the London Stock Exchange (34–44% of all the FTSE 100 firms) for the sample period. This arguably minimises the possibility of sample selection bias.

To-date, different types of units of measurement have been used in content analysis. These include counts of: words (Li et al. 2008; Subbarao and Zeghal 1997); lines (Abeysekera and Guthrie 2004, 2005); sentences (Abeysekera and Guthrie 2004); and paragraphs (Guthrie et al. 2004). Previous studies have primarily considered the presence or absence of information, frequency of reporting, text size (Abeysekera and Guthrie 2004, 2005; Subbarao and Zeghal 1997), and coverage in terms of breadth (Beattie 2005) of reporting. Very few studies, though, have considered the “depth” of reporting (Bouten et al. 2011), and this has long been recognised a limitation within the literature (Michelon et al. 2015). This is akin to Michelon et al.’s (2015) idea of inferring “managerial orientation” (towards corporate social responsibility) from disclosure data.

Informed by these HRD studies (Abeysekera 2008; Abeysekera and Guthrie 2004, 2005; Ax and Marton 2008; Huang et al. 2013; Subbarao and Zeghal 1997), disclosure themes have been selected and validated via a pilot study of a sample of UK annual reports (20 reports each from 2008 and 2009) prior to the main data collection exercise. A pilot study, which started with a broader list of items (39), helped refine the final disclosure index to a 22-item list. This was done through an iterative process of classifying and re-classifying items in order to develop the current HRD index, inclusive of its two categories, “sustainable” and “procedural” HRD. The HRD for a particular item is then measured in terms of its presence and absence (1 and 0, respectively), followed by an assessment of the informational availability of HRD items at each level of reporting. Briefly, “objective” refers to the availability of disclosures related to the objectives under each individual HRD item (e.g. our plan is to increase employees’ engagement). “Measurement” refers to the availability of disclosures related to measurement criteria to assess under each item (e.g. employee engagement score). “Target” refers to the availability of information regarding the periodic targets under each HRD item (e.g. increase the employee engagement score by 5% during the year). “Initiative” refers to the availability of disclosures related to the action plans under each HRD item (e.g. organise workshops and training programmes to reinforce and encourage employee engagement). Finally, “achievement” refers to the availability of disclosures related to the periodic achievement of the targets under each HRD item or related outcomes (e.g. this enables the firm to engage employees to deliver to customers and to increase business performance).

Hence, the disclosure score under each HRD item can vary from 0 (no disclosure) up to 5 (comprehensive overall disclosure integrating all the five levels). A comprehensive disclosure (score of 5) for any item in the HRD index reflects the substantive nature of the information, thereby allowing stakeholders and other societal actors to make an informed judgement on the nature of a firm’s HR practices (Marr et al. 2004).

The first research hypothesis is investigated via a descriptive/trend analysis. The second research hypothesis consists of examining the relationship between the firm’s employee relation ideology and HRD using multiple regression analysis. Key variables and the measurements used in the model are provided in Table 10 in Appendix.

The initial analyses focus on the firms’ disposition toward HRD in FTSE 100 firms in terms of: (i) percentage of firms that disclose any information in relation to each of the HRD items measured for each financial year and for the overall sample (Table 2); (ii) mean disclosure score for individual HRD items for each financial year and for the overall sample (Table 3); and (iii) percentage of firms that substantively report for each HRD item, in which case only the firms that report on all five levels or at least 4 levels would qualify as being engaged in substantive reporting (Table 4). Results illustrated in Table 2 reveal that each and every item considered under the disclosure index was seen to be important, albeit to varying extents. The percentage of firms that disclosed the various HRD items over the period appears to be very high and is relatively stable.

Items such as work environment and culture, employee know-how and competency, human resource section and function, employee numbers, employee appreciation, employee involvement, and employee experience were the most popular items to be disclosed by the firms (i.e. all the firms disclosed these items on several occasions). Value added and revenue per employee and HR director committee were the least popular HRD items since less than 50% of the firms disclosed information on these two aspects.

Furthermore, according to the mean HRD scores over the period (Table 3), both sustainable and procedural HRD increased over time. The majority of most disclosed items related to sustainable HRD (i.e. work environment and culture, employee involvement and engagement, employee training programmes, and employee expenses and pension). The most reported procedural HRD was on employee numbers, and employee health and safety. Most of the least disclosed items, such as HR director committee, employee motivation, and employee featured, related to procedural HRD. Value added or revenue per employee was the least disclosed of all the items. Clearly, however, in terms of the quality of the reported information, more significant variations can be noticed.

In determining the substantive or symbolic nature of the disclosures (Table 4), only two items (health & safety and employee/involvement engagement) could be classified as comprehensive and even in such cases, this only applied to 55% of the surveyed reports. Figure 3 presents a comparative picture of the HRD index (HR sustainable-related and Procedural HRD), presenting HRD items with the highest to lowest mean disclosure values and the highest to lowest percentage of firms substantively reporting under each HRD item. The most reported HRD items in terms of the mean disclosure value are similar to those comprehensively disclosed by a large number of companies. The HRD items, which are higher in terms of the mean disclosure value as well as substantively reported by a large number of firms, are items related to long-term HR investment, with only a few exceptions (Fig. 3). In effect, at the dichotomous measurement stage, HRD practices appear quite frequently across the sample. However, over time and once the depth of disclosure was considered, more selectivity emerged from the disclosures.

Procedural HRD items mostly relate to the categories least reported in terms of the mean HRD value and were substantively disclosed by the least number of firms, indicating that more firms appeared to be involved in a selective disclosure strategy. Several procedural HRD items, such as employee health and safety, and employee diversity and equity, accounted for a larger part of the mean HRD score, reflecting the considerable attention of stakeholders and societal actors on health and safety, and a processual form of moral legitimacy. For certain HRD items, such as value added or revenue per employee and employee experience, selective or non-disclosure was relatively frequent and reflective of a strategy to not reveal the more financial-led metrics of employee contributions (e.g. comparing value-added to average salaries in the company) (Amir and Lev 1996). In other words, there are indications of a more instrumental form of stakeholder management and moral legitimacy seeking practices than of a more ethically driven notion of accountability to stakeholders.

As evidence of the corporate response to regulatory attempts by the then Labour government, our results show a clear increase in the number of firms that started to disclose information relating to our HRD items between 2005 and 2006, except for two notable items (HR director committee and value added/revenue per employee). The drops in disclosure in 2007 and 2008 might have been attributable to the unfavourable economic conditions arising from the financial crisis. There was a generally corresponding increase in procedural and sustainable HRD, indicative of an increase in the breadth of disclosure as well. According to the annual averages presented in Table 5, with the exception of 2008, on average, the HRD indices have increased over the period, albeit marginally. This can be seen as an attempt to respond to stakeholder and government-led demands to engage with the Accounting for People agenda, and is in accordance with the stakeholder salience and moral legitimacy perspectives (Abeysekera 2008; Abeysekera and Guthrie 2004, 2005; Ax and Marton 2008; McCracken et al. 2018).

The overall HRD index revealed an increasing trend in terms of the amount disclosed. This indicates that the UK’s policy attempts to formally recognise human resources has had a positive, albeit selective, impact on some aspects of HRD in the annual reports. In addition, the eventual failure of the government’s attempts to formalise an HRD framework could have signalled to UK firms that such disclosure has become less important and/or societal attention has shifted. Similar organisational responses were noted in the case of Soobaroyen and Ntim (2013). Over the period under study however, it can be concluded that the nature of the disclosures has gradually shifted from a selective form to a more comprehensive one, and more substantive information is disclosed in the annual reports (Marr et al. 2004), thereby supporting H1, and reflecting stakeholder salience and moral legitimacy perspectives. In particular, the existence of more substantive information in relation to sustainable HRD suggests that there has been a change in the salience of the stakeholder arising from the government’s disposition towards HR reforms. At the same time, organisational variations do exist, and the next section attempts to explain this changing behaviour by considering the relationship between the company’s employee relation ideology and HRD.

A complete set of regression diagnostics was performed before hypothesis testing. The correlation matrix is presented in Table 6. All the financial variables are winsorized at the 1st and 99th percentile in order to minimise the effect of possible outliers. Results of the multiple linear regression to test for H2 are summarised in Table 7. The main analysis takes into account the year and the firm fixed-effects. Results for the robustness check with random-effects model and the pooled-linear regression are presented in Tables 8 and 9, respectively.

According to the results reported in Table 6, there is a strong positive association between the extent of the unitarist employee relation ideology of the firm and the level of HRD. The results also hold for categorical HRD indices for procedural and sustainable HRDs. This implies that H2 is supported, and thus offers new insight in terms of how a firm’s managerial orientation towards a unitarist employee relations ideology can motivate more substantive and comprehensive disclosures. As per our theoretical expectations, an environment that is characterised by both mutual dialogue, respect, and trust between management/owners and employee unions/labour (Brown 2000), and agreement on common objectives results in management being more likely to engage HRDs that are substantive in nature. The firm-specific and the corporate governance related control variables used in previous empirical studies explaining the non-financial disclosure in general, however, did not reveal any significant association with HRD when firm- and year-fixed-effects were accounted for. This suggests that HRD, compared to other social and environmental disclosures, is less (if at all) influenced by traditional organisational factors. However, when relying on a random-effects model and pooled-linear regressions, the firm-specific variables (firm size, liquidity, and profitability), governance related control variables (board independence, remuneration committee size, and nomination committee meeting), and the financial crisis dummy variable are significantly associated with HRD.