Employer Silencing in a Context of Voice Regulations: Case Studies of Non-Compliance

This article, drawing on the latest insights into organisational silence, considers how employers seek to withhold information and circumvent meaningful workplace voice when confronted with regulatory requirements. It offers novel theoretical insights by redefining employer silencing as characterised by the withholding of information and the restriction of workplace dialogue. In outlining three principal routes of non-com-pliance — avoidance, suppression, and neglect — we empirically illustrate the path to silence in the regulatory context of the European Union Directive establishing a general framework for informing and consulting employees. Rather than considering how employers utilised the regulations, as existing research considers, we look at how employers circumvented the regulatory space in three case studies in the United Kingdom and Ireland and the significant role of employer silencing as a tool for explaining this dynamic.


2014)
. Although the silence-voice literature traditionally focuses on how employees withhold information and refrain from speaking up, we focus on how employers withhold information and restrict workplace dialogue.
Indeed, to conceptualise silence in this way offers an important analytical tool in enhancing understanding of employee voice. Whereas the latter looks at the ways and means in which employees can express their views within the workplace, employer silencing, in contrast, considers the agency of how employers silence employee voice by failing to either disclose information or abstaining from genuine dialogue, which is a vital precursor for meaningful employee voice to occur in the first instance. Employer silencing is understood here as management engaging in acts of silencing workers through the withholding of information and the restriction of workplace dialogue. Employer silencing is analytically distinct from extrapolating the presence or absence of employee voice policies and practice, and it differs from workers remaining silent in response to management. As such, this article breaks new ground by advancing the concept of employer silencing of employee voice to improve understanding of the nuanced duality of voice-silence dynamics in employment relations and human resource management (HRM) scholarship.
Regulatory context is essential in enabling or constraining employer silence. The broader regulatory context in many jurisdictions can constrain employer efforts to engage in silence on a variety of matters (Emmenegger, 2015;Marchington, 2015). For example, there has been growth in many countries of "hard" and "soft" regulations providing for more social dialogue in the workplace. The most notable developments in recent decades is the European Union (EU) Directive Establishing a General Framework for Informing and Consulting Employees (2002/14/EC, hitherto the Information and Consultation (I&C) Directive; Gold, 2010). These regulations potentially constrain employer attempts to withhold information and avoid consultation with the workforce.
The existence of regulatory constraint raises a question over how employers preserve silence in response to regulatory requirements intended to do the opposite, that is, legislation aimed at establishing information disclosure and workforce dialogue. Current research on the I&C Directive, for example, examines how employers use the regulations to provide information and create consultation mechanisms. In this article, however, we focus on how the Directive's intended benchmarks are subject to a typology of non-compliance behaviours, effectively routes to employer silence. Specifically, our typology of regulatory non-compliance identifies three ways by which employers withhold information and restrict dialogue despite legal requirements to establish their precise opposite; these are "avoidance," "suppression," or "neglect." The article illustrates these three routes empirically by using qualitative case studies derived from studying how regulations providing for greater information disclosure and dialogue via the I&C

WHAT IS CURRENTLY KNOWN ABOUT THE SUBJECT MATTER
• Existing research sees silence as employee-led/workers choose to withhold voice; • Existing research examines how the European Union Directive for Informing and Consulting Employees, and national laws, is utilised.

WHAT THE PAPER ADDS TO THIS
• Organisational silence can be employer-led ("employer silencing"); • Novel insights in defining employer silencing as "management engaging in acts of silencing workers"; • New typology of employer silencing of Information and Consultation regulations voice requirements: avoidance, suppression, and neglect.

THE IMPLICATIONS OF THE STUDY FINDINGS FOR PRACTITIONERS AND POLICY-MAKERS
• Employer silencing causes a democratic voice deficit; • Problem of non-compliance with regulatory requirements for voice; • Information and Consultation regulations, too weak to institutionalise voice at work, need re-casting.
Directive were subject to employer circumvention and marginalisation. The evidence derives from the United Kingdom and Ireland, where some scholars believed the Directive would potentially hold most implication and potential impact (Dundon, Dobbins, Hickland, Cullinane, & Donaghey, 2014). This approach follows Kaufman's (2015) call for greater understanding of employer activity designed to stifle employee voice.
The contribution and value of the article are at two levels. First, our conceptualisation of employer silencing offers an expanded understanding of the possibilities for voice than merely focusing on voice and voice mechanisms alone. Voice cannot meaningfully function if denied from the information upon which valid and reliable decisionmaking occurs. The concept of employer silencing, therefore, offers a useful conceptual appendage to both the voice and silence field of study more generally. Second, the paper contributes to our understanding of the regulatory context of the I&C Directive. Our analysis offers a new understanding of how the Directive failed in its objectives to enhance voice and social dialogue at workplaces through our conceptualisation of employer silencing as a form of regulatory non-compliance characterised by the typology of avoidance, suppression, and neglect. The next section reviews some existing literature on silence and voice, followed by a section outlining the specific regulatory context of the EU I&C Directive.

| SILENCE AND VOICE
Although the notion of employee voice is well established and understood, organisational behaviour literatures have recently added the concept of "silence" to the conceptual repertoire of the field (; Morrison, 2011;Tangirala & Ramanukam, 2008). Morrison and Milliken (2000) first presented the term "organisational silence" as a discrete construct. These authors understood organisational silence as a collective level dynamic wherein employees withhold their opinions and concerns about organisational problems. It is motivated by what they call a "climate of silence", that is, a widely shared view among employees that speaking up about issues or problems can be useless or indeed perilous (see also Milliken, Morrison, & Hewlin, 2003). Pinder and Harlos (2001) complemented this path-breaking research by introducing the concept of "employee silence." Employee silence offers an individual-focused lens assessing how individual employees withhold information on work problems to persons who are perceived capable of effecting redress. Later, Van Dyne, Ang, and Botero (2003) extended this and, building on Pinder and Harlos, proposed different forms of employee silence derived from varying motives, for example, acquiescent silence based on resignation, defensive silence based on fear of consequences, and pro-social silence based on protecting proprietary knowledge to the benefit of the organisation (see also Brinsfield, 2014;Edmondson, 2003;Huang, Van de Vliert, & Van der Vegt, 2005). The organisational behaviour literature's insights later migrated into the field of employment relations (Nechanska, Hughes, & Dundon, 2018). Donaghey, Cullinane, Dundon, and Wilkinson (2011), for example, treated silence as not simply a product of employee communicative choices but as actively engineered by employers who fail to provide voice structures or provide those which are ineffectual, impotent, and rule out discussion on issues of concern to employees (see also Barry et al., 2018). We might add to this the concept of employer silencing that is where employers not only withhold information from the workforce (see on information disclosure) but also prevent dialogue with the workforce on issues that encompass employment contract matters, job prospects, work organisation, and the firm's economic circumstances. Arguably, employer silencing offers a more robust consideration of the quality of employer-employee interaction than only "voice" : voice concerns itself with the provisions and mechanisms for, as well as the calibre of, the upward provision of employee views. However, employer silencing also looks at what employers do not themselves provide or voice via information disclosure. Failures by employers to disclose information, not typically considered as a strictly voice dynamic, will nonetheless impact the capacity for informed dialogue to occur and hold significant ramifications for informed voice. There is consequently much utility in advancing analysis of organisational silence in workplace studies, by incorporating employer silencing as the act of withholding information and curtailing workplace dialogue. However, as Donaghey et al. (2011) and Barry et al. (2018) have noted, silence, like voice, is contextually and situationally specific; it can exist on some issues and not others in one organisation. For example, an employer may be willing to discuss pay and provide a voice mechanism for such purposes but not entertain dialogue with the workforce on introducing new workplace technology. Although silence is likely to occur where voice structures are denied to workers altogether, it can still occur where voice mechanisms exist. Employers can dominate voice arrangements through wielding decision-making prerogatives and self-interested agenda-setting (Butler, 2005) as well as via indirect hegemonic power (Lukes, 2005). That employers might intentionally want silence rather than voice reflects that information disclosure and dialogue is bound-up with distribution and exercise of power inside firms. There are constraints on employer information disclosure due to the costs not only of information processing but also the loss of advantage vis-a-vis others. Kleiner and Bouillon (1988) suggest that there can be strong incentives for employers to withhold information because employees in possession of greater information can use it to extract a greater share of rents: The more information disclosed by employers, the higher the wages but the lower firm profitability and productivity. Dialogue can reduce speed of decision-making and thwart employer strategies insofar as it provides employees with opportunities to express voice.

| REGULATING FOR INFORMATION DISCLOSURE AND DIALOGUE: THE I&C DIRECTIVE
Although some employers have an incentive to circumvent dialogue and withhold information, preferences are not unconstrained. Employers face pressures to provide information and engage in workforce discussions via external regulatory compulsion (Eurofound, 2016;Rogers & Streeck, 2009). Thus, regulations can provide structures for dialogue, as in a recognised trade union, or can require employers to provide information in the context of a transfer of undertaking or redundancy consultation. A prominent example of regulatory requirement in this context is the European Information & Consultation Directive transposed across EU member states in the 2000s (Adam, Purcell, & Hall, 2016;Butler, Lavelle, Gunnigle, & O'Sullivan, 2018;Dobbins, Dundon, Cullinane, Hickland, & Donaghey, 2017;Hall, Hutchinson, Purcell, Terry, & Parker, 2013;Hall, Purcell, Terry, Hutchinson, & Parker, 2015). The Directive seeks to promote social dialogue via timely information and consultation through workforce-agreed structures. Under its Standard Rules, provision of information and consultation, with a view to reaching an agreement, is required on the recent and probable development of the undertaking's or the establishment's activities and economic situation; thus it applies to any envisaged or probable changes that are likely to lead to substantial alterations in work organisation or contractual relations within an undertaking or establishment.
Decentralized flexibility was allowed across member states in implementing the Directive. Due to preexisting legislation in most member states, many countries made little or no changes (Donaghey, Carley, Purcell, & Hall, 2013). The United Kingdom and Ireland were exceptions and were important country test cases given their weak traditions of information sharing and consultation relative to other member states (Dobbins et al., 2017;Dundon et al., 2014;Hall & Purcell, 2012). The two jurisdictions are the focus for our examination of regulatory impact on silence due to the Directive's significance in these contexts. Although there are differences in the detail, two broadly similar features of the implementing legislation in the United Kingdom and Ireland are noteworthy. First, in neither state is a positive obligation placed on business to establish employee representative bodies. Instead, employers only respond to and establish a body when 10% of workers sign a request for such a structure. Second, in both countries, consistent with their voluntarist traditions, provisions allow Pre-Existing Agreement (PEA) priority over the Standard Rules. Consequently, companies that recognise trade unions or a pre-existing non-union employee representation scheme can claim this as the mechanism for fulfilling the Directives' requirements, provided it is agreed with workforce representatives (although in Ireland a cut-off point for claiming a PEA before legislative enactment was required). Where PEAs do not exist, negotiations commence to agree to a structure; where negotiations fail within a specified timeframe (see Table 1), a default to the Standard Rules is applied. The intent, structure, processes, and content implied by the Directive, and written into the transposed laws in both jurisdictions, are summarised in Table 1.
Studies of the Directive in the United Kingdom and Ireland show limited impacts on information and consultation practices Dobbins et al., 2017;Hall et al., 2013Hall et al., , 2015. The work of Hall & Purcell, 2012) sought to distinguish between firms that actively responded to the regulations but used the regulations in diverse ways. Some firms became "active consulters," positively engaging with the regulations; others took an instrumental view and used it for downward communication. Other companies adhered to the regulations by implementing appropriate structures, but these fell into disuse over time and became defunct. The focus in this paper is different to that of existing work; however, it is not concerned with describing the way different employers used the regulations to construct consultation structures and disclose information, but rather with how the intended benchmark of the Directive could be circumvented by employers so that silences prevailed on items under regulatory remit. In effect, we show how the regulations became marginalised through a plethora of non-compliance routes constituting employer silencing of employee voice, enabling employers to dominate and occupy this regulatory space.
We propose three principal ways the Directive was diluted or negated along a spectrum of activity, ranging from simply hoping to remain unnoticed by regulatory agencies to engaging in the act of regulatory pretence. In effect, employers were able to withhold information and circumvent dialogue despite regulatory requirements intending to establish their precise opposite. The categories of regulatory non-compliance we propose are avoidance, suppression, and neglect. By avoidance, we mean employers can elude the regulations to ensure the type of information sharing and dialogue it stipulates does not take place. In undertaking avoidance, the employer will hope to go unobserved or under the radar to avoid regulatory oversight (see, for example, Bloor & Sampson, 2009;Jaehrling & Mehaut, 2012). However, in those cases where employers can no longer go unnoticed from regulatory requirements, recalcitrant employers may counter-mobilise by seeking to suppress attempted compliance via obstruction and flouting (see for example Gall, 2010).
In contrast, other employers, instead of avoidance and suppression, may engage in the act of neglect (Farrell, 1983;Kolarska & Aldrich, 1980). Apathy and passive disengagement exemplify the route of neglect as employers' T A B L E 1 Benchmark for information and consultation as set by Directive and National Regulations for United Kingdom and Ireland Intent • For employer to work in a spirit of co-operation with workforce representatives and with due regard for their reciprocal rights and obligations in providing agreed structures for information and consultation. commitment to the spirit and practice of the regulations is low and not in accord with principles of acting in good faith. However, employers in this scenario are likely to be caught in a half-way house between wanting to ignore the regulations entirely and needing to maintain some pretence of compliance, if only to avoid punitive penalisation via regulatory oversight. A minimalist level of acknowledgement is thus likely to exist, but in substantive terms, regulatory compliance is marginalised and abandoned over time.

| CASES AND METHODS
The article uses illustrative cases from a study of employer responses to the information and consultation regulations in the United Kingdom and Ireland to demonstrate how avoidance, suppression, and neglect could preserve employer silencing of employee voice despite regulatory requirements to do otherwise. Specifically, the paper presents three cases on a post facto basis that exhibit how regulatory circumvention operationalised in response to the I&C Directive. Fieldwork for all cases derives from the period between 2008 and 2014.

| The cases
The first case is an American-owned non-union multinational company operating in the contract catering sector in the Republic and Northern Ireland. This organisation, called CateringCo., employed 4,000 employees in over 400 sites. The majority of workers are catering assistants and paid minimum wage rates on a mix of full-time, part-time, temporary, and zero-hours contracts. The HRM function was centralised at a head office in Dublin, with auxiliary support in three regional offices. Site managers at local level tended to deal with HRM issues as they arose. Although the company was non-union, one of the larger general unions attempted to organise workers at some of its sites. Although the employer did not grant recognition, unofficially recognised union representatives in a tiny minority of sites actively dealt with individual grievances. Collective consultation was non-existent, and information was predominantly downward communication from management. The case illustrates regulatory avoidance.
The second case is from the medical device manufacturing sector in the Republic of Ireland. ManufactureCo.
is an American-owned non-union multinational company employing 3,000 employees at the facility researched.
Eighty percent of the staff were semi-skilled assembly line operatives, with the remainder in back-office support functions. This employer pursued many elements of "best practice" HRM, such as selective hiring, teamwork,

| Fieldwork methods
The fieldwork deployed two research instruments: semi-structured interviews and documentary content analysis.
Interviews were with employer representatives (mainly, the human resource director, senior, and line management), employees, and where existing, union or non-union representatives. Interviews focused on information and consultation practices in use, why they were in use, awareness and assessment of the regulations, actions taken in response to the regulations, if any, and evaluation of information and consultation practices in situ at the company/site. Documentation such as management circulars, minutes of meetings, and constitutions of information and consultation forum provided evidence on the structure, process, and content of workplace dialogue. Table 2 summarises the evidence sources. At CateringCo., information derives from interviews with 16 senior managers, including the human resource director, human resource support staff at the national level, 3 regional HR managers, and 11 site managers. Across four of the locations, the research team interviewed 27 catering assistants as well as 3 union representatives (2 site stewards and 1 sectoral officer with responsibility for the company). Documentation on internal HRM policy, procedure, and examples of staff communication was also acquired.
At ManufactureCo., a dispute over employee attempts to trigger the regulations via the 10% request generated contact with several employee representatives, who provided a sample of internal documentation on the company's response to the regulations. The documentation included up to 50 documents outlining the dispute over the regulations, the employer's response to statutory bodies, and information regarding the workforce and employees initiating the trigger. Some of this documentation was in the form of emails, postal correspondence, internal circulars, and intranet documentation. The documents form the basis for the material presented in the case alongside interviews with seven employees active in the dispute and who represented the trigger campaign.
In ConcreteCo., the fieldwork incorporated 20 interviews. The schedule of interviews included eight management representatives, including the national human resource director, one regional human resource manager and six site managers. Four employee representatives, three union and one non-union, were also interviewed, whereas in one site, eight direct labourers, selected by management to represent employee categories on-site, were interviewed. The fieldwork also accessed documentation on company HRM policies and internal joint consultative committee minutes.

| Analysis
Data analysis followed protocols appropriate to case study design, suggested by Eisenhardt (1989) and Yin (2018).
Conceptual categories became subject to a coding protocol of identifying cases consistent with the categories of regulatory circumvention categories. The coding was a refining process that underwent several rounds of review.
The research checked and cross-referenced the coding of both interview reports and documentary results after the fieldwork to ensure validity, accuracy, and consistency of interpretation and analysis. Cases cover prior background information on information and consultation before the Directive; the response of employers to regulations and their intentions for information and consultation; and the resulting structures, processes, and content of information and consultation.

| CASE FINDINGS
Earlier, we outlined the Directive's regulatory benchmarks for the meaningful delivery of information sharing and dialogue in the workplace. Such benchmarks sought to foster behavioural patterns on intent, structures, processes, and content. This section outlines how employers could circumvent these benchmarks via our typology of avoidance, suppression, and neglect, amounting to forms of employer silencing. Table 3 provides a comparative summary of regulatory objectives and actual results in each firm per the type of regulatory non-compliance the employer pursued.

T A B L E 3 Benchmarking employer information provision and workforce dialogue against regulatory requirement
Regulatory requirement (see also Our typology of non-conformance indicated that avoidance results in silence where an employer has successfully dodged the regulations to ensure the type of information sharing and dialogue it stipulates never occurs. Employer non-compliance goes unnoticed by all other interested parties, enabling the preservation of structures, processes, and associated content that are out of sync with regulatory expectations. In our fieldwork, employer representatives at CateringCo. were aware of the regulations in the period before transposition and enactment but opted not to alter practices. The senior human resource (HR) manager for operations in the Republic claimed the company had been "nervous" about the intent of the regulations when first noted in the early 2000s. CateringCo. intentions towards information and consultation and their own preferred practice in this context were confined to an exclusively management determined and controlled structure of direct downward communication with staff on content relating to individual performance, operational schedules, and working time rotas. The employer accepted that there was no agreed structure on information and consultation, so they did not claim to have a PEA. HRM respondents maintained that they were unclear how such an agreement would occur given the geographical spread of site locations and the relatively small numbers of employees at each site; logistical issues of coordinating an agreement were claimed as problematic. Thus, the process by which downward communication occurred was via a mixture of individual one-on-ones, team briefings at the start of a shift, site notice boards, and formal appraisals. There was no structure or process in place for collective employee voice, whereas avenues for individual employees to raise matters varied greatly at site level, being contingent on individual management receptiveness. In a few sites, unofficially recognised union representatives existed, participating in grievance or disciplinary hearings if a member so requested. However, any management-union meetings occurred off-site, and shop stewards could not circulate any union material on site.
Local management did not inform American headquarters of potential developments associated with the regulations. Local management, as the HR Director put it, did not want to "alarm" American headquarters, which might lead to more significant intervention from abroad, something management at national-level wanted to avoid. The employer's intent towards the Directive was to take a "wait and see" approach. Ultimately, the HR Director described the final form of the transposition process as "allaying fears" among CateringCo. management about maintaining the status quo: No change would arise unless requested, enabling the above status quo in structure, process, and content to prevail. The employee population was assumed by management to be uninterested in the regulatory rights and employer representatives at the site level reported confidence in proceeding with their own employer determined structures, processes, and associated content unless otherwise challenged. Absence of any meaningful regulatory oversight meant that avoidance was a low-risk option for the employer.
Despite the claim that local-level interest among employees in the regulations would be low, the HR Director reported concern that unions might use the regulations in the Republic of Ireland to establish a foothold at site-level by organising the workforce to trigger the regulations, introducing a more formal collective structure to strengthen union presence in the company. At the time of research, a general union was organising in the catering and hospitality sector and recruited members in the company. Yet the employers' concern proved unfounded; field research at various sites observed that aside from only one shop steward, union activists were unaware of the regulations and the full-time officer who played an adjunct role assisting CateringCo. stewards displayed no interest in the Directive, leaving the employer free to avoid regulatory purview. Nonetheless, the employer remained wary of the union using the regulations and an indicator of that intent was their vetting of the research interview schedules to ensure fieldwork avoided raising the issue of regulatory rights with staff. Furthermore, the national headquarters changed the interview schedules to remove any reference to "consultation," which the research conceded for access. Representatives from the national HR office explained this action because they did not wish to create "unmanageable expectations" among employees.

| Suppression at ManufactureCo.
Whereas the actions by CateringCo. evidences an example of an employer who went unnoticed, other employers' non-compliance can be exposed and subjected to requests for change in the direction of the regu- followed by a revised version of the constitution removing the date altogether, and the process and content of the body were rephrased. The Forum was first presented as a body "to facilitate two-way communication," later revised to "facilitate communication of information and consultation." It was not agreed by employee representatives nor was there evidence on how employees approved it. The employer ignored a written request by the Court to have the PEA voluntarily verified, as the latter did not have powers to compel evidence under the legislation. The employer later confirmed in documentation to the activists that it "could not find" a signed, dated copy of the alleged PEA but that this did not "disprove" it did not have one. The employer ignored requests to initiate negotiations on an information and consultation structure in line with regulatory requirements and presented its preferred Forum as a fait accompli. Appointment, rather than election, of representatives transpired; the majority of whom were staff operating in support functions rather than shop floor operatives. Thus, employer action in this case was characterised first by attempted obfuscation to undermine the legitimate trigger request before turning to a reliance on management fiat to steamroll a preferred course of non-cooperation with the trigger campaign and avoidance of negotiation to introduce an I&C structure under regulatory auspices.
The employer appointed two of the employees involved in requesting compliance to sit on the Forum. Their participation, however, proved disputatious: They obstructed forum meetings from proceeding smoothly because the body was not compliant with the I&C regulations. Employee representatives contended that the period of 3 months for negotiating had elapsed and the Standard Rules should now apply. The employer responded by suspending the Forum, reporting its intentions to hold a plant-wide referendum to settle the matter on whether the company's preferred Forum should remain or the Standard Rules formula should apply. The employer circulated, before the referendum, an information sheet outlining a series of questions and answers on the purpose of the referendum. The same sheet also detailed apparent differences between the I&C structures of the Information Forum and the Standard Rules and their intended content. Differences were minor, although the definition of consultation processes was stronger in the Standard Rules, emphasising "consultation with a view to reaching agreement," a prescription absent in the Information Forum's constitution. The referendum process, however, was reported by activists to be highly tainted by employer threats that if the Standard Rules were adopted, the American owners' investment in the plant would suffer.
No formal declaration on/of the referendum results was ever provided by the employer, apart from digital notice boards on-site reporting that the majority of those who voted, voted for the Information Forum. An attempt by the activists to secure figures on voting distributions was unsuccessful, as was a request to the external auditor who oversaw the election. The only documentary evidence on voting distribution available is a report by the company in later correspondence with an employment tribunal on a separate matter relating to an unfair dismissal claim by one of the employees behind the request for I&C rights. The correspondence claimed 66% of staff voted for the Forum over the Standard Rules. Within 12 months of the initial request, the company dismissed two employees linked with the trigger campaign for alleged work absences, although they later claimed victimisation by the employer (securing an out of tribunal settlement in both cases). Their exit from the organisation ended the campaign for regulatory compliance.

| Neglect at ConcreteCo.
Avoidance and suppression represent, in their respective ways, active forms of regulatory non-compliance.
However, as outlined earlier, employers can adopt a more passive form of non-compliance characterised by minimalist acknowledgement of the regulatory requirement and but not conforming with the regulatory spirit.
ConcreteCo. constituted a case of regulatory neglect. Employer representatives in both Northern Ireland and the Republic reported awareness of the Directive but were sceptical of its value or expressed disinterest, particularly when it became clear that alterations to existing I&C practices were voluntary or required initiatives from below. When asked if any action was taken in response to the regulations to ascertain employer intent, employer representatives maintained the company did not need to undertake new actions. The employer's view was that ConcreteCo. had in place appropriate structures and processes at a national and local level that would suffice as "pre-existing agreements": We wouldn't have gone through a process of having to put in place arrangements to comply with the Information and Consultation Directive. It hasn't had a major effect on us here. We were aware of it, but felt we didn't need to do anything with it. That legislation was badly worded and it didn't really give clear guidance. We were already doing most of it anyway.
(HR Director, Northern Ireland) The employer representatives acknowledged that, apart from health and safety, consultation was not, as assessed by the HR Director in Northern Ireland, necessarily "well developed" in the company and regarded as an "overhead"; employer communication tended to dominate or, for the unionised sites, centralise pay bargaining. In explaining intent and existing structures, employer representatives identified that each site was autonomous on information sharing and workplace dialogue practices, adopting their own structures for information disclosure and dialogue. Such practices typically derived from site-level inherited traditions and the intent of the local-level management as to whether information share and dialogue were "a good thing or not" (HR Director, Northern Ireland).
The structures were thus unilaterally determined by management, were not formally agreed with workforce representatives, and were ultimately dependent on the vagaries of different local manager's intent as to how often the structure met and whether it existed at all. Some sites evidenced "works committees," typically forums for addressing content on production and operational matters like the site-level layout of stock and materials. Other sites had health and safety committees, which tended towards formalised reviews and discussions on the content of risk assessments, associated regulatory requirements, and the reporting of defects on site by either management or the health and safety representative. In terms of composition, some committees had an employee representative-nearly always an appointed health and safety officer-and others did not, whereas some were entirely composed of management or senior engineers and supervisors.
Although health and safety representatives were available on nearly all sites, content involving information disclosure, or dialogue on content-like financial circumstances, employment developments and contractual matters was non-existent in the non-union sites, whereas work organisation was discussed only in considering changing operational or layout requirements for new production. Yet, where this occurred, the process was reported by both employer and employee representatives as a process of downward communication rather than collective consultation. Even in the unionised sites, process characteristic of information disclosure and dialogue on content outside of pay was negligible. Despite relatively high density, site-level unionism was confined to individual grievance and disciplinary issues. Pay negotiations were centralised at the national level with bi-annual wage agreements balloted on by members. Trade unionists, however, reported that consultation was absent on non-pay issues like organisational change and new technology. Despite this, unions showed no inclination in utilising the I&C regulations to effect changes in practice: An attitude driven by a combination of union disinterest in (and lack of awareness of) the regulations and a view that initiating new patterns of consultation would likely disrupt settled collective bargaining relationships with the employer.

| DISCUSSION AND CONCLUSION
Existing organisational silence-voice literature (Barry et al., 2018;Brinsfield, 2014;Morrison, 2011) largely confines itself to focusing on how employees withhold information and refrain from speaking up, arguably resulting in the mistaken perception that organisational silence is employee-led. As a corrective, the novel conceptual insight in this article is to extend the silence concept to reveal how employers withhold information and curtail workplace dialogue and opportunities for employees to have a say. It is the first research to present the concept of employer silencing.
Significantly, this is analytically distinct from extrapolating the presence or absence of employee voice policies and practice and also differs from workers remaining silent in response to management. The agency and process of employer silencing of employee voice produce weaker voice outcomes than may be intended by state regulatory requirements, facilitating employer dominance and occupation of regulatory space (Hancher & Moran, 1989).
The article considered how employers responded to regulatory requirements intended to encourage dialogue and employee voice at work, by pursuing the opposite outcome by silencing workers' voice opportunities. Specifically examining the European Information & Consultation Directive and national regulations in the United Kingdom and Ireland, we proposed three routes to employer silencing: avoidance, suppression, and neglect. These were then evidenced in practice to three matched cases of employer responses. The intent of the Directive was to encourage information disclosure and consultation in a spirit of cooperation to be realised via agreed structures, processes, and content agreed with workforce representatives (Butler et al., 2018;Dobbins et al., 2017;Hall et al., 2013). Instead of assessing how the regulations were used, as existing research has already examined, the evidence in this article considered how the regulatory space for I&C was circumvented by employers in three case studies. Although each firm adopted different means, none of the three cases could be said to align to key elements of the regulatory requirements, although some cases do so more than others.
Rather, each case perpetuated different traditions and practices characterised by weak to non-existent information sharing on employment relations matters and avoidance or minimisation of consultation. Despite the aspirational ambit of the Directive to encourage information and consultation on substantive items, employers could avoid or limit this from occurring in the cases studied, partly through weak and minimalist regulatory design at national level and via the different forms of non-compliance adopted at organisational level. The organisational silence that traditionally existed in these firms could be largely preserved or moderated only slightly, whereas structures for information disclosure and consultative dialogue could be avoided or neutered. A crucial influence in driving the pattern of regulatory non-conformance at firm level is the employers' intent. Negative intentions towards the Directive provide the space in which non-conforming structures, processes, and associated content prevail. It is not so much that a particular intent of avoidance, suppression, or neglect dictate a specific pattern of structure, process, and content, which themselves are a product of employer preference and organisational custom, but rather that the intent of nonconformance guarantees the aspirations of the Directive will fall short. Silence on issues specified by the legislation thus go unobserved. The employer at CateringCo. saw the Directive as damaging the preferred status quo of management controlled downward communication and was therefore keen to evade its infringement into the workplace.
Opting for avoidance was viable given the light-touch monitoring by regulatory agencies and improbability of an employee trigger in this context. Avoidance ensured that existing employer preferences for limited information share and non-consultation remained. 'At ManufactureCo. the Directive became a live issue at firm level when employees attempted to utilise the trigger mechanism to establish an independently elected employee forum as set out in the I&C regulations.' The employer, seemingly wedded to unilateral prerogative in designing internal information share and dialogue structures, opted for suppression of regulatory procedure initiated by the employee trigger. The intention to suppress the possibility of a joint regulatory space on information and consultation resulted in the employer steamrolling through procedural requirements to negotiate. Even though nominally the firm ended up with an I&C structure that mimicked much of the Directive, it was principally a creature of the employer rather than a jointly negotiated one. Given the context of suppressionary intent in which the structure was born, the likelihood that it will replicate the Directive's spirit of cooperation and good faith dialogue seemed remote. Employer awareness at ConcreteCo. that the Directive had limited regulatory teeth alongside weak prioritisation of workforce consultation encouraged a lackadaisical approach. Characterising this approach were claims that existing practice was sufficient, enabling employers to neglect the Directive and any readjustment of their existing behaviour. The result of such complacency and disinterest was the continued provision of ad hoc structures, processes, and content driven by local managers perceived needs. These structures tended to be underwhelming relative to what the Directive championed. The employer neither felt the need to avoid the Directive's purview actively nor faced any requirement for suppression. In the face of the relatively non-intrusive regulations, casual regulatory neglect could persist.
Thus, in none of the companies could the benchmark set out by the Directive be found: Organisational silence persisted in the realm of structures and processes for social dialogue, whereas the types of content aspired to by the Directive for dialogue went undiscussed, or as in the case of Manufacture Co., were raised under conditions of veiled threats about job loss. The intent of the Directive was not achieved because employers were not committed to the regulatory spirit of meaningful social dialogue and opted for avoidance, suppression, and neglect.
The implications of employers perpetuating the withholding of information and neutering social dialogue are workplaces with a significant democratic voice deficit (Hyman, 2016;Timming, 2015). Several important implications arise, which could inform both future research and policy. In terms of research, the responses of avoidance, suppression, and neglect could be operationalised further to determine why one is adopted over the other, that is, why do some employers calculate at least the pretence of regulatory compliance and others simply avoid it? Is the adopted form contingent on the type of regulation? Or could different combinations of response be simultaneously utilised in the same situation? Our study of the I&C Directive and transposed national regulations suggests prevailing workplace traditions, and employer views on the legitimacy of disclosure and dialogue are important influencers of subsequent responses. Employer responses of avoidance, suppression, and neglect might be further considered vis-à-vis more robust (or harder) regulatory forms. Other regulations may be less amenable to local-level creativity in how conformance is secured and may be more stringently monitored and enforced. The I&C Directive, via national-level transposition in two liberal market economies, provided ample scope for employer innovation at local level; other regulations might be less permissive. Would some regulations minimise the capacity for avoidance or become prone to neglect?
This links to policy, because if non-compliance can occur in the forms outlined here, it is necessary to ask what would make regulatory conformance more likely? As neglect and suppression imply, conformance is no easy matter of simple verification and suggests close local-level observation, and monitoring are necessary. Yet this may be beyond the resources of most regulatory authorities. Local actors like workers or unions may not have the inclination or capability to carry this burden. For example, employers, in opting for avoidance, are operating on the premise that there is a lack of countervailing forces at the firm level to constrain such behaviour. Where suppression is favoured, employers are likely to calculate that the balance of forces locally favour their ability to circumvent potential threats to the status quo. The act of neglect would appear to work on the assumption that there is little appetite or momentum to alter existing arrangements. As regulatory actors, unions could conceivably attempt to alter such patterns.
Where unions find avoidance, they could expose the prevailing regime of silence by highlighting alternative and legally backed possibilities. Faced with suppression, unions could counter-mobilise members at firm-level around the incursion on employee rights or call upon the aid of state powers to bolster their position locally. Confronted with neglect, unions could direct energies towards revamping deficient bodies by encouraging local activism or making a case to management around potential mutual gains or the risks of their transgression. However, such responses to regulatory non-conformance types assume unions have an interest in policing the prevalence of silence on these matters. In our cases, unions were mostly disinterested in regulatory compliance on this issue. It is already documented that unions can potentially see I&C Forums associated with the Directive as a potential competitor and that their commitments lay towards collective bargaining rather than information disclosure and non-bargaining forms of workplace dialogue (Dundon et al., 2014;Hall et al., 2015). Indeed, in the neglect case, unions were happy to allow this situation to exist for fear of disrupting exist bargaining arrangements on pay, whereas in the avoidance case, the union evidenced disinterest in the regulations as a whole (which seems to be why local activists remained ignorant of the legislation). As indicated by Bales, Bogg, and Novitz (2018), other regulations may prove more successful in securing employer compliance; where there are considerable transgression costs, workplace actors are informed of their relevant rights and responsibilities and the assessment of compliance is more transparent and identifiable.
Employers themselves might be mindful that although non-compliance can be expedient, the risk of exposure if regulatory requirements are breached might not only be subject to transgression costs and fines but potentially negative employment and public relations consequences.
To conclude, this article has advanced novel theoretical insights on employer silencing of voice.
Organisational silence is much more than simply workers choosing whether or not to speak out but is driven by employer preferences regarding information sharing and workplace dialogue. Therefore, employer silencing provides an important new research avenue, which other employment relations and HRM scholars could also explore in different empirical contexts. The article also highlights that employer silencing and regulatory noncompliance are an important public policy matter for the state and its agencies, as well as for HR and employment relations practitioners.