kaufman-2015-human resource management journal

Upload: tantanle

Post on 07-Jul-2018

215 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    1/22

    Theorising determinants of employee voice: an

    integrative model across disciplines and levels of

    analysis

    Bruce E. Kaufman, Department of Economics, Georgia State University, and Centre

    for Work, Organization and Wellbeing, Department of Employment Relations and

    Human Resources, Griffith University

    Human Resource Management Journal, Vol 25, no 1, 2015, pages 19–40

    This article critiques organisational behaviour (OB) research on employee voice and presents a

    broader-based conceptual model that integrates ideas and concepts across employment relationship

    disciplines and levels of analysis. OB studies err by taking an overly individualistic, psychological,

    managerialist and de-institutionalised perspective on employee voice. This criticism is documented andillustrated with numerous examples from the OB literature. To provide a constructive step forward, the

    article presents an enlarged model of employee voice that not only includes OB but also brings in

    important contributions from the HRM, industrial relations, labour economics and labour process fields.

    The model provides an integrative framework for theoretical and empirical studies of voice and yields a

    number of research and practice implications.

    Contact:  Bruce E. Kaufman, Department of Economics, Georgia State University, Atlanta, GA

    30303, USA. Email:  [email protected]

    Keywords: employee voice; HRM; employment relationship; organisational behaviour; indus-

    trial relations

    INTRODUCTION

    The employee voice concept was popularised in the research literature by Freeman and

    Medoff (1984), although its roots go back more than two centuries (Kaufman, 2014a). Thenumber of books and articles on employee voice began to mushroom 10 years later and

    spread into industrial sociology, labour law and behavioural science fields, such as

    industrial-organisational psychology, organisational behaviour (OB) and HRM. Illustratively,

    symposia on employee voice have been featured in the   Journal of Management Studies(September 2003),  Socio-Economic Review   (May 2006),   Human Relations   (March 2010),   Human

    Resource Management   (January 2011) and   Industrial Relations   (January 2013). Also recently

    published is a 29-chapter  Handbook of Research on Employee Voice   (Wilkinson  et al., 2014).Naturally, researchers in these many different fields and disciplines explore the voice topic

    from different frames of reference and with different concepts and methods. Also, the ongoingtrend towards academic specialisation both across and within fields tends to fragment the

    research literature on employee voice into self-contained and self-referential silos (Wilkinson

    and Fay, 2011) – referred to by Suddaby (2012), editor of the  Academy of Management Review, as

    the ‘balkanization of management theory’ (p. 7).This article seeks to break down some of these employee voice silos and foster a more

    integrative and cross-disciplinary research dialogue. No field is free of excessive specialisation

    and narrowness of approach; however, the OB segment of the voice literature seems

    particularly divorced from the historical mainstream of the subject and theories and findingsin other employment-related fields (Godard, 2014; Mowbray  et al., 2014; Pohler and Luchak,

    bs_bs_banner

    doi: 10.1111/1748-8583.12056

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 19

    © 2014 John Wiley & Sons Ltd.

    Please cite this article in press as: Kaufman, B.E. (2015) ‘Theorising determinants of employee voice: an integrative model across disciplines andlevels of analysis’.   Human Resource Management Journal  25: 1, 19–40.

    mailto:[email protected]:[email protected]

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    2/22

    2014). Hence, this article proceeds in a two-step process of critique and reformulation. The first

    step is to outline the OB model of employee voice, using recent articles by Morrison (2011) and

    Klaas  et al. (2012) as a focal point, and identify its important shortcomings (also see Brinsfield,

    2014). The second step is to take the OB perspective and build on it a more integrative and

     broad-based voice model by incorporating key concepts and ideas from other fields, such asHRM, industrial relations (IR), labour economics (LE) and the labour process (LP) part of 

    industrial sociology. These perspectives receive chapter-length treatment in the Wilkinson  et al.

    (2014) research handbook, and this article can be viewed as an attempt to formalise a greaterconceptual unity among them in the spirit of the original cross-disciplinary and multilevel

    employment relations (ER) field (Edwards, 2003; Budd, 2004; Kaufman, 2004).

    OB: CONCEPTUALISATION AND MODEL OF EMPLOYEE VOICE

    The OB-centred literature on employee voice was recently summarised and synthesised in two

    lengthy review articles in top-tier journals by Morrison (2011) and Klaas  et al. (KOBW, 2012).1

    Both articles are billed in their titles as integrative and include in their bibliographies more than

    two hundred studies. These articles do not speak for all OB researchers; a search of theliterature can always find exceptions to the criticisms and generalisations made here, and the

    two articles and literature they summarise may have a discernible American slant.

    The place to start is the definition and conceptualisation of employee voice. Morrison (2011)

    states, ‘I offer the following integrated conceptualization of voice: discretionary communication

    of ideas, suggestions, concerns, or opinions about work-related issues with the intent toimprove organizational and unit functioning’ (p. 375). She also lists these specific features of 

    employee voice: verbal expression, individual choice, face-to-face, prosocial and constructive

    intent. Her article focuses on psychological antecedents, processes and outcomes of voice at theindividual and small group levels, and excludes from coverage the voice literatures inorganisational justice, HRM and IR (LE is not mentioned) on the grounds that they ‘have not

    considered discretionary voice  behaviors, nor the causes or consequences of this behavior . . .

    [but] a wide range of formal mechanisms’ (p. 381, emphasis in original).

    Unexpectedly for a review article, KOBW do not provide a definition of the voice construct.However, their perspective is also centred in micro OB, and in particular psychological

    determinants and processes. For example, KOBW describe voice as a discretionary choice made

     by the individual employee that typically consists of communicating ideas, opinions and

    preferences upward to superiors in the organisation. They also observe that the dominant

    assumption that voice is functional for the organisation (p. 337) brings benefits to managers

    (p. 328) and takes a prosocial orientation (pp. 337–338). They cite seven determinants of voice(p. 316), and these have a heavy micro OB orientation, such as trait-like individual differences,

    satisfaction-commitment-loyalty and organisational culture. Unlike Morrison, however, KOBW

    give some consideration to formal and collective forms of voice, and posit not only a prosocialorientation but also a justice orientation rooted in dissatisfaction and potential conflict of 

    interests. This facet leads KOBW to briefly discuss grievance systems and mention collective

     bargaining; they do not, however, specifically introduce HRM or IR into their integrative review

    (LE is not mentioned).Based on her review and synthesis of the micro organisational literature, Morrison depicts

    in a diagram the core independent and dependent variables and causal orderings for a model

    of employee voice. It is reproduced as Figure 1. KOBW do not present a diagram, but thediscussion in their article closely follows Figure 1.

    Integrative model of employee voice

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 201520

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    3/22

    Does the OB model depicted in Figure 1 capture the strategic ‘must have’ elements necessary

    for a manager or researcher to adequately understand the basics of employee voice? Aspreliminary evidence, consider these 10 features.

    1. Employees are included as active decision makers, but organisations and managers are

    treated as autonomous contextual factors (top left-hand box).

    2. The amount and type of voice is explained without reference to the goals, strategy andperformance of organisations and their managers (none included in the Contextual Factors

     box).

    3. Voice is explained without reference to programmes, policies and strategies that companies

    create for HRM, employee involvement and workplace governance.

    4. Also omitted is the influence on voice of conditions in the organisation’s external

    environment, such as prosperity level of the economy, laws governing employee voice,extent of trade unionism and individual versus collective cultural attitudes (none are

    included in the top two boxes).

    5. The assumption is that employees want to speak up to help managers and the organisation, but this predisposition is never explained (lower left-hand box).

    6. The ‘benefit the organisation’ motive appears to rest on an implicit assumption that the ER

    is typically unitarist, cooperative and mutual gain.

    7. Economic considerations, such as the effect of voice on the organisation’s profit andemployee’s wages, are not included as contingencies in the choice process or voice

    outcomes (the outcome ‘employee rewards’ being closest to an exception).

    8. Only individual voice is modelled; collective forms of voice are omitted and no distinctionis made between direct and indirect (representative) voice.

    Figure 1  Morrison’s voice model

    Bruce E. Kaufman

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 21

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    4/22

    9. Ignored are other dimensions of voice besides communication and problem-solving, such as

    exercising power, providing democratic governance, solving organisational failures,

    maintaining justice, and protecting employee rights.

    10. Voice is portrayed as typically good for organisational performance (the numerous plus

    signs in the organisational outcome box) so managers should revise organisational practicesto encourage more of it, but no positive feedback loop is included and no explanation given

    where the upward voice cycle stops.

    No model can include all relevant factors; further, a model has to be tailored to the level of 

    analysis and research question. This article’s contention, however, is that these 10 items

    represent strategic omissions even at the individual level of voice research. The amount, quality

    and value-added of individual employee voice, for example, surely depend on the firm’s

     business strategy, use of traditional or high-performance HRM practices, availability of formal/group voice mechanisms, cooperative versus adversarial tenor of employee relations,

    employee cultural attributes (e.g.   power distance, individual vs. collective orientation), and

    whether alternative jobs are plentiful or scarce. As argued by Godard (2014),Mowbray et al. (2014), Pohler and Luchak (2014), and Cullinane and Donaghey (2014), a broaderand more inclusive model that gives balanced consideration to all stakeholder roles and

    interests is needed.

    ALTERNATIVE ER MODEL

    Voice takes place within an employment relationship that crosses traditional disciplinary borders.

    HRM is often framed as management of the employment relationship, and for this reason is said

    to be cross-disciplinary (Wilkinson  et al., 2010). Academic researchers, however, are trained in

    narrow areas of specialisation, and the broader cross-disciplinary dimension is often neglectedand sometimes defensively greeted – per the observation by Burris (2012) that ‘[v]oice that

    fundamentally challenges is likely to be met with resistance’ (p. 853). Given this lacuna, Figure 2

    presents an alternative model of employee voice with roots in the cross-disciplinary field of ER.

    The model is necessarily a generalised and broad-brush representation with only the moststrategic features of voice included from the various literatures. It is, nonetheless, a  model  – as

    opposed to a classification system or catch-all list of variables culled from each field – because the

    component parts are conceptual entities linked together in a cause–effect structure that yields

    hypotheses and explains patterns in the dependent variable.The model starts at the top and proceeds to the bottom. The final step in the structure is the

    determination of the outcome variable, employee voice. The most important causal and

    feedback relationships are indicated with arrows; many links are omitted, however, to keep thediagram legible. Individual components, such as the firm’s desired amount of voice and therole of ER climate in the firm, are theorised in greater detail in Kaufman (2014b,c).

    Employee voice: defined and conceptualised

    Employee voice is the dependent variable to be explained. One reason the voice literature

    remains fragmented is that the voice construct is not commensurate across authors and fields.

    OB writers, for example, define voice as individual expression for problem-solving, while manylabour economists define voice as synonymous with unions and collective bargaining (e.g.Addison, 2005).

    An alternative approach is to ask managers how they conceive voice. Wilkinson  et al. (2004)conclude from field interviews that voice encompasses  consultation, communication, and say, and

    Integrative model of employee voice

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 201522

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    5/22

    varies along dimensions of  form  (direct vs. indirect),  agenda (shared vs. contested) and  influence(power or ‘muscle’). Accordingly, employee voice to managers has both integrative/

    pie-growing and distributive/pie-sharing dimensions, and takes place in highly variegated

    forms and settings ranging from an informal 5-minute problem-solving discussion to collective bargaining. Voice also includes not only face-to-face talk but also other forms of 

    communication, such as emails, grievance filings and striking.

    This conceptualisation of employee voice is far wider than commonly used in OB. However,

    it accords with the definition used by other researchers who take a broad ER perspective. The

    editors of the voice research handbook, for example, define voice as ‘the ways and meansthrough which employees attempt to have a say and potentially influence organizational affairs

    relating to issues that affect their works and the interests of managers and owners’ (Wilkinsonet al., 2014).

    This definition encompasses all employment relationships and includes both thecommunication and influence dimensions of voice. Also, this conceptualisation can be

    measured in workplace surveys, such as along the dimensions of form, agenda and influence.

    These data can then be converted into a low-to-high ranking, similar to the way individual

    HRM practices are combined into a composite index for firm performance studies andrepresented as a voice frequency distribution (Kaufman and Miller, 2011; Kaufman, 2014b).

    External environment

    The model starts with an organisation’s  external environment. This category includes factors suchas economy (e.g. macro cycles, industry growth),  legal (e.g. employee rights, union organising and

    Figure 2  Employment relations (ER) model of employee voice determinants

     

    Internal Contingencies

    EmployerProduction Technology

    Managerial Quality

    Organisational CultureEmployee

    Psycho-Social Dispositions

    Knowledge, Skills, Abilities

    Demographics

    External Environment

    Economic Legal Social-Cultural

    Organisational

    Configuration

    Governance

    Structure

    Employer

    Goal: Efficiency and ProfitBenefits

    Costs

    Strategy

    EmployeeGoal: Enhanced Well-Being

    BenefitsCosts

    Strategy

    Voice System

    Individual versus Collective

    Topics and Agenda

    Communication versus Influence

     

    Voice System

    Individual versus Collective

    Topics and Agenda

    Communication versus Influence

    Interaction Process

    Markets

    Management

    Individual

    DesiredVoice

    Demand

    Desired

    VoiceSu l

    Employee Voice

    Employment Relationship

    L  a  b  o ur M

     a r k  e  t     L  a   b

      o  u  r   M  a  r   k  e   t

    ER Climate

    Conflict/Adversary

    Cooperation/

    Partner

    Authority

    Bruce E. Kaufman

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 23

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    6/22

     bargaining regulation, co-determination) and   cultural-social   (e.g.   individualist vs. collective

    attitudes; authority-conformity norms).

    Morrison and KOBW almost entirely omit external environment contingencies on the

    argument their focus is on individual choice about an extra-role behaviour outside of formal

    voice structures. Surely, however, the type and intensity of workplace problems, the mannerworkers choose to present them to management, and the financial ability of firms to

    constructively address them are significantly influenced by whether the economy is in recession

    or at full employment, the law allows companies to easily terminate workers or switch frompermanent to temporary/contingent workers, and if workers are located in Sweden, China or

    Brazil. In support of this position, the authors of the introduction chapter to the  Sage Handbook 

    of Organizational Behavior  (Hoption  et al., 2008) state that organisations have to be theorised as

    open systems and ‘any explanation of individual-level employee behavior that ignores these

    external events will likely result in a narrow perspective of OB’ (p. 8).Illustratively, Ng and Feldman (2012) perform a meta-analysis of 55 studies on the

    relationship between job stress and voice, find a negative association, and suggest in the

    managerial implications section that to get more voice managers should try to reduce ‘highlevels of work and organizational demands’ (p. 230). They fail, however, to square thisrecommendation with the widespread conclusion in the management literature that firms over

    the last several decades are facing intensifying market competition and therefore have to get

    more performance from employees at the same time they hold down labour and HR costs (e.g.Paauwe et al., 2013; Kaufman, 2014d). Indeed, LP scholars (Marks and Chillas, 2014) argue thatit is the essence of capitalism that markets continually erode profit margins and increase the

    pressure/stress on owners-managers to ‘get more for less’ from employees. In this situation,

    constructive prosocial voice is likely to atrophy, while self-oriented voice from greater job stress

    and work intensification creates a growing volume of complaints, criticisms and demands.

    The lack of attention in micro OB studies to the external environment of organisations makesthe models and conclusions particularly susceptible to ethnocentric bias because so much of 

    their explanatory power comes from potentially culture-specific motivations, attitudes, norms

    and psychological frames. These considerations are omitted by Morrison (2011) and mentioned

    in one place by Klaas  et al. (2012: 321), implying either these articles are implicitly meant forlargely a North American audience or the psycho-social determinants of voice have universal

    fit across regions of the world.

    Also important are complementarities among external environment variables that create

    distinct national employment/voice systems (Katz and Darbishire, 2000; Wood  et al., 2009).

    Voice breadth, depth and form, for example, are likely considerably different in a liberal marketeconomy, such as America, versus a coordinated market economy, such as Germany, because

    the former emphasises individual employee exit into competitive labour markets as the routeto improved job satisfaction, while the latter emphasises collective voice in regulated labour

    markets through works councils, trade unions, co-determination and other organisationalforms.

    Organisational configuration

    Positioned below the external environment in Figure 2 is a roof. It represents the form and

    structure of an organisation. At the ends of the roof are written, respectively,   organisational

    configuration  and  governance structure. Their left/right position on the roof is not important for

    the components underneath.

    Organisational configuration represents alternative organisational designs and managementsystems, such as size, division and coordination of tasks, hierarchical versus flattened structure,

    Integrative model of employee voice

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 201524

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    7/22

    command/control versus participative management, and standardisation-formalisation of 

    processes (Short et al., 2008). Organisational configurations exhibit large diversity, starting with

    mechanistic versus organic designs and extending to more finely grained forms such as

    represented by simple, machine, bureaucracy and commitment models (Cordery and Parker,

    2007). Component parts of organisations, such as the employment system, HR function andwork team, also have configurational structures (Verburg  et al., 2007).

    Research shows that the amount and form of employee voice vary across these

    configurations, being low and informal in simple structures, and high and formalised incommitment organisations (Toh   et al., 2008). Employee voice, for example, is considered a

    linchpin practice in firms with a high-involvement system, and according to Lawler (1986) is

    encouraged by greater delegation to employees of four critical voice ingredients: information,

    knowledge, rewards and power (p. 3). These resources for voice have large variation not only

    across firms (e.g.   fast food restaurants vs. consulting firms) but also within firms (e.g.   clericalvs. technical), per emphasis in HRM on strategic differentiation of high-performance practices

    across work groups (Huselid and Becker, 2011). KOBW omit consideration of organisational

    configuration; Morrison lists ‘organisational structure’ in the contextual factors  box but discussesit in the narrow context of how bureaucracy and hierarchy stifle face-to-face upward voice.

    Governance structure

    On the other side of the roof is   governance structure. Every organisation is not only a formal

    structure of positions, departments and functions coordinated by a management system but

    also a political structure that contains executive, legislative and judicial rule-making and

    enforcement functions. All three functions may be performed by a single person, such as theowner/entrepreneur, or may be delegated and decentralised within the organisation, such as

    through various voice institutions (e.g.   works council, arbitration board). The governance

    structure determines decision-making rights, the distribution and exercise of authority andpower, and the rights and responsibilities of subordinates (Kaufman  et al., 1995; Rousseau andShperling, 2003; Budd, 2004). It also determines whose interests get included in decision

    making, how much influence and participation employees are given, and expresses an

    organisation’s ethical values and moral commitments to employees (Van Den Berg, 2004;

    Verhezen, 2010). A governance structure consideration omitted by Morrison and KOBW, forexample, is the effect of a shareholder versus stakeholder model of the firm on the amount and

    form of employee voice behaviour (presumably more voice in the latter). More generally, both

    sets of authors observe that employees are fearful of speaking up to superiors for fear of 

    retaliation which, presumably, suggests strengthening the governance structure.

    One set of writers, such as Burris (2012), follow Morrison and are silent on the governance

    subject and instead recommend increased training for managers and employees so the formerare more open to voice and the latter tailor voice to be more constructive and non-threatening

    (referred to as ‘high-quality’ voice by Morrison). Another set, such as Takeuchi  et al. (2012),

    follow KOBW and recommend that managers consider increased protection for employees whouse voice, principally in the form of a grievance or alternative dispute resolution programme.

    OB writers in both groups, partly due to their emphasis on individual and direct types of voice,

    do not go further and recommend revising the governance system to include a form of 

    collective/representative voice or greater legal protection, such as an employee council orlegislative just-cause termination requirement. Actually, these topics are ruled out by Morrison

    (2011) when she delimits OB research to discretionary  behaviours  and not   formal mechanisms  (p.

    381) – although later in the article she observes that ‘individuals are more likely to voice . . .where there are structural mechanisms for providing input’ (p. 386).

    Bruce E. Kaufman

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 25

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    8/22

    A rationale for employee voice going back more than a century is that it balances power in

    organisations and introduces democratic elements of representation, participation, influence

    and due process into otherwise authoritarian top-down organisations (Webb and Webb, 1897;

    Leiserson, 1922; Foley, 2014). Present-day researchers in IR, HR and LE carry forward this

    perspective, noting that organisations use different governance structures that provideemployees with varying degrees of power (and empowerment) and opportunities for

    individual/collective and informal/formal voice and participation (Gospel and Pendelton,

    2010; Lewin, 2014; Willman  et al., 2014). Also central is the idea that employment relationshipsinevitably contain conflicts of interest among participants, and workers on moral and ethical

    grounds deserve voice to protect and promote their well-being (Budd, 2004; Kaufman, 2005;

    Guest, 2008).

    Employment relationship

    Under the roof is the employment relationship, also called the employee–organisation relationship(Shore  et al., 2012). The ER comes into existence when a business owner or managerial agent

    hires a person who agrees for a wage or salary to work as directed. The employmentrelationship may include as few as two people or as many as hundreds of thousands.

    The employment relationship utilises two alternative coordination modes with, respectively,horizontal and vertical dimensions. One coordinating mode is the labour market where firms

    go to get new employees (or return employees no longer needed) and workers go to find jobs.

    In the diagram, the labour market is outside the walls of the organisation and the flow of labour

    in and out of the firm is depicted by the horizontal arrows. The matching process through

    demand and supply establishes a market-wide pattern of wages and other terms and conditionsfor jobs and workers of different characteristics and provides incentives for both parties to live

    up to their side of the agreement. For example, dissatisfaction leads to turnover, which imposes

    a cost on both firm and the worker.The second coordination mode is the organisation where production takes place and

    managers direct employees in the performance of their jobs. Per the organisational

    configuration discussion above, most organisations have a hierarchical pyramid structure with

    coordination achieved by a vertical command and control process where authority and

    decision-making rights are located at the top and through directives and administration arepassed down (or delegated) from managerial order-givers to production worker order-takers.

    Employment contract law, earlier known as the law of master and servant, reinforces the

    vertical authority/command structure (the downward arrow from authority) by stipulating that

    as a condition for keeping the job the employee must faithfully follow the directions and

    policies of management. Inside the firm, the visible hand of HRM, rather than the invisible

    hand of demand/supply, sets wages and other terms and conditions, recruits, develops andmotivates employees, and monitors and enforces work performance.

    Also fundamental to the employment relationship is the incomplete nature of the labour

    contract (Simon, 1951; Marsden, 1999; Edwards, 2003). The employer and worker reach anunderstanding, sometimes formally articulated in words or writing, about the contents of the

    exchange, such as pay, work hours and job duties, but many aspects are left implicit and filled

    in later as the job proceeds. Here originate three important constructs in OB voice theory:

    psychological contracts, social exchange and organisational citizenship behaviours (OCBs).

    Because the labour contract is incomplete, the employer and the employee develop mental beliefs about the expected terms and conditions of the exchange and what each side owes the

    other. These expectations involve not only the economic dimensions of the exchange but alsosocial dimensions, such as respect and fair treatment. Social exchange is regulated by a norm

    Integrative model of employee voice

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 201526

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    9/22

    of reciprocity, and if one side fails to deliver on expectations or obligations the other perceives

    a violation of the psychological contract and reduces discretionary contributions (Farndale  et al.,

    2011). Many discretionary contributions are conceived of as OCBs, such as conscientiousness,

    hard work and loyalty. Thus, the incomplete nature of the labour contract provides an

    opportunity for employee voice to help solve unanticipated problems and contributeproductive ideas, but employees are only motivated to provide this discretionary extra-role

    OCB if managers not only honor their side of the psychological contract but also reciprocate

    with extra-role inducements. The end result, if all goes well, is a virtuous win-win – voice helpsfirms perform better and firms reward employees with valued economic/social goods.

    Given this outline of the employment relationship, consider implications. To begin, note that

    Morrison’s model represents a highly truncated version of an employment relationship. Only

    the employee side is represented in Figure 1 as an active agent in the voice process. The model

    also omits the labour market half of the ER, and thus the economic part of the exchangeprocess. Indeed, omitted are  all  external influences on voice that originate outside the walls of 

    the organisation (none in the Contextual Factors box). To put this feature in perspective, the

    parallel situation is if an economist constructed a voice model that includes only markets andthe external environment, and leaves blank in Figure 2 everything inside the walls of theorganisation.

    To further accent this lacuna, consider omission of the labour market part of the employment

    relationship. Both Morrison and KOBW emphasise that employees choose to provide the

    amount/type of voice based on a cognitive evaluation of the benefits/costs of speaking up. Thecosts principally involve negative consequences for the employee from management retaliation,

    such as lower performance ratings and being fired. The state of the labour market, say as

    measured by unemployment rate in the worker’s local area, is surely an important cost

    determinant. For example, if the labour market is strong and many jobs are available, workers

    perceive a lower cost of speaking up since if retaliated against they can more easily switchemployers (the exit option); high unemployment, however, tilts employee choice towards

    silence (Kaufman, 2014b).

    The state of the labour market should, therefore, be included as a context variable in Figure 1

    and control variable in empirical studies. Van Dyne and Lepine (1998), for example, perform across-sectional and longitudinal analysis (21 organisations over 6 months) of individual

    employee’s voice contributions. Six control variables are used but none pertain to the determi-

    nants in the external environment. It seems plausible, however, that if the unemployment rate in

    (say) Ohio in 1998 is twice as high as in Nebraska in 1997, the effect should be a lower level of 

    employee voice – and also lower critical message type – in the former.Going further in seriousness, taking into account external trends in the employment

    relationship seriously weaken OB’s social exchange and OCB pillars. HRM writers, for example,frequently argue that markets have become more competitive, thus necessitating the use of 

    high-performance practices (Kaufman, 2014d). A way to conceptualise this trend is that theexternal labour market (ELM) outside the organisation in Figure 2 is growing larger and more

    powerful, while the internal labour market (ILM) inside the firm is growing smaller and

    weaker. This result is said to have given rise to a ‘new employment relationship’. Tsui and Wu

    (2005: 116) observe, ‘the new employment relationship is a quasi-spot contract . . . Employers. . . are interested primarily in a high level of employee task performance without requiring

    commitment . . . Additionally, employees do not expect the employer to provide long-term job

    security’.

    What are the implications of the new employment relationship for employee voice? First,with the new employment relationship, power shifts in favour of firms and they want less, not

    Bruce E. Kaufman

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 27

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    10/22

    more, employee voice. Hollowed and shrunken ILMs need less voice to facilitate efficient

    management coordination, and likewise spending money to promote voice via a fair and just

    workplace becomes an increasingly poor investment as turnover increases and job tenure and

    organisational loyalty fall. Second, employees also have less desire for organisational voice in

    a more market-mediated employment relationship because they increasingly rely on exit if dissatisfied, correctly realise they can be more quickly and inexpensively replaced if speaking

    up irks a manager, and see no pay-off to helping the company with discretionary OCBs when

    they are not likely to be around to share in the gains. Finally, the norm of reciprocity uponwhich rests the power of social exchange theory shrinks in lockstep with the disappearance of 

    employer–employee commitment to cooperate together in a long-term mutual gain

    relationship. Following the theoretical insights of Nobel laureates Coase (1937), Simon (1951)

    and Williamson (1985), one can package these ideas as an argument that technological change

    and market deregulation have lowered transaction cost of using markets, and accordinglyELMs, transactional employment contracts and exit are supplanting ILMs, relational contracts

    and voice (Kaufman, 2010, 2013; Willman  et al., 2014). Also implied is that social exchange

    theory in general, and OCB theory of prosocial voice in particular, have shrinking relevance andexplanatory power.

    A focus on the employment relationship provides two other important insights omitted from

    OB voice research. Owners of capital only hire employees if doing so is expected to generate

    an economic surplus above cost. Employers gain from employee cooperation, therefore, because

    the extra discretionary work effort and OCBs increase the surplus and amount available forprofit. However, workers expect a share in return for their contribution, such as provision of 

    prosocial voice.

    OB writers assume this sharing process takes place through the social exchange process and

    norm of reciprocity. Omitted, however, are these considerations. First, the reciprocity norm may

     be a weak-to-non-existent force in short-term employment relationships or hire/fireemployment systems; when high unemployment makes threat of job loss a cheaper and more

    potent cooperation motivator than investing money to build good will and commitment; and

    when employers can substitute a technological process for getting discretionary effort (e.g.

    electronic surveillance of workers in call centres) in place of a more expensive psychologicalprocess (Marks and Chillas, 2014). Second, employers are incented by competition, financial

    markets and self-interest to surplus-share with employees only to the minimal extent it helps

    profit; hence, the indeterminacy of the labour contract creates an inherent distributive conflict

    over how the larger pie from cooperation is divided (Edwards, 2003). The OB conception of 

    voice is, in this respect, one-sided and employer-friendly because it restricts the subject to thepie-growing function from cooperation, omits (or denigrates as self-centred and harmful) the

    role of voice as a bargaining/influence tool to gain for employees more of the fruits of cooperation, and focuses on the voice type (individual prosocial communication) which is least

    powerful for employees and least threatening for firm profit and managerial control. Third, thelabour contract is not only incomplete but also terminable at-will (absent legal restriction),

    which gives firms an incentive to opportunistically terminate employees in order to keep more

    of the surplus, such as from lay-offs made possible by productivity improvements and culling

    out long-service employees close to pension. Employee voice, in this case, serves a protectivefunction and needs muscle to be effective (Holland  et al., 2012).

    Decision making and voice choice

    The next step is to go to the individual   employer   and   employee   levels and specify thedecision-making process by which both parties make  choices  about their desired level and form of 

    Integrative model of employee voice

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 201528

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    11/22

    voice. Decision making and choice are particularly salient theory topics for an OB model of 

    voice because the focus is on the individual and discretionary behaviour. As Morrison (2011)

    states, ‘individuals choose whether or not to engage in this behavior’ (p. 375). Since decision

    making is a much-researched subfield within in OB (e.g.  the journal Organizational Behavior and

     Human Decision Processes), one would expect this dimension of employee voice to be wellarticulated in OB studies. Surprisingly, it is only lightly discussed by Morrison and KOBW and

    problematically presented.

    The Morrison model can usefully be thought of in a regression equation context:V =  β0 + β1CF +  β2IF +  β3CF·PCE + β4IF·PCE, where V is voice amount, CF is the vector of 

    variables in the contextual factors box (Figure 1), IF is the vector of factors in the individual

    factors box, and PCE is a moderator variable (specified as an interaction term) representing the

    factors in, respectively, the perceived costs and efficacy boxes. This representation brings to

    attention an anomalous aspect of her model. She states that the  primary driving motive  behindvoice (p. 382) is a person’s desire to benefit the organisation. This motive is taken as a ‘given’

    for the individual, is therefore outside the person’s decision-making process, and hence goes

    into the constant term β0 – with the maintained hypothesis  β0 > 0. Variation in prosocial motivesacross individuals thus explains part of voice variation through up and down movements inthe intercept of the regression line but explains  zero of the  discretionary part. A similar problem

    applies to the numerous personality traits that both Morrison and KOBW cite as voice

    determinants. KOBW, for example, argue that part of voice variation is explained by individual

    differences in stable personality traits, such as self-efficacy and proactivity. Like the prosocialmotive, these individual trait differences (constants at the employee level) explain intercept

    changes but none of the discretionary variation in voice.

    A theory of discretionary voice, therefore, has to include a model, however simple, of human

    choice and decision making. Thus, KOBW state that ‘the decision to use voice contains a

    significant calculative element’ (p. 323), and likewise Morrison argues that the discretionarypart of voice comes from an ‘expectancy like calculus’ (p. 384) and cites Vroom’s expectancy

    theory of motivation.

    It seems surprising in light of the strong emphasis in OB on explaining discretionary voice

    and the large OB literature on decision making that authors such as Morrison and KOBW givethis subject such cursory and ill-formulated treatment. Of course, the issue is whether the

    omission makes a substantive difference for management science and practice on voice. To

    provide an affirmative answer, and also point a way forward, Figure 2 outlines elements of a

    serviceable model of voice choice, drawing inspiration from models in both IR and LE. Theargument is much less in favour of any  particular  choice model, however, and much more to

    suggest that OB research on voice will remain muddled without   some  articulated model.

    The decision making model in Figure 2 has four elements, briefly described below and withapplications to voice. The model, in the ER tradition, is built on a rational choice foundation

     but with behavioural amendments (Kaufman, 1999; Budd, 2004).

    Action goal   The first component for decision making is specification of each agent’s action goalor end-state objective. This step is important in order to predict choice because it is otherwise

    impossible to say if a voice outcome represents a perceived benefit or cost to the decisionmaker. For example, Morrison argues that workers will not provide voice if it is seen as futile

    (no benefit), but whether an action is futile depends on the desired outcome. For example,

    writing a negative blog post about management or demonstrating outside a store may be futile

    for personal job advancement but effective for exposing social injustice. Economic voice modelssolve this problem by including the agent’s diverse end goals (e.g. personal advancement, social

    Bruce E. Kaufman

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 29

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    12/22

     justice) as variables in a utility function (utility = well-being), specify determinants of each end

    goal’s benefits and costs, and derive the equilibrium solution. This choice model can be

    modified to incorporate behavioural elements, such as uncertainty, imperfect information and

    fairness. Details aside, the salient point is that to explain and predict an employee’s

    discretionary voice decision, some assumption has to be made about the goal that motivates behaviour – as is well recognised in OB theories of motivation (e.g.  Barrick  et al., 2013).

    Morrison’s model only includes a role for   employee  decision making. Surely, however, the

     breadth, depth and form of voice are also critically influenced by decisions of managers. To usean apt metaphor from economist Marshall (1890), a model that purports to explain the amount

    of voice in organisations but includes only the employee side is akin to trying to cut a piece

    of paper with one-bladed scissors. Accordingly, in keeping with the employment relationship

    concept in Figure 2, underneath it is included   two  decision-making parties: the employer (left

    side) and employee (right side).Individual well-being is the end goal for employees, but according to Sirmon  et al. (2007)

    ‘[t]he primary pursuit of business is creating and maintaining  value’ (p. 273, emphasis added).

    Value is multidimensional and firms have numerous goals (Boxall, 2007), but ultimately theytelescope into running the firm efficiently to maximise the capitalised value of profit and returnon capital – of necessity given the role of profit/loss as capitalism’s selection criterion for firm

    survival and demise. Figure 2, therefore, lists efficiency and profit as the goal objective for the

    employer decision maker. (An objective of higher performance begs the question of how

    performance is measured and evaluated.)Here surfaces a hugely important omitted variable in OB voice models: profit-making. Voice

    form and amount are explained by individual differences in employee motives and

    psychological dispositions, but the financial side of voice is neglected. Illustratively, the word

    ‘profit’ is not mentioned by either Morrison or KOBW, her diagram excludes financial

    determinants of voice, and voice is limited to prosocial forms which by definition benefit firmperformance. But in a business world where companies decide resource investments, including

    in people and voice, on the size of the prospective rate of return, how is it possible to accurately

    theorise the amount and form of employee voice with attention only to psychology and none

    to finance, profit-making and the stock price? Workers in two firms may be identical on everymeasureable psychological trait and disposition, but the voice amount will nonetheless show

    large variance and systemic effects if it leads to larger profit in one (e.g. Costco, Southwest) and

    lower profit in another (e.g. Walmart, Ryan Air). Likewise, a profit perspective, much more than

    a psychological perspective, explains why managers systematically welcome individual and

    constructive voice more than collective and confrontational voice (Cullinane and Donaghey,2014).

    Benefits and costs   The second set of items in the choice process depicted in Figure 2 specifies

    the   benefits and costs   of voice, as subjectively estimated by employer and employee. Theyinclude both pecuniary and psychological-social consequences and are influenced by the

    cognitive and affective states emphasised in OB studies. The benefits of voice for an employer,

    for example, include improved productivity and a more positive organisational climate, while

    the costs include higher wages and diluted authority. For employees, a benefit of voice is higherwages (the mirror image of the cost to the employer) and praise from the boss, while costs are

    potential job termination and interpersonal conflict.

    To predict behaviour, it is not enough to simply list the benefits and costs in a box-arrow

    diagram (e.g.   last two boxes in Figure 1). Benefits and costs get explanatory traction as theyinfluence the ability of employee and employer to attain their utility and profit goals. Here

    Integrative model of employee voice

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 201530

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    13/22

    enters a decision-making insight from economics, which greatly improves the predictive

    content of voice theory but is neglected in OB (and HRM). It is the   marginal method. Assume

    for sake of argument that, in line with OB research, greater employee voice improves firm

    financial performance. This finding, however, is surely not open-ended (e.g.   4 hours of 

    employee voice per day?), yet without the marginal method the optimal stopping point in voicecannot be determined. The decision rule for both employee and employer is similar: increase

    voice as long as the marginal gain exceeds the marginal cost and when the two are equal utility

    and profit are maximised (Freeman and Kleiner, 2000; Kaufman, 2014b). From this perspective,one reason many firms feature little employee voice is that managers have calculated that it

    soon costs more than it brings in.

    Strategy   Another determinant of choice omitted by Morrison and KOBW is, respectively,

    employer and employee strategy  (Dundon and Gollan, 2007). In an uncertain and fast-changingenvironment, employers and employees cannot flawlessly peer into the future and foresee the

     best route to reach their goals. Hence, they must formulate a strategy. Some employees, for

    example, conclude that they are more likely to get ahead by ‘staying under the radar’ andhence will seldom voluntarily offer opinions or voice dissent; others, however, conclude theopposite and use a strategy of using voice to ‘push the envelope’. Others may have little desire

    for voice because their career strategy is to acquire human capital by frequently moving from

    company to company (Royer   et al., 2008). Similarly, one employer may choose a low-cost

    strategy to attain competitive advantage, along with a complementary HRM strategy of hire/fire, and therefore desire small-to-little employee voice. A different employer, however,

    uses a product quality strategy, along with a high involvement HRM strategy, and hence desires

    considerable employee voice (Lepak et al., 2005; Chi  et al., 2011). Of an entirely different nature,

    an organisation’s strategy towards unions (e.g.   hostile vs. accommodative; suppression vs.

    substitution) also influences voice amount and form (Cooper and Briggs, 2009; Marginson et al.,2010).

    Internal contingencies

    Benefits, costs and strategy are not themselves independent variables but mediating variables

    influenced by a variety of other factors. Some of these factors are contained in other parts of the model, such as external environment, organisational configuration and governance

    structure. Another block of determinants in Figure 2 (not entirely separate) is   internal

    contingencies, that is, factors internal to the organisation and employee, respectively. A

    representative sample of factors for each is identified in the middle box under the roof. Internal

    organisational contingencies include, for example, production technology, managerial quality

    and organisational culture. Internal employee contingencies include factors such as theworkforce’s psycho-social attitudes towards work and management, knowledge, skills and

    abilities, and demographics.

    The internal contingencies box of Figure 2 subsumes the factors in the ‘contextual factors’and ‘individual factors’ boxes in Morrison’s diagram, and in this respect is a repackaging rather

    than substantive new addition. (The boxes below them on perceived costs and efficacy are

    embedded in the decision making boxes of Figure 2.) With respect to the KOBW article, the

    internal contingencies listed for the employer in the top part of box is new; that is, KOBW giveconsiderable attention to the role of the employee’s psychological structure (traits, dispositions,

    etc.) as voice determinants but mostly omit the firm’s organisational structure (production

    technology, organisational configuration). Whether the firm has an assembly line, self-managedteam or bureaucratic form of production surely has as much – probably a great deal more –

    Bruce E. Kaufman

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 31

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    14/22

    influence on the amount and form of discretionary voice than whether workers are low or high

    in self-efficacy and extroversion (i.e. the assertion is that for a one standard deviation in CF and

    IF,  β1 > β2).

    Further, since workers with heterogeneous traits and attitudes sort among firms to find the

     best person/job fit, much of the cross-section statistical evidence presented in OB studies ontraits and dispositions as determinants of voice is systematically biased and unreliable. For

    example, Crant  et al. (2011) find that proactive personality is the strongest predictor of voice.

    They inform managers (pp. 294–295):

    ‘Organizations that rely on employee suggestions as a source of continuous

    improvement are advised to identify the extent to which job applicants possess

    proactive personality, and to evaluate the proactivity of current employees as part

    of promotion decisions . . . [and] as part of a broader selection process’.

    Neglected is the fact that the positive effect of proactivity on voice is statistically inflated

     because proactive people sort into firms providing more opportunity for voice. The advice that

    firms alter selection procedures to target applicants with proactive personalities is also flawed.If this trait is in limited supply and gives firms higher productivity, the workers with this traitgain through demand/supply competition a higher wage – implying that the extra

    performance value of the trait brought into the firm through the selection process in the end

    accrues not to the firm (higher profit) but the worker (higher wage), as theorised in the

    resource-based view of the firm (Barney and Clark, 2007: ch. 2).At this point in Figure 2, the employer and employee, influenced by all the factors and

    influences described above, have decided on the amount of individual voice they desire.

    Morrison and KOBW effectively end the theory of voice determinants here and proceed to voice

    outcomes (the last two boxes in Figure 1). A more integrative and complete model, however,

    requires four more parts.

    ER climate

    Morrison and KOBW mention climate at several places as a contextual and mediating factor,

    and in a different article Morrison and colleagues give explicit attention to conceptualising

    climate and identifying its effect on voice (Morrison  et al., 2011). They say climate as a genericconstruct refers to ‘collective beliefs or perceptions about the practices, behaviors, and activities

    that are rewarded and supported in a given work environment’ (p. 184). They then define voice

    climate as a two-dimensional construct encapsulating shared beliefs about, respectively,

    whether speaking up is safe versus dangerous, and effective versus futile. As noted earlier,

    these boxes in Figure 1 are also central to her voice decision making model (utilising Vroom’s

    expectancy model of motivation), and thus climate affects voice rather narrowly throughindividual perception of benefits and costs of speaking up.

    A climate construct is also used in voice studies in other fields, such as HRM/IR. Pyman et al.

    (2010), for example, define IR climate as a subset of organisational climate, and say that it refersto ‘the atmosphere, norms, attitudes and behaviors reflecting and underpinning how workers,

    unions, and managers interact collectively with each other in the workplace’ (p. 463). Other

    studies equate climate with organisational culture (Clapham and Cooper, 2005) or distinguish a

    specific component of organisational culture, such as justice climate (Goldberg et al., 2011).All of these conceptualisations are helpful but also too narrowly tailored to specific voice

    dimensions and organisational contexts. Figure 2 introduces a broader conceptualisation, called

    ER climate. ER climate is the perceived tenor or ‘good-bad’ quality of relations betweenmanagement and workers in organisations. A way to model the ER climate is to use the ER

    Integrative model of employee voice

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 201532

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    15/22

    ‘four frames of reference’ model (Budd and Bhave, 2008; Kaufman, 2014c). Employment

    relationships fall into four categories – market, unitarist, pluralist and critical – based on how

    the two sides perceive their relation. End points are 100 per cent conflict/adversary (worst) and

    100 per cent cooperation/partner (best). These frame of reference end points create similar ER

    climate end points, as indicated in Figure 2. Bad ER climate is associated with the critical frame,good climate with the unitarist frame, and a mix of conflict/cooperation and adversary/partner

    with the pluralist frame. [The market frame is not relevant to climate since the employment

    relationship is entirely transactional and impersonal (no climate).] The distribution of organisations among these three frames creates a similar frequency distribution of ER climates

    (depicted as a bell-shaped curve). This distribution can be further refined by introducing, in

    addition to broad frames of reference, alternative employment systems (e.g.  simple, machine,

     bureaucratic, commitment, adhocracy); also, the distribution changes shape and shifts left/right

    in reaction to external and internal contingencies.The hypothesis advanced is that the tenor of ER climate influences the amount and type of 

    employee voice (Kaufman, 2014c). Since OB studies recognise this link, the contribution made

    here is to broaden and generalise it. When viewed against the entire climate frequencydistribution and its determinants, one sees, for example, that the OB field has structured theanalysis to include only-to-mostly voice situations and outcomes in the high-end tail of the ER

    climate distribution. KOBW, for example, note that OB studies of prosocial voice posit a

    unitarist-type employment relationship featuring mutuality of interests (p. 327), thus helping

    create a positive ER climate (top end of the distribution) that, in turn, feeds into higher jobsatisfaction for workers and, via the reciprocity norm, a desire to provide helpful other-related

    voice that benefits the organisation.

    Morrison  et al. (2011) qualify this perspective by noting that some firms have a ‘climate of 

    silence’. A broad ER model of voice, as in Figure 2, suggests that many firms deliberately engineer

    a climate of silence, even if it puts them on the low side of the ER climate distribution, becausea ‘lean and mean’ HR strategy (pay cuts, work intensification, downsizing) is more profitable –

    particularly in a short-term finance-driven economy with intense competition – than a long-term

    partnership-mutual gain strategy (Cullinane and Donaghey, 2014; Kaufman, 2014d). These

    ER-level factors are omitted by Morrison and co-authors, however, and instead they link a climateof silence to the shortcomings of individual leadership style and behaviour. The contrary

    hypothesis is that a climate of silence arises not from the shortcomings of individual managers

    (although in individual cases it may) but from their strategic decision to tightly control/suppress

    voice to restrict employee power resources and keep a lid on unwelcome and potentially costly

    complaints and demands – paradoxically, a consideration well recognised in the OB literature onorganisational power (Anderson and Brion, 2014). Illustrative of the unbalanced managerialist

    slant in this literature, KOBW devote a section to the ‘dark side of voice’ but does not discuss init managers using coercion to silence employees (Thompson and Harley, 2007; Gall and Dundon,

    2013) but workers using voice for revenge, harassment and embarrassment of managers.

    Voice system

    The next step in the voice model is  voice system  for employer and employee, respectively. The

    employer and employee have made a choice, based on benefits and costs as shaped by strategies,

    external contingencies and internal contingencies, regarding their desired level of voice (a behaviour). The level of voice cannot be separated from form and type of voice (a structure),

    however, because these dimensions are inextricably connected. In particular, if all of the

    enumerated factors cause the employer and employee to desire a high voice workplace, then thetwo sides cannot simply rely on extensive extra-role face-to-face communication, as focused on

    Bruce E. Kaufman

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 33

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    16/22

    in micro OB studies. Rather, as desire for voice moves from low to high, it necessitates that both

    parties to the employment relationship introduce broader and more formalised voice procedures

    and structures (Gollan   et al., 2014). Wilkinson  et al. (2010: 11) conceptualise this idea as an

    ‘escalator of participation’. Included in this process is transition from individualised and informal

    voice procedures, such as direct voice methods of worker–supervisor discussion and open-doordispute resolution, to progressively more formalised and representative voice structures, such as

    elected employee plant councils, multi-step arbitration-mediation systems, collective bargaining

    and co-determination. Viewed in this light, it seems paradoxical for OB writers to limit voice tothe lowest step of the escalator not only because collective voice forms can benefit firms but also

     because individual voice behaviours are readily studied in formal voice structures – as

    researchers in employment psychology, industrial sociology and IR teamed up to do more than

    a half century ago (Kornhauser et al., 1954). (Morrison’s claim that HRM/IR studies do not study

    individual behaviours is a very large caricature – see, for example, Luchak, 2003; Pyman  et al.,2010; Marchington and Suter, 2013).

    Also, part of individual decision making for both employee and employer is about whether

    to keep voice at the person level or expand it to the collective level. This decision goes backto the end objective of voice, benefits and costs, and strategies. Even if voice has only aconstructive organisational improvement goal, it may be more effectively delivered and acted

    upon in group form, such as quality circle or company-wide employee council, if production

    has significant organisational or knowledge interdependencies. Also affecting the

    individual/collective choice are various organisational failures and social traps (Miller, 1991),such as under-supply of individual voice due to the free rider problem and the lack of trust that

    the other party would not behave opportunistically (per the prisoner’s dilemma game). Another

    important variable affecting employees’ desire for collective voice is perceptions of 

    organisational fairness (Goldberg   et al., 2011) – a variable with a strong psychological

    foundation but not listed in the Morrison voice model.Particularly salient to the individual/collective voice decision, but omitted from OB studies,

    is the distributive pie-sharing and ‘muscle’ element. Worker voice is able to exert more

    influence and power as it is collectively organised, and thus provides workers with more

    effective say, protection and leverage over work terms and conditions. Employers, therefore,may find a productivity advantage from expanding voice up the escalator to a collective level

     but nonetheless keep voice at a low level from dual fears. The first fear (Freeman and Lazear,

    1995; Marsden and Cañibano, 2010; Kaufman, 2014b) is that collective voice at some point starts

    to raise costs and reduce profit by giving employees more ability to improve wages, job security

    and work conditions (i.e. pie-sharing starts to dominate pie-growing); the second is that givingworkers training in and access to collective voice may backfire and lead them to want even

    stronger voice in the firm of an independent union with collective bargaining (Timur  et al.,2011). These dual costs to employers have been identified as major reasons why voice is

    systematically under-supplied in workplaces, evidenced in survey data that show that a large‘participation–representation gap’ exists within and across nations (Freeman  et al., 2007). The

    costs to firms in profit and managerial control have also been linked to empirical findings that

    managers retaliate against employees who voice a complaint, file a grievance or express a desire

    for a union (Gall and Dundon, 2013; Lewin, 2014).

    Voice demand and supply

    The confluence of all of these considerations in Figure 2 leads employer and employee(s) to

    formulate decisions about desired amount, form and type of voice. The employer’s desiredamount is labelled  voice demand   in the diagram, and the employees’ decision, both individual

    Integrative model of employee voice

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 201534

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    17/22

    and collective, is labelled  voice supply. (Alternatively, employers can be depicted as supplying

    and employees demanding voice; the two views are opposite sides of the same coin.) In a

    non-union firm with no formal voice structure, voice on the employee side is scattered and

    informal; as the firm establishes more formal involvement and participation structures, the

    employees’ voice is increasingly aggregated, centralised and formalised, such as in a workscouncil, collective bargaining and co-determination.

    For a variety of collective action and organisational failure reasons, the sum of individual

    demands and supplies that go into the voice system boxes may not aggregate to the amountsthat come out in expressed demand and supply. For example, individual plant managers may

    see the performance advantage of voice, but the corporate chief financial officer sees mostly the

    expense side and uses his/her superior power to nix it. In a similar way, labour unions, such

    as through a median voter process, aggregate the voices of individual workers, and some voices

    therefore get more weight than others (Kaufman, 2012).

    Adjustment-interaction process

    One last step is required. The organisation’s demand for voice usually differs from theemployee(s)’ desired supply in terms of amount, form and type. This divergence alsocharacterises other items in the employment contract, such as wages, work speed and schedule

    flexibility. Accordingly, these dual preferences for voice have to be adjusted, negotiated and

    reconciled into a single amount, perhaps in writing as part of a formal agreement, or more

    likely as an implicit understanding in a psychological contract.

    This adjustment-interaction process takes place through a variety of methods. Workers may exita low-voice organisation and take a job at a high-voice organisation; at the same time, the

    low-voice firms come under pressure to increase voice supply due to expensive turnover and

    desire to not increase wages to fill empty positions (Riordan  et al., 2005). In these examples, the

    labour market helps coordinate voice demand and supply. Actions by company managers and behaviour changes by employees are an alternative method. Workers who feel dissatisfied, for

    example, provide low voice, and other OCBs and managers react by identifying and correcting

    the sources of dissatisfaction, which then encourage more voice. On the other hand, if managers

    do not respond, then dissatisfied workers have other options, such as organising a union forcollective bargaining (Timur  et al., 2011), or electoral politics and street protests to pressure

    government to mandate more workplace voice.

    A variety of channels, therefore, help match the organisation’s voice demand with its

    employees’ voice supply. As earlier indicated, however, empirical studies find that on average

    workers want more voice and influence than firms provide, suggesting that the equilibratingprocess is slow and only partially successful (Freeman  et al., 2007). This finding also suggests

    that when employer and employee voice preferences do not match, it is the employer’spreference that typically predominates and employees bear the gap (employees are on the ‘long

    side’ of the voice market). The size of the voice gap depends on relative power positions andseriousness of market and organisational failures (Kaufman, 2014b). Although the different

    contingencies in Figure 2 appear to typically favour employers, in some situations the power

    position favours employees and they effectively set the level of voice. An example is a strong

    union bargaining with a small firm.

    Voice outcome

    Out of the interaction of employer and employees in the employment relationship, conditioned

     by numerous external and internal contingencies, emerges at the bottom of the diagram theamount, form and type of  employee voice. Voice, in turn, travels back through feedback loops to

    Bruce E. Kaufman

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 35

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    18/22

    affect higher parts of the model (e.g.   management decision making, the national legislative

    process) in a recursive adjustment process, indicated by the upward arrows.

    Voice is the dependent variable and can be conceptualised as a low-high frequency

    distribution, with individual/informal/cooperative voice at one end and collective/

    formal/adversarial voice at the other. An empirical example is the frequency distribution of voice forms, ranging from ‘no formal voice’ to ‘union voice’, provided by the WERS (Work

    Employment Relations Survey) data set for Great Britain (Willman  et al., 2009) and depicted in

    Kaufman (2014b,c). The purpose of theorising is to identify the factors that determine thematrix of employee voice characteristics, its variation in a cross-section or over time, economic

    and social outcomes, and feedback effect of voice on higher order parts of the model (two

    upward arrows). Consistent with this perspective, Morrison (2011) states that the research goal

    is ‘identifying factors that increase or decrease  the amount of voice that an employee engages in’

    (p. 385, emphasis added), and KOBW claim that their integrative review ‘allows for us toexplore the different factors that have been found to   inhibit or encourage   forms of workplace

    voice’ (p. 322, emphasis added).

    The central point of critique is that these OB studies frame employee voice in an overlyconstricted manner and model voice with numerous omitted variables and mis-specified causalrelationships. The end product is a voice model with low R-square, biased coefficients and poor

    predictive performance in research, and a paucity of actionable, reliable and quantitatively

    significant managerial implications. A revealing indicator is trying to use the OB model to

    understand, predict and improve employee voice in real-life organisations. Other fields, suchas IR and HRM (e.g. Gollan et al., 2014), feature many case studies on voice with concrete action

    points for managers. OB research, on the other hand, features an ever-expanding set of 

    psychological constructs with complex statistical analysis of survey data sets (frequently

    obtained from students) to explain voice but with little illustration or evidence from case

    studies (e.g.   no case studies are discussed by Morrison, KOBW or Brinsfield, 2014).

    CONCLUSION

    Huselid and Becker (2011) recently observed that ‘[a] significant divide between the micro andmacro levels of theory and analysis is evident in many areas of the organizational sciences’, and

    concluded that ‘the consequences of this trend can be unfortunate’ (p. 421). This article has been

    a case study of both propositions with respect to the subject of employee voice. Specialisation

    and division of labour are productive in all areas of life, including academic research onemployee voice. The OB side of voice research, however, has carried specialisation too far with

    the result its models and empirical studies are truncated entities inhabiting a narrow

    psychological, individualist, managerialist and scholastic silo having little contact with forcesand ideas in the larger outside world, research in other voice-related fields, or the practice of employee voice in real-life organisations. In order for this article to be not just critical voice, I

    have also sketched the basic building blocks for a more integrative model combining micro and

    macro dimensions in the spirit of the broad and inclusive field of ER.

    Notes

    1. Also see Morrison (2014) – a modestly different but complementary survey article on voicewhich the author learned of only after this article had gone through the review process and

    was in close-to-final form. The most important difference between the two articles regarding

    theorisation of voice is that she modestly reconfigures the diagrammatic model (Figure 1 in both articles) so the box ‘Motive to Help the Organisation’, which is independently repre-

    Integrative model of employee voice

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 201536

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    19/22

    sented in the earlier article, is made one of three elements in a ‘Motivators’ box in the later

    article. Since the two articles are quite similar in terms of substance, no significant inaccuracy

    is introduced by restricting focus only to Morrison (2011).

    REFERENCES

    Addison, J. (2005). ‘The determinants of firm performance: unions, works councils, and employeeinvolvement/high performance work practices’.  Scottish Journal of Political Economy, 52: 3, 406–450.

    Anderson, C. and Brion, S. (2014). ‘Perspectives on power in organizations’.   Annual Review of Organizational Psychology and Organizational Behavior, 1: 1, 67–97.

    Barney, J. and Clark, D. (2007).  Resource Based Theory, New York: Oxford University Press.Barrick, M., Mount, M. and Li, N. (2013). ‘The theory of purposeful work behavior: the role of 

    personality, higher-order goals, and job characteristics’.   Academy of Management Review, 38: 1,132–153.

    Boxall, P. (2007). ‘The goals of HRM’, in P. Boxall, J. Purcell and P. Wright (eds),  Oxford Handbook of  Human Resource Management, New York: Oxford University Press.

    Brinsfield, C. (2014). ‘Employee voice and silence in organizational behavior’, in A. Wilkinson, J.Donaghey, T. Dunson and R. Freeman (eds),  Handbook of Research on Employee Voice, Northampton,MA: Elgar.

    Budd, J. (2004).   Employment with a Human Face, Ithaca, NY: Cornell University Press.Budd, J. and Bhave, D. (2008). ‘Values, ideologies, and frames of reference in industrial relations’, in P.

    Blyton, N. Bacon, J. Fiorito and E. Heery (eds), Sage Handbook in Industrial Relations, London: Sage.Burris, E. (2012). ‘The risks of rewards of speaking up: managerial responses to employee voice’.

     Academy of Management Journal, 55: 4, 851–875.Chi, W., Freeman, R. and Kleiner, M. (2011). ‘Adoption and termination of employee involvement

    programs’. Labour (Committee on Canadian Labour History), 25: 1, 45–62.Clapham, S. and Cooper, R. (2005). ‘Factors of employees’ effective voice in corporate governance’.

     Journal of Management and Governance, 9: 3–4, 287–313.Coase, R. (1937). ‘The nature of the firm’.   Economica, 4: 16, 386–405.Cooper, R. and Briggs, C. (2009). ‘ “Trojan horse” or “vehicle for organizing”? Non-union collective

    agreement making and trade unions in Australia’. Economic and Industrial Democracy, 30: 1, 93–119.Cordery, J. and Parker, S. (2007). ‘Work organization’, in P. Boxall, J. Purcell and P. Wright (eds),

    Oxford Handbook of Human Resource Management, New York: Oxford University Press.Crant, J., Kim, T. and Wang, J. (2011). ‘Dispositional antecedents of demonstration and usefulness of 

    voice behavior’.  Journal of Business Psychology, 26: 3, 285–297.Cullinane, N. and Donaghey, J. (2014). ‘Employee silence’, in A. Wilkinson, J. Donaghey, T. Dunson

    and R. Freeman (eds),  Handbook of Research on Employee Voice, Northampton, MA: Elgar.Dundon, T. and Gollan, P. (2007). ‘Re-conceptualizing voice in the non-union workplace’.  International

     Journal of Human Resource Management, 18: 17, 1182–1198.

    Edwards, P. (2003). ‘The employment relationship and the field of industrial relations’, in P. Edwards(ed.),  Industrial Relations: Theory & Practice, 2nd edn, London: Blackwell.

    Farndale, E., Ruiten, J., van Kelliher, C. and Hope-Hailey, V. (2011). ‘The influence of perceivedemployee voice on organizational commitment: an exchange perspective’.   Human Resource

     Management, 50: 1, 113–129.Foley, J. (2014). ‘Industrial democracy in the twenty-first century’, in A. Wilkinson, J. Donaghey, T.

    Dunson and R. Freeman (eds),  Handbook of Research on Employee Voice, Northampton, MA: Elgar.Freeman, R., Boxall, P. and Haynes, P. (2007).  What Workers Say: Employee Voice in the Anglo-American

    Workplace, Ithaca, NY: Cornell University Press.Freeman, R. and Kleiner, M. (2000). ‘Who benefits most from employee involvement: firms or

    workers?’  American Economic Review, 90: 2, 219–223.

    Freeman, R. and Lazear, E. (1995). ‘An economic analysis of works councils’, in J. Rogers and W.Streeck (eds),   Works Councils, Chicago, IL: University of Chicago Press.

    Bruce E. Kaufman

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 2015 37

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    20/22

    Freeman, R. and Medoff, J. (1984).  What Do Unions Do?  New York: Basic Books.Gall, G. and Dundon, T. (2013).   Global Anti-Unionism, London: Routledge.Godard, J. (2014). ‘The psychologisation of employment relations?’   Human Resource Management

     Journal, 24: 1, 1–18.Goldberg, C., Clark, M. and Henley, A. (2011). ‘Speaking up: a conceptual model of voice responses

    following the unfair treatment of others in non-union settings’.   Human Resource Management, 50:1, 75–94.

    Gollan, P., Kaufman, B., Taras, D. and Wilkinson, A. (2014).  Voice and Involvement at Work , London:Routledge.

    Gospel, H. and Pendelton, A. (2010). ‘Corporate governance and employee participation’, in A.Wilkinson, P. Gollan, M. Marchington and D. Lewin (eds),   Oxford Handbook of Participation inOrganizations, New York: Oxford University Press.

    Guest, D. (2008). ‘Worker well-being’, in P. Blyton, N. Bacon, J. Fiorito and E. Heery (eds),   Sage Handbook of Industrial Relations, London: Sage.

    Holland, P., Cooper, B., Pyman, A. and Teicher, J. (2012). ‘Trust in management: the role of employeevoice arrangements and perceived managerial opposition to unions’.  Human Resource Management

     Journal, 22: 4, 377–391.Hoption, C., Christie, A. and Barling, J. (2008). ‘Introduction’, in J. Barling and C. Cooper (eds),  Sage

     Handbook of Organizational Behavior, Vol. 1, London: Sage.Huselid, M. and Becker, B. (2011). ‘Bridging micro and macro domains: workforce differentiation and

    strategic human resource management’.  Journal of Management, 37: 2, 421–428.Katz, H. and Darbishire, O. (2000). Converging Divergences: Worldwide Changes in Employment Systems,

    Ithaca, NY: Cornell University Press.Kaufman, B. (1999). ‘Expanding the behavioral foundations of labour economics’.   Industrial and

    Labour Relations Review, 52: 3, 361–392.Kaufman, B. (2004). ‘Employment relations and the employment relations system: a guide to

    theorizing’, in B. Kaufman (ed.),  Theoretical Perspectives on Work and the Employment Relationship,Champaign, IL: Industrial Relations Research Association.

    Kaufman, B. (2005). ‘The social welfare objectives and ethical principles of industrial relations’, in J.Budd and J. Scoville (eds),  The Ethics of Human Resources and Industrial Relations, Champaign, IL:Labour and Employment Relations Association.

    Kaufman, B. (2010). ‘The theoretical foundation of industrial relations’.  Industrial and Labor RelationsReview, 64: 1, 74–108.

    Kaufman, B. (2012). ‘An institutional economic analysis of labour unions’. Industrial Relations, 51: S1,438–471.

    Kaufman, B. (2013). ‘The optimal level of market competition: neoclassical and newinstitutional conclusions critiqued and reformulated’.   Journal of Economic Issues, 47: 3, 639–672.

    Kaufman, B. (2014a). ‘Employee voice before Hirschman: its early history, conceptualization, and

    practice’, in A. Wilkinson, J. Donaghey, T. Dunson and R. Freeman (eds),  Handbook of Research onEmployee Voice, Northampton, MA: Elgar.

    Kaufman, B. (2014b). ‘Explaining breadth and depth of employee voice across firms: a voice factordemand model’.   Journal of Labor Research, 35: 3, 296–319.

    Kaufman, B. (2014c). ‘The future of employee voice in the USA: predictions from an employmentrelations model of voice’, in S. Johnstone and P. Ackers (eds),   Finding a Voice? EmployeeRepresentation in the New Workplace, Oxford: Oxford University Press.

    Kaufman, B. (2014d). ‘Market competition, HRM and firm performance: the conventional paradigmcritiqued and reformulated’.  Human Resource Management Review, forthcoming.

    Kaufman, B., Lewin, D. and Adams, R. (1995). ‘Workforce governance’, in G. Ferris, S. Rosen and D.Barnum (eds),  Handbook of Human Resources Management, New York: Blackwell.

    Kaufman, B. and Miller, B. (2011). ‘The firm’s choice of HRM practices: economics meets strategic

    human resource management’.  Industrial and Labor Relations Review, 64: 3, 526–557.

    Integrative model of employee voice

    HUMAN RESOURCE MANAGEMENT JOURNAL, VOL 25 NO 1, 201538

    © 2014 John Wiley & Sons Ltd.

  • 8/18/2019 Kaufman-2015-Human Resource Management Journal

    21/22

    Klaas, B., Olson-Buchanan, J. and Ward, A.-K. (2012). ‘The determinants of alternative forms of workplace voice: an integrative perspective’.  Journal of Management, 38: 1, 314–345.

    Kornhauser, A., Dubin, R. and Ross, A. (1954).   Industrial Conflict, New York: McGraw-Hill.Lawler, E. (1986).   High-Involvement Management, San Francisco, CA: Jossey-Bass.Leiserson, W. (1922). ‘Constitutional government in American industry’.  American Economic Review,

    12: 1, 56–79.Lepak, D., Bartol, K. and Erhardt, N. (2005). ‘A contingency framework for the delivery of HR

    practices’.  Human Resource Management Review, 15: 2, 139–159.Lewin, D. (2014). ‘Individual voice: grievance and other procedures’, in A. Wilkinson, J. Donaghey,

    T. Dunson and R. Freeman (eds),  Handbook of Research on Employee Voice, Northampton, MA: Elgar.Luchak, A. (2003). ‘What kind of voice do loyal employees use?’   Industrial Relations, 41: 1, 115–134.Marchington, M. and Suter, J. (2013). ‘Where informality really matters: patterns of employee

    involvement and participation (EIP) in a non-union firm’.   Industrial Relations, 52: Suppl. S1,284–313.

    Marginson, P., Edwards, P., Edwards, T., Ferner, A. and Tregaskis, O. (2010). ‘Employeerepresentation and consultative voice in multinational companies operating in Britain’.   British

     Journal of Industrial Relations, 48: 1, 151–180.Marks, A. and Chillas, S. (2014). ‘Labour process perspectives on employee voice’, in A. Wilkinson,

     J. Donaghey, T. Dundon and R. Freeman (eds),   Handbook on Research on Employee Voice,Northampton, MA: Elgar.

    Marsden, D. (1999).  A Theory of Employment Systems, Oxford: Oxford University Press.Marsden, D. and Cañibano, A. (2010). ‘An economic perspective on employee participation’, in A.

    Wilkinson, P. Gollan, M. Marchington and D. Lewin (eds),   Oxford Handbook of Participation inOrganizations, New York: Oxford University Press.

    Marshall, A. (1890).  Principles of Economics, London: Macmillan.Miller, G. (1991).   Managerial Dilemmas, New York: Cambridge University Press.Morrison, E. (2011). ‘Employee voice behavior: integration and directions for future research’.

     Academy of Management Annals, 5: 1, 373–412.

    Morrison, E. (2014). ‘Employee voice and silence’.   Annual Review of Organizational Psychology andOrganizational Behavior, 1: 1, 173–197.

    Morrison, E., Wheeler-Smith, S. and Kamdar, D. (2011). ‘Speaking up in groups: a cross-level studyof group voice climate and voice’.  Journal of Applied Psychology, 96: 1, 183–191.

    Mowbray, W., Wilkinson, A. and Tse, H. (2014). ‘An integrative review of employee voice: identifyinga common conceptualization and research agenda’.   International Journal of Management Reviews,DOI: 10.1111/ijmr.12045.

    Ng, T. and Feldman, D. (2012). ‘Employee voice behavior: a meta-analytic test of the conservation of resources framework’.  Journal of Organizational Behavior, 33: 2, 216–234.

    Paauwe, J., Guest, D. and Wright, P. (2013).  HRM & Performance: Achievements & Challenges, New York:Wiley.

    Pohler, D. and Luchak, A. (2014). ‘The missing employee in employee voice research’, in A.Wilkinson, J. Donaghey, T. Dundon and R. Freeman (eds),  Handbook of Research on Employee Voice,Northampton, MA: Elgar.

    Pyman, A., Holland, P., Teicher, J. and Cooper, B. (2010). ‘Industrial relations climate, employee voiceand managerial attitudes to unions: an Australian study’.  British Journal of Industrial Relations, 48:1, 460–480.

    Riordan, C., Vandenber