Open Doors: Union Access Rights, Employee Voice, and How Leaders Can Build Constructive Labor Relations
Download the article here:
Listen to a review of this article:
Abstract: From 30 October 2026, trade unions in Great Britain gain a statutory right to access workplaces, physically and digitally, to meet, recruit, organize and support collective bargaining. The change, introduced by the Employment Rights Act 2025, arrives alongside simplified recognition rules, new protections against detriment for industrial action and stronger rights for union representatives. This article explains what the access regime provides, where details still depend on codes of practice and secondary legislation, and why the shift matters for leaders well beyond the United Kingdom. Drawing on research about the voice and monopoly faces of unions, meta-analytic evidence on productivity, studies of employee voice and silence, and labor-management partnership, it argues that organizations face a strategic choice between defensive compliance and the deliberate design of voice systems. Five evidence-based responses are examined through organizational examples from technology, health care, logistics, automotive manufacturing, retail and platform work, followed by guidance on building long-term relational capability.
For much of the past four decades, labor relations strategy in many Anglo-American firms has been organized around a quiet assumption: that the best union relationship is no union relationship. Executives hired consultants to keep organizing drives out, trained managers to spot early signs of collective activity, and treated engagement surveys and open-door policies as substitutes for independent representation. In Great Britain, that assumption is about to meet a new legal reality. From 30 October 2026, independent trade unions acquire a statutory right to request access to workplaces, both on site and through digital channels, for the purposes of meeting, supporting, representing, recruiting and organizing workers and facilitating collective bargaining (Department for Business and Trade, 2026c; Faragher, 2026). Employers who cannot agree terms with a union may find the Central Arbitration Committee (CAC), the independent tribunal that already handles statutory recognition, deciding the terms for them.
The access right is one piece of a broader package in the Employment Rights Act 2025, which received Royal Assent on 18 December 2025 (Employment Rights Act 2025). Other trade union measures arriving on the same October date include a revised code of practice on recognition, reasonable paid time off and facilities for union equality representatives, and a prohibition on employers subjecting workers to detriment designed to deter or penalize protected industrial action (Acas, 2026; Department for Business and Trade, 2026c). Earlier in 2026 the statutory recognition process was simplified, removing the requirement that 40% of the whole bargaining unit vote in favor (Peyton & Penman, 2026). A duty to tell workers in writing about their right to join a union follows from 1 January 2027 (Department for Business and Trade, 2026c).
Leaders outside the United Kingdom may be tempted to treat this as a local compliance matter. That would be a mistake. Multinationals with British operations will need access protocols by the end of October, and the underlying questions travel well. How should an organization respond when workers seek a collective voice? What is the business case, and what are the risks, of moving from avoidance toward engagement? And how can leaders build voice systems that workers trust, whether or not a union is present?
Public sentiment is also shifting. In the United States, Gallup's August 2026 survey found 71% of adults approving of labor unions and a record 47% wanting unions to have more influence, the highest share in a quarter century (Saad, 2026). Yet U.S. union membership stood at 10.0% of wage and salary workers in 2025, half its 1983 level (U.S. Bureau of Labor Statistics, 2026). The gap between approval and membership is one reason high-profile organizing contests at Amazon, Starbucks, Volkswagen and Mercedes-Benz have drawn so much attention.
This article sets out what the new British regime requires, summarizes the research on what unions and collective voice actually do to organizations and individuals, and then turns to five evidence-based responses for leaders. The goal is practical. The legal detail is summarized carefully but is not legal advice, and several operational points will depend on the final Code of Practice and regulations. The larger argument is strategic: an access right opens a door, and leaders choose whether that door leads to a courtroom or to a better-run organization.
The Union Access and Employee Voice Landscape
Defining Key Terms
Employee voice is a broad term for the ways employees express views, concerns and ideas about work and influence decisions that affect them. In organizational behavior, Morrison (2011) defined voice behavior as discretionary communication of ideas, suggestions, concerns or opinions intended to improve organizational or unit functioning. Her later review emphasized that voice is constructive yet challenges the status quo, and that its opposite, employee silence, is the withholding of such input rather than simply having nothing to say (Morrison, 2014). Industrial relations scholars use the term more structurally. Dundon et al. (2004), studying eighteen organizations in England, Scotland and Ireland, distinguished voice as individual dissatisfaction, as collective organization, as contribution to management decision-making, and as mutuality. They found that managers mostly described voice in efficiency terms and played down rights, and that how deeply voice practices were embedded mattered more than how many schemes existed.
Collective voice refers to representation through an independent body, usually a trade union, that speaks for a group. Hirschman (1970) provided the classic framing: when dissatisfied, members of an organization can exit (leave) or exercise voice (try to change things from within), and loyalty shapes which they choose. Freeman and Medoff (1984) applied this lens to unions, arguing that unions have two faces. The monopoly face raises wages above competitive levels and can impose costs. The collective voice/institutional response face gives workers a safe channel to raise problems, reduces quits, and can push management toward better practices.
A right of access in the new British sense is a statutory entitlement for an independent trade union to request entry to a workplace, and to communicate with workers digitally, for defined purposes. Under the framework described in government guidance, access can be physical, digital, or both, with digital access meaning access through electronic means (Department for Business and Trade, 2026c). The right does not extend to organizing industrial action (Peyton & Penman, 2026; Rowlands et al., 2026). Statutory recognition is the separate CAC-run process by which a union can compel an employer to bargain collectively over pay, hours and holidays for a defined group. A neutrality agreement, more common in the United States, is a voluntary pledge by an employer not to campaign against unionization and to follow agreed rules during organizing.
Finally, labor-management partnership describes arrangements in which an employer and its unions commit to joint problem-solving on issues beyond the contract, such as quality, safety and change. In much of continental Europe, a parallel channel exists through works councils, elected employee bodies with legal information, consultation and in some cases co-determination rights, and through board-level employee representation.
Prevalence, Drivers, and Distribution
Union membership has declined across most advanced economies, but the decline is uneven. In the United Kingdom, 22.4% of employees were union members in 2025, up slightly from 22.0% in 2024 but well below the 32.4% recorded in 1995 (Department for Business and Trade, 2026a). Membership is heavily concentrated in the public sector, where density was 48.5% in 2025, against 12.1% in the private sector. Roughly 6.6 million employees were members. Collective agreements reach further than membership: according to the Annual Survey of Hours and Earnings, about 39.9% of jobs had pay set by a collective agreement in 2025, though only 19.5% in the private sector (Department for Business and Trade, 2026a). Younger workers are much less likely to be members, with density of 8.7% among employees aged 16 to 24 compared with 26.6% among those aged 50 and over (Department for Business and Trade, 2026a).
U.S. figures are lower still. The Bureau of Labor Statistics reported 14.7 million union members in 2025, a membership rate of 10.0%, compared with 20.1% in 1983 (U.S. Bureau of Labor Statistics, 2026). Public sector workers had a rate of 32.9%, private sector workers 5.9%. Median weekly earnings for full-time union members were $1,404, against $1,174 for nonunion workers, a raw gap that reflects differences in occupation and industry as well as any union effect. The 2025 annual estimates rest on 11 months of data because the October survey was not collected during the federal government shutdown, so they are not strictly comparable with earlier years (U.S. Bureau of Labor Statistics, 2026).
Several drivers explain why the access question has become pressing now. The first is legal. The Employment Rights Act 2025 implements a deliberate policy of strengthening collective rights, and the access regime is its most operationally visible element for non-union employers. The process, as described in the government's consultation outcome and in practitioner summaries, runs roughly as follows. A union submits a request on a standard form. The employer has 15 working days to respond, then the parties have 25 working days to negotiate in good faith. If they cannot agree, the matter can go to the CAC, which decides whether access should be granted and on what terms (Bevan Brittan, 2026; Rowlands et al., 2026). The CAC can only impose an agreement where the employer has 21 or more workers (Department for Business and Trade, 2026c), and practitioner guidance notes that workers of associated employers in Great Britain are counted together (Peyton & Penman, 2026). Access agreements last for a maximum of two years. Government model terms contemplate access up to weekly, with minimum notice for visits, and employers must make existing facilities available but are not expected to make significant structural or technological changes (Bevan Brittan, 2026; Rowlands et al., 2026). Where an employer breaches an access agreement, the CAC may impose penalties of up to £75,000 for a first breach, £150,000 for a second and £500,000 for a third or subsequent breach (Peyton & Penman, 2026; Rowlands et al., 2026). Figure 3, later in the article, maps this sequence.
Several details remain provisional. The government consulted on the access framework between October and December 2025, receiving more than 1,600 responses, and published its response on 8 April 2026 (Department for Business and Trade, 2026d). It then consulted on a draft statutory Code of Practice, which drew more than 793 responses, and laid a revised draft before Parliament on 6 July 2026 together with secondary legislation, expected to come into force in October subject to approval by both Houses (Department for Business and Trade, 2026b). Leaders should therefore treat any summary, including this one, as a starting point and check the final instruments. The trade union measures apply in England, Scotland and Wales but not Northern Ireland (Department for Business and Trade, 2026c).
The second driver is the changing recognition landscape. Since April 2026, unions seeking statutory recognition no longer need to show that a majority of the bargaining unit is likely to support them, and a recognition ballot is won by a simple majority of those voting rather than also requiring 40% of the whole unit (Acas, 2026; Peyton & Penman, 2026). The 10% membership threshold for applications can be lowered by secondary legislation to as little as 2% (Rowlands et al., 2026). From 30 October, a revised recognition code applies unfair-practice rules from the point the CAC accepts an application, and suggests access meetings as frequently as once every five working days, with a minimum length of 45 minutes (Department for Business and Trade, 2026c). Taken together, the access right and the recognition changes lower the cost to unions of reaching workers and of converting support into bargaining rights.
The third driver is workforce expectation. Approval of unions in the United States has stayed between 67% and 71% for six years, compared with an all-time low of 48% in 2009 (Saad, 2026). High-profile campaigns at companies with strong consumer brands have shown that organizing is no longer confined to traditional industries. The fourth driver is the nature of work itself. Distributed, hybrid and platform-mediated work makes traditional "organizing at the gate" less relevant and digital access more so, which is why the British framework explicitly covers electronic communication.
Organizational and Individual Consequences
Organizational Performance Impacts
The evidence on what unions do to organizational performance is more nuanced than either advocates or opponents tend to admit. Freeman and Medoff (1984) argued that the collective voice face could raise productivity by reducing turnover, improving communication and prompting management to tighten practices, even as the monopoly face raised labor costs. Subsequent research tested this proposition many times with mixed results.
Doucouliagos and Laroche (2003) brought order to this literature through meta-analysis and meta-regression. They found that much of the variation in published estimates reflected differences in how studies were specified. Once those differences were accounted for, the association between unions and productivity was negative in the United Kingdom and positive in the United States, particularly in U.S. manufacturing. The lesson is not that unions are good or bad for productivity in general, but that the effect depends heavily on context, industry and, by implication, the quality of the relationship.
Hirsch (2008) offered a more cautious reading for the U.S. private sector. He argued that in competitive and dynamic markets, the combination of a union "tax" on earnings and slow responses to economic shocks has been associated with weaker performance in union companies, using automotive and airline industries as illustrations. His conclusion was not that worker institutions are doomed, but that they flourish when they add value and allow firms to perform at levels comparable with nonunion practice. For leaders, this frames the strategic question well: the aim is a relationship that makes the organization more adaptive, not less.
Evidence from co-determination points in a similar direction. Jäger et al. (2021) studied a German reform that ended mandatory one-third worker representation on the boards of stock corporations incorporated after August 1994. Comparing firms just before and after the cutoff, they found no effects of board-level co-determination on wages, wage structure, rent sharing or the labor share, and if anything shared governance was associated with greater capital formation. Fears that formal worker voice in governance automatically leads to rent extraction or underinvestment were not supported in that setting.
Organizational consequences also arise from the absence of voice. Willman et al. (2006) treated the choice of voice regime as an employer decision, distinguishing union, non-union, dual-channel and no-voice arrangements and asking which employers choose silence. The practical point is that "no voice" is itself a choice with costs: problems surface late, through exit or conflict, rather than early, through dialogue. Morrison (2014) summarized research showing that employees weigh whether speaking up will be effective and whether it is safe, and that silence often arises automatically when people sense that hierarchy makes voice risky. Where internal channels are weak, a statutory access right gives workers an external route into the conversation that leaders can no longer close.
Finally, the strategic costs of conflict are visible in recent examples. When roughly 75,000 workers represented by a coalition of unions at Kaiser Permanente staged a three-day strike in October 2023, it was described as the largest health care strike in U.S. history (Isidore & Delouya, 2023). Prolonged first-contract disputes, litigation and public campaigns impose reputational and management-time costs that rarely appear in labor-cost models.
Individual Wellbeing and Stakeholder Impacts
For individuals, voice matters because it affects fairness, control and the sense that problems can be fixed. Hirschman (1970) emphasized that voice is costly to exercise, and that people use it when they believe it can work. The research on employee voice behavior shows that perceived efficacy and safety shape whether people speak, and that voice also connects to identity and career motives (Morrison, 2011, 2014).
A recurring puzzle in the union literature is that union members often report lower job satisfaction than comparable nonmembers. Freeman and Medoff (1984) interpreted this partly as a voice effect: members are more willing to articulate dissatisfaction because they have a channel to do so. Bryson et al. (2004) tested the relationship using linked employer-employee data from the 1998 British Workplace Employee Relations Survey. They found a large raw satisfaction gap between union and non-union workers, but once they accounted for the fact that workers sort into unionized jobs for reasons that also affect satisfaction, the gap disappeared. Selection rather than a causal effect explained the relationship. For leaders, the implication is that complaints raised through collective channels are signals, not evidence that representation makes people unhappy.
Wage and security effects are the most tangible individual outcomes. The BLS earnings gap between union and nonunion workers cited earlier is a raw comparison, but first contracts in new bargaining units often include explicit pay floors, grievance procedures and limits on arbitrary dismissal. The tentative agreement reached by quality assurance workers at Microsoft-owned ZeniMax in 2025 included higher minimum salaries, grievance procedures, protections against arbitrary dismissal and commitments about advance notice of artificial intelligence uses that could affect workers (Kilkenny, 2025). These provisions address the uncertainty that drives much contemporary organizing.
Stakeholder impacts extend beyond employees. In health care, labor relations shape staffing, safety culture and continuity of care. In consumer-facing businesses, customers increasingly notice labor disputes. Gallup's data indicate that a majority of the U.S. public views unions favorably (Saad, 2026), which means aggressive avoidance campaigns can carry brand risk. For investors, persistent conflict signals governance problems; for regulators, it invites scrutiny. Conversely, stable partnership arrangements can become a source of legitimacy with commissioners, public funders and communities.
Evidence-Based Organizational Responses
The research and recent experience point toward five responses that help leaders move from avoidance toward constructive labor relations. They are not a sequence that every organization must follow, and several can be pursued whether or not a union is present. Table 1 summarizes the evidence, practices and illustrative examples for each.
Table 1: Evidence-based responses to union access rights and collective employee voice
Response | Evidence Base | Key Practices | Illustrative Examples |
Adopt a principled neutrality stance | Collective voice face of unions; costs of conflict | Board-level labor principles; no anti-union campaigning; fair, private choice for workers | Microsoft–CWA neutrality (2022); Volkswagen Chattanooga vs. Mercedes-Benz Alabama (2024) |
Build access-ready governance and protocols | Statutory framework; voice requires efficacy and safety | Named owners; 15-day response playbook; physical and digital access plans; manager briefing | UK access regime (2026); Amazon Coventry recognition ballot (2024) |
Bargain first contracts in good faith | Productivity effects depend on relationship quality | Early agreement on process; focus on scheduling, discipline, technology; avoid delay tactics | Apple Towson (2024); ZeniMax QA (2025); Starbucks (2021–2025) |
Institutionalize labor-management partnership | Partnership research; voice embeddedness | Joint governance bodies; unit-based teams; shared performance metrics | Kaiser Permanente Labor Management Partnership; NHS Staff Council |
Integrate collective and direct voice | Dual-channel voice research; co-determination evidence | Combine union, works council and direct channels; close the loop on feedback | Deliveroo–GMB agreement (2022); Volkswagen worker representation |
Adopt a Principled Neutrality Stance
The first response is a deliberate decision about posture. Most organizations do not consciously choose a point on that spectrum. Their position is set by habit, by the advice of whichever law firm they call when a request arrives, or by the instincts of local managers. A principled neutrality stance means deciding at the top that the organization will not campaign against workers' choice to organize, will ensure that choice is made fairly and privately, and will judge any union relationship by whether it helps the organization perform.
The evidence supports this posture on two grounds. First, Freeman and Medoff (1984) and the subsequent meta-analytic literature indicate that the productivity consequences of unions vary with context and relationship quality (Doucouliagos & Laroche, 2003), which means a hostile relationship forfeits the potential gains of the voice face while still incurring costs. Second, the statutory changes in Great Britain shift the economics of avoidance. Once unions can obtain regular access and recognition is decided by a simple majority of those voting, investment in resisting organization buys less and risks more.
Effective approaches:
Board-endorsed labor principles: Publish a short statement of how the organization will respond to organizing, approved at board or executive committee level, so that local managers are not left to improvise.
Neutral communication rules: Commit that managers will provide factual information but will not campaign, threaten or make promises tied to the outcome of an organizing effort.
Private, accessible choice: Ensure workers can decide about union membership without surveillance or pressure, including through digital means.
Fast, fair dispute resolution: Agree in advance how disagreements about conduct during organizing will be resolved, including expedited arbitration where appropriate.
Separation of roles: Keep legal compliance advice distinct from relationship strategy so that a defensive legal posture does not become the default business posture.
Microsoft offers the most cited corporate example of this approach. In June 2022, the company and the Communications Workers of America (CWA) announced a labor neutrality agreement covering Activision Blizzard employees, to take effect 60 days after the acquisition closed (Microsoft, 2022). Microsoft committed to take a neutral approach when employees expressed interest in joining a union, to allow employees to talk with coworkers and union representatives without business disruption, to provide a technology-supported process for choosing whether to join, to let employees keep their choice confidential, and to use an expedited arbitration process for unresolved disputes. Quality assurance workers at Microsoft-owned ZeniMax later became the first group to organize under the company's neutrality commitments, and after roughly two years of negotiation the parties reached a tentative first contract in May 2025 (Kilkenny, 2025).
Volkswagen and Mercedes-Benz illustrate how posture can shape outcomes within the same industry and region. In April 2024, workers at Volkswagen's Chattanooga plant voted 2,628 to 985 to join the United Auto Workers, with 73% of ballots in favor, after unsuccessful votes in 2014 and 2019 (Krisher & Megerian, 2024; National Public Radio, 2024a). Volkswagen said it respected employees' right to decide through a democratic process, and the plant had been the only one of the company's 120 sites worldwide without worker representation (National Public Radio, 2024a). A month later, workers at Mercedes-Benz near Tuscaloosa, Alabama, rejected the union 2,642 to 2,045. Workers described company videos warning about unions at the start of shifts and small-group meetings led by outside consultants, and the union filed unfair labor practice charges alleging intimidation and complaints in Germany under supply chain human rights law (National Public Radio, 2024b). The outcomes differ, and many factors shaped them, but the contrast shows how visibly an employer's conduct becomes part of the story.
A common pitfall is to declare neutrality at headquarters while local managers continue to behave as though the organization opposes unions. Workers judge posture by what supervisors say on the shop floor and in team chats. Neutrality without training, monitoring and accountability can be worse than no statement at all, because it invites accusations of hypocrisy.
Build Access-Ready Governance and Protocols
The second response is operational. Under the British regime, an employer that receives an access request has a fixed period to respond and a further fixed period to negotiate (Department for Business and Trade, 2026c). An organization that has not decided who owns the response, how sites will be mapped, and which grounds for limiting access might genuinely apply will spend those weeks scrambling. Access-readiness is not about finding ways to say no. Practitioner guidance stresses that refusals will generally be difficult and must be evidenced (Peyton & Penman, 2026). It is about responding competently, predictably and in good faith.
The behavioral research on voice adds a further reason for preparation. Employees assess whether speaking up is safe and whether it will make a difference (Morrison, 2014). When access visits are handled grudgingly, with managers hovering or meetings recorded, workers infer that voice is unsafe. Practitioner summaries of the government's approach indicate that employers should respect the privacy of access meetings and should not record them or send in management representatives (Rowlands et al., 2026). A well-run protocol signals the opposite of hostility.
Effective approaches:
Clear ownership: Name an accountable executive and a working group spanning HR, legal, facilities, security and IT, with authority to respond within the statutory window.
Headcount and site mapping: Confirm whether the organization and its associated employers exceed the 21-worker threshold in Great Britain, and map meeting spaces, shift patterns and safety rules for each site (Peyton & Penman, 2026).
Digital access design: Decide in advance how a union could communicate digitally, for example through an all-staff message or a dedicated channel, and how data protection obligations will be met (Peyton & Penman, 2026).
Front-line briefing: Train reception, security and site managers to recognize and escalate requests and to behave neutrally during visits.
Evidence discipline: Where there are genuine constraints, such as safety-critical areas or confidentiality, document them in advance rather than inventing them after a request arrives.
Relationship-first framing: Treat the first access negotiation as the opening of a relationship, not the first round of litigation.
Amazon's experience in Coventry shows why process and conduct matter. In July 2024, workers at the company's Coventry warehouse voted narrowly against statutory recognition of the GMB union, with about 50.5% voting no and the union falling short by fewer than 30 votes (FashionNetwork.com, 2024). The GMB's organizer accused Amazon of pressuring workers through hours of anti-union seminars, while Amazon said it valued engaging directly with employees and pointed to its investment in pay, benefits and skills (FashionNetwork.com, 2024). The ballot predated the new regime. Under the 2026 rules, a union in a similar position could seek regular access to the workforce from the point the CAC accepts an application, and unfair-practice rules would apply from that earlier point (Department for Business and Trade, 2026c). Employers who relied on intensive campaigning in the old system will need to reconsider both its legality and its wisdom under the new one.
A common pitfall is to hand the entire access process to external counsel and treat it as a dispute. Legal advice is essential on scope, thresholds and data protection, but if every interaction is conducted through solicitors' letters, the organization will have chosen a posture by default. The CAC will apply statutory principles, and penalties for breaching agreed terms escalate sharply with repetition (Peyton & Penman, 2026). An adversarial start makes such breaches more likely.
Bargain First Contracts in Good Faith
The third response concerns what happens after workers choose representation. In the United States, a union election victory is only the start: securing a first contract often takes years, and many units never reach one. In Great Britain, statutory recognition leads to a bargaining method that the CAC can impose if needed, but the quality of the relationship still depends on the parties. The research reviewed earlier suggests that the performance consequences of collective bargaining depend on its quality (Doucouliagos & Laroche, 2003; Hirsch, 2008). A first contract negotiated quickly and constructively can establish routines of joint problem-solving. One dragged out through delay and litigation teaches both sides to expect conflict.
Effective approaches:
Agree the process first: Before tabling substantive proposals, agree ground rules, schedules, information sharing and how sticking points will be escalated.
Prioritize what workers care about: First contracts in new units frequently focus on scheduling, transparent discipline, job security and the use of technology rather than headline pay alone.
Address technology explicitly: Commit to advance notice and discussion of automation and artificial intelligence that could affect jobs.
Avoid two-tier signals: Be cautious about giving improvements only to nonunion staff, which can be read as retaliation and can poison bargaining.
Invest in bargaining capability: Equip managers with interest-based negotiation skills and accurate business information.
Apple's first contract with retail workers illustrates a relatively contained, practical agreement. Workers at the company's Towson, Maryland, store voted to organize with the International Association of Machinists' Coalition of Organized Retail Employees in June 2022. In July 2024, the parties announced a tentative three-year agreement covering roughly 85 employees, with average raises of 10% over the life of the contract, higher starting pay for most job classifications, scheduling improvements intended to protect work-life balance, limits on contract workers, a severance clause, a more transparent disciplinary process, maintenance of existing benefits, and a commitment to bargain over future changes (The Register, 2024).
ZeniMax shows the same pattern in a technology setting. The tentative agreement reached in May 2025 for more than 300 quality assurance workers across sites in Maryland and Texas included wage increases, a credit policy recognizing testers' contributions to games, grievance procedures and protections against arbitrary dismissal. It also built in a 2023 union-studio understanding on artificial intelligence, requiring advance notice of AI uses that could affect members and negotiation over their impact (Kilkenny, 2025). The two-year timeline and a strike authorization vote shortly before the deal are reminders that even a neutrality framework does not guarantee speed.
Starbucks presents the contrasting case. After a Buffalo, New York, store became the first company-owned union shop in December 2021, organizing spread rapidly, and so did conflict, including accusations of retaliation and litigation between the company and Workers United (In These Times, 2024). In February 2024, the parties announced a framework for future negotiations intended to guide talks at nearly 400 union stores and resolve outstanding litigation (In These Times, 2024). Yet by November 2025, with about 550 of the company's roughly 10,000 U.S. company-operated stores unionized, no contract had been agreed and workers voted to strike (Associated Press, 2025). Nearly four years of conflict have consumed management attention and generated public criticism that a faster settlement might have avoided.
A common pitfall is to treat delay as a strategy. Delay may postpone cost increases, but it signals to employees that voice is ineffective, the condition under which research suggests silence and exit grow (Hirschman, 1970; Morrison, 2014). It also extends the period in which disputes play out publicly.
Institutionalize Labor-Management Partnership
The fourth response moves beyond the contract to joint governance of change. Labor-management partnership aims to channel the collective voice face of unions into improvements in quality, safety and efficiency, while giving workers real influence over how work is organized. Kochan et al. (2009), who tracked the partnership between Kaiser Permanente and the Coalition of Kaiser Permanente Unions for nearly a decade, described it as the largest and most complex such partnership in the United States, involving more than two dozen unions. Their account documents both innovative agreements and recurring tensions, a useful corrective to the idea that partnership eliminates conflict.
Partnership is consistent with the finding of Dundon et al. (2004) that the embeddedness of voice practices matters more than their number. A partnership that exists only as a steering committee will have little effect; one that reaches into everyday work through joint teams can change how problems are solved.
Effective approaches:
Joint governance bodies: Establish forums at enterprise and site level where management and union leaders share information and make decisions on defined issues.
Front-line joint teams: Create co-led teams that work on quality, safety and service improvement, so that partnership is felt at the point of work.
Shared metrics: Agree a small set of performance and workforce indicators that both sides track and discuss.
Explicit rules for disagreement: Accept that bargaining conflicts will continue and agree how partnership work will be protected during them.
Leadership continuity: Treat partnership as an enterprise strategy that survives changes of executive and union leadership.
Kaiser Permanente launched its Labor Management Partnership with the Coalition of Kaiser Permanente Unions in 1997, after decades of strikes and discontent (Labor Management Partnership, n.d.). In 2005, more than 3,500 unit-based teams, co-led by managers and front-line workers, began working to improve care, service and affordability, and the partnership has since been governed through successive national agreements with its union coalitions (Labor Management Partnership, n.d.). The partnership did not prevent the October 2023 strike by roughly 75,000 coalition-represented workers. A tentative agreement reached shortly afterwards included a 21% wage increase over four years and new minimum wages for health care workers (Isidore & Delouya, 2023). The lesson is that partnership is not a guarantee of peace, but it provides infrastructure for returning to joint work after conflict.
The NHS Staff Council provides a British public-sector example at national scale. It brings employer and trade union representatives together to maintain the Agenda for Change pay system, negotiate changes to core terms and conditions, and help local employers and unions interpret the national agreement (NHS Employers, n.d.). Its joint referral process requires local employers and unions to submit interpretation disputes together, and joint groups on health, safety and wellbeing and on equality, diversity and inclusion work in partnership (NHS Employers, n.d.). Private-sector leaders often overlook such models, yet they demonstrate how institutional design can make joint problem-solving routine.
A common pitfall is to launch partnership as a communications initiative rather than a governance change. If union representatives are invited to meetings but decisions are made elsewhere, workers will quickly see through it, and the partnership will lose credibility with both members and managers.
Integrate Collective and Direct Voice
The fifth response recognizes that most organizations will operate several voice channels at once: unions where workers choose them, non-union representative forums, works councils in some jurisdictions, and direct channels such as surveys, team meetings and suggestion systems. Research on voice regimes distinguishes union-only, non-union, dual-channel and no-voice arrangements (Willman et al., 2006). The aim is not to use direct voice as a substitute for independent representation, which workers and regulators increasingly see as union avoidance by other means, but to design a coherent system in which each channel does what it does best.
Co-determination research suggests that formal employee voice in governance need not undermine performance. Jäger et al. (2021) found no adverse effects on wages or investment from board-level representation in Germany. Morrison (2014) shows that direct voice depends on psychological safety and perceived efficacy, conditions that a credible collective channel can reinforce by protecting those who speak up.
Effective approaches:
Map existing channels: Inventory every route by which employees can raise concerns and influence decisions, and identify gaps, overlaps and populations left out.
Clarify roles: Distinguish what is negotiated collectively from what is consulted on and what is handled directly with teams.
Close the loop: Report back on what changed as a result of employee input, through both collective and direct channels.
Include hard-to-reach workers: Design digital and in-person channels that reach shift, remote, agency and platform workers.
Respect independence: Avoid creating management-controlled bodies that appear designed to pre-empt independent unions.
Deliveroo's 2022 agreement with the GMB shows both the potential and the controversy of integrating collective voice into a platform model. Announced on 12 May 2022, the voluntary recognition agreement covered more than 90,000 self-employed riders and gave the GMB consultation rights on benefits and on health, safety and wellbeing, collective bargaining rights over pay, and the right to represent member riders in individual disputes (GMB, 2022). The agreement recognized riders as self-employed, in line with several court rulings. The Independent Workers' Union of Great Britain, which had been pursuing statutory recognition through the courts, called it a public relations move and argued that it undermined riders' ability to enforce rights and to organize independently (Computer Weekly, 2022). The episode illustrates that collective voice arrangements are judged not only by their terms but by whether workers see the chosen partner as genuinely representative.
Volkswagen's global footprint offers a contrasting model. Before the 2024 vote, Chattanooga was the only one of the company's 120 plants worldwide without worker representation (National Public Radio, 2024a). In most of its operations, especially in Germany, collective bargaining coexists with works councils and board-level representation, the dual-channel arrangement that the co-determination evidence suggests can be compatible with investment and competitiveness (Jäger et al., 2021). For multinationals, the question is why voice arrangements accepted as normal in one jurisdiction are resisted in another.
A common pitfall is to use direct voice tools, such as pulse surveys or listening sessions, as a substitute for engaging with a union that workers have chosen. Such tools are valuable, but when they compete with independent representation they erode trust in both.
Building Long-Term Labor Relations Capability
The five responses address immediate choices. Sustaining constructive labor relations over time requires capabilities that many organizations allowed to atrophy during decades of declining membership. Three are especially important.
Leadership Mindset and Manager Capability
The most durable barrier to constructive labor relations is often managerial belief rather than law. Many managers absorbed the idea that a union presence signals leadership failure. A more useful frame, consistent with Freeman and Medoff (1984) and Hirschman (1970), is that collective voice is information: it tells leaders what workers value and where problems lie, often before those problems appear as turnover or safety incidents. Executive teams should discuss this frame explicitly, because front-line managers take their cues from what senior leaders say and reward.
Manager capability is the practical counterpart. Many supervisors have never worked alongside a union representative, conducted a grievance meeting or explained a business decision to a collective body. Development programs should cover the legal framework at a level appropriate to each role, interest-based problem-solving, the conduct expected during organizing and access visits, and how to respond to concerns raised through any channel without defensiveness. Research on voice suggests that managerial openness shapes whether employees speak (Morrison, 2011), and in unionized settings the same openness determines whether representatives bring problems early or escalate them. Organizations should also assess managers on the quality of their employee relations, not only on avoiding disputes, since a manager whose team never raises concerns may be presiding over silence rather than satisfaction.
Measurement, Transparency, and Trust
Constructive relationships rest on shared facts. Partnership arrangements such as Kaiser Permanente's work best when both sides look at the same data on quality, staffing and financial performance (Kochan et al., 2009). Organizations preparing for greater union access should consider which information they can share routinely, such as workforce composition, turnover, safety indicators and the results of employee surveys, and which requires confidentiality.
Leaders should also measure the health of their voice system directly. Useful indicators include the proportion of employees who believe they can raise concerns without negative consequences, the time taken to resolve grievances and collective issues, the share of employee suggestions acted upon, and the frequency of disputes that reach external bodies such as the CAC or employment tribunals. In Great Britain, the time limit for most employment tribunal claims is extending from three to six months for events from 1 October 2026 (Faragher, 2026), which increases the window in which unresolved grievances can turn into formal disputes. Tracking these indicators over time, and discussing them with employee representatives, turns voice from a slogan into a managed system.
Transparency extends to the organization's own conduct. Employers that publish their labor principles, report on how access requests were handled, and invite union and employee representatives to review those reports send a strong signal of good faith. Those that keep these matters entirely within legal and HR functions risk having their posture defined by the most adversarial incidents.
Governance, Strategy, and Cross-Border Consistency
Labor relations is increasingly a board-level issue. The combination of new statutory rights, reputational exposure and the cost of conflict means that boards should understand the organization's labor relations strategy, its risk profile and the indicators described above. Evidence from German co-determination suggests that formal employee voice in governance can coexist with investment and competitiveness (Jäger et al., 2021), and although board representation is not required in Britain, boards elsewhere can learn from the routines that such systems create for hearing the workforce perspective.
For multinationals, cross-border consistency is a strategic test. The Volkswagen and Mercedes-Benz cases illustrate how practices accepted in one country can sit awkwardly with conduct elsewhere, and how home-country supply chain and human rights rules can bring that conduct back into view (National Public Radio, 2024b). The British access regime will create similar comparisons, since a global employer may find unions requesting access in Great Britain while it resists organizing elsewhere. A coherent global labor principles statement, applied with sensitivity to local law, reduces this exposure and helps managers act consistently.
Finally, labor relations strategy should be integrated with workforce strategy more broadly. Decisions about automation, artificial intelligence, scheduling technology and the use of contractors increasingly shape the issues on which workers organize, as the ZeniMax and Apple agreements show (Kilkenny, 2025; The Register, 2024). Organizations that involve employee representatives early in such decisions can surface implementation risks sooner and build legitimacy for change. Those that announce changes and then face collective resistance will find the new British access rules make that resistance easier to organize.
Conclusion
The statutory right of access that takes effect in Great Britain on 30 October 2026 is, in practical terms, a set of timelines, thresholds and procedures. Unions can request physical and digital access; employers must respond and negotiate within fixed periods; the CAC can impose terms on employers with 21 or more workers; and repeated breaches can be penalized heavily (Department for Business and Trade, 2026c; Peyton & Penman, 2026). Together with simplified recognition, protection against detriment for industrial action and new rights for representatives, these changes alter the cost-benefit calculation that has long underpinned union avoidance.
The more important change, however, is strategic. The research does not support simple claims that unions always damage or always improve organizations. Effects on productivity depend on context and relationship quality (Doucouliagos & Laroche, 2003; Hirsch, 2008); apparent dissatisfaction among union members largely reflects selection (Bryson et al., 2004); and formal worker voice in governance has not produced the harms often predicted (Jäger et al., 2021). What matters is how leaders choose to engage. The organizational examples, from Microsoft and Apple to Kaiser Permanente, the NHS, Deliveroo and the contrasting outcomes at Volkswagen, Mercedes-Benz, Amazon and Starbucks, show that posture, process and persistence shape results.
Leaders inside and outside the United Kingdom can use this moment to make a deliberate choice: to treat collective voice as information and partnership as a capability, rather than to wait for the first access request and respond defensively.
Decide your posture deliberately: Agree at board or executive level how the organization will respond to organizing, and make neutrality real through training and accountability.
Prepare for access now: Name owners, map sites and digital channels, confirm thresholds and brief front-line staff before the first request arrives.
Bargain to build, not to delay: Treat first contracts as the foundation of a working relationship and address scheduling, discipline and technology directly.
Invest in partnership infrastructure: Create joint bodies and front-line teams with real decision rights and shared metrics.
Integrate voice channels: Use direct voice to complement, not compete with, independent representation, and close the loop on what changes.
Measure voice system health: Track whether employees feel safe and effective in speaking up, and review these data with representatives.
Stay current on the law: Check the final Code of Practice and regulations, since several operational details remain subject to parliamentary approval.
Research Infographic
References
Acas. (2026). Employment Rights Act 2025. Acas.
Associated Press. (2025, November 6). Starbucks' union workers plan strike next week unless company agrees to contract. ABC News (WPVI).
Bevan Brittan. (2026, April 20). Union knocking: What the government's access rights response means for employers. Bevan Brittan LLP.
Bryson, A., Cappellari, L., & Lucifora, C. (2004). Does union membership really reduce job satisfaction? British Journal of Industrial Relations, 42(3), 439–459.
Computer Weekly. (2022, May 19). Deliveroo accused of "soft union busting" with GMB deal. Computer Weekly.
Department for Business and Trade. (2026a). Trade union membership, UK, 1995 to 2025: Statistical bulletin. GOV.UK.
Department for Business and Trade. (2026b). Make Work Pay: Draft code of practice on trade union right of access [Consultation outcome]. GOV.UK.
Department for Business and Trade. (2026c). Trade union reform: New employment rights guidance for employers. business.gov.uk.
Department for Business and Trade. (2026d). Make Work Pay: Trade union right of access [Consultation outcome]. GOV.UK.
Doucouliagos, C., & Laroche, P. (2003). What do unions do to productivity? A meta-analysis. Industrial Relations, 42(4), 650–691.
Dundon, T., Wilkinson, A., Marchington, M., & Ackers, P. (2004). The meanings and purpose of employee voice. International Journal of Human Resource Management, 15(6), 1149–1170.
Employment Rights Act 2025, c. 36. (2025). legislation.gov.uk.
Faragher, J. (2026, September 28). Employment law changes in October 2026. Personnel Today.
FashionNetwork.com. (2024, July 18). UK Amazon workers urging union recognition lose by 29 votes. FashionNetwork.com.
Freeman, R. B., & Medoff, J. L. (1984). What do unions do? Basic Books.
GMB. (2022, May 12). GMB and Deliveroo sign historic recognition deal. GMB Union.
Hirsch, B. T. (2008). Sluggish institutions in a dynamic world: Can unions and industrial competition coexist? Journal of Economic Perspectives, 22(1), 153–176.
Hirschman, A. O. (1970). Exit, voice, and loyalty: Responses to decline in firms, organizations, and states. Harvard University Press.
In These Times. (2024, February 29). A first contract is finally in sight for unionized Starbucks workers. In These Times.
Isidore, C., & Delouya, S. (2023, October 13). Kaiser Permanente reaches tentative deal with union workers after largest-ever US health care strike. CNN Wire / ABC7 News.
Jäger, S., Schoefer, B., & Heining, J. (2021). Labor in the boardroom. Quarterly Journal of Economics, 136(2), 669–725.
Kilkenny, K. (2025, May 30). Video game workers reach tentative agreement with Microsoft's ZeniMax. The Hollywood Reporter.
Kochan, T. A., Eaton, A. E., McKersie, R. B., & Adler, P. S. (2009). Healing together: The labor-management partnership at Kaiser Permanente. ILR Press.
Krisher, T., & Megerian, C. (2024, April 19). Tennessee Volkswagen employees overwhelmingly vote to join United Auto Workers union. Associated Press.
Labor Management Partnership. (n.d.). History of the LMP. Kaiser Permanente and Coalition of Kaiser Permanente Unions.
Microsoft. (2022, June 13). CWA, Microsoft announce labor neutrality agreement [Press release].
Morrison, E. W. (2011). Employee voice behavior: Integration and directions for future research. Academy of Management Annals, 5(1), 373–412.
Morrison, E. W. (2014). Employee voice and silence. Annual Review of Organizational Psychology and Organizational Behavior, 1, 173–197.
National Public Radio. (2024a, April 19). Volkswagen workers vote yes to unionizing, igniting UAW's push to organize the South. NPR.
National Public Radio. (2024b, May 17). Mercedes workers vote no to union; UAW says they were illegally intimidated. NPR.
NHS Employers. (n.d.). NHS Staff Council. NHS Confederation.
Peyton, D. L., & Penman, A. (2026, September 30). How UK employers can prepare for new trade union rights that take effect in October. McGuireWoods.
Rowlands, R., Bracebridge, C., Mann, S., & Michaelides, A. (2026). UK Employment Rights Act 2025: A new era for UK trade unions. Are employers ready? Covington & Burling.
Saad, L. (2026, September 1). Record-high 47% want unions to have more influence. Gallup.
The Register. (2024, July 30). First unionized Apple store in the US reaches tentative labor agreement. The Register.
U.S. Bureau of Labor Statistics. (2026). Union members — 2025 (USDL-26-0229) [News release].
Willman, P., Bryson, A., & Gomez, R. (2006). The sound of silence: Which employers choose no employee voice and why? Socio-Economic Review, 4(2), 283–299.
Jonathan H. Westover, PhD, Chief Research Officer (Nexus Institute for Work and AI); Co-Founder & Chief Workforce and Learning Officer (Future State University); Founder & CEO (Human Capital Innovations); Professor of Organizational Leadership & Change (UVU). Read Jonathan Westover's executive profile here.
Suggested Citation: Westover, J. H. (2026). Open Doors: Union Access Rights, Employee Voice, and How Leaders Can Build Constructive Labor Relations. Human Capital Leadership Review, 40(1). doi.org/10.70175/hclreview.2020.40.1.5
