Bridging the work governance divide: Pluralism and performance
Funding information: This study did not receive funding from any specific grant provided by public, commercial or not-for-profit sectors.
Abstract
This article introduces a new direction of studies that looks at the Workplace of the Future through enlarged interdisciplinary lenses. This article bridges the divide between different traditions – human resource management, industrial relations and economic democracy – arguing theoretically and demonstrating empirically their complementarity in designing a workplace that better responds to the challenges posed by increasingly conflicting demands, such as flexibility and safeguards; quality of life and quality of productive output; diversity of workforce; and diversity of contexts and institutions. We hypothesise and find (using the largest available dataset on work-related systems CRANET) that the response may lie in adopting ‘pluralist’ work governance models, blending mechanisms that differ in kind and are recommended by different disciplinary traditions: ‘highly powered incentives’, worker representation and voice rights, communitarian practices and formal personnel systems. We also show that pluralist systems in the ‘private governance’ sphere are complementary with strong labour-related institutions in the ‘public governance’ of work; and indicate how those findings can lead to a better understanding of how to take into account institutions, such as governments, regulations and industries, in shaping and designing the workplace of the future. In particular, the finding that a pluralist configuration consistently outperforms other configurations across industries and countries suggests that a pluralist trait may be a universally valid attribute of workplace governance, even though the specific mechanisms that create that variety are different and context-dependent.
INTRODUCTION
The relationship between individuals and organisations, once at the heart of organisation and management, is regaining centrality (Jooss, McDonnell, & Conroy, 2021; Magrizos et al., 2022). After a couple of decades in which work lost weight in both economic reality and general theoretical discussions, the value of work and the threats to a sustainable working life are again recognised as a core issue (Howe, Hentrup, & Menges, 2021; Kochan, 2015). Is management prepared to support the redesign, in theory and practice, of the systems that govern the relationships between people and firms under current and foreseeable future conditions? In fact, there are signs that the state of the art could be improved.
Among the major issues and challenges facing labour service providers are rising poverty and inequality (Atkinson, 2015), increasing risk exposure as a result of corporate decisions and actions over which they have no voice (Kochan, 2002), and job insecurity due to ‘atypical contracts’ that reduce investment in the training of future human resources (Blair, 1999), to name but a few. Arguably, many of these problems stem from a lack of worker voice and bargaining power (Kochan, 2015). Indeed, the power of trade unions has notoriously and markedly declined, while other possible mechanisms for representing workers' knowledge and interests have not sufficiently compensated. In fact, the institutional representation of work in the ‘private governance’ of enterprises remains relatively rare – as our empirical results confirm – and many countries are characterised by weak ‘public governance’ (Hardy & Ariyawansa, 2019), in other words, ‘weak labour related institutions’ (Cuevas-Valenzuela, 2015).
Does the field of human resource management (HRM hereafter) address these problems? Only to a limited extent, we submit. In fact, the dominant concern has been to increase ‘flexibility’ (Bal & Izak, 2021) and motivation, either through social mechanisms of communitarian identity building or economic benefits through ‘highly powered incentives’ (Roberts, 2004). Further extensions considered the provision of social benefits and protection designed by corporate management through social responsibility and welfare programmes (Barrena-Martínez, López-Fernández, & Romero-Fernández, 2017, 2019). However, the set of practices considered, despite being expanded and very large (Dyer & Reeves, 1995), seems to be typically ‘truncated’ in a certain direction: notably strong participatory devices and democratic governance. Strong voice-giving participatory practices are often excluded, such as institutional and unionised worker representation and property rights diffusion (Lavelle, Gunnigle, & McDonnell, 2010). Representation and bargaining devices have instead been a core concern in fields such as industrial relations (IR) (Kaufman, 2010), work organisation (Trist, 1981), economic democracy (Dubb, 2016) and organisational justice (Folger & Greenberg, 1985). In fact, as Morley et al. (1996) observed, these traditions have become increasingly divorced from HRM: ‘The distinction between the fields of study of HRM and industrial relations has been heightened, primarily because of their perceived incompatibility’ (p. 646). But are they really incompatible? The main point of this paper is to argue theoretically and demonstrate empirically that the main mechanisms considered in these divorced traditions are linked by remarkable complementarities, arguably thanks to compensatory effects.
There are also some methodological reasons for revisiting the repertory of human resource systems. In recent comprehensive reviews, we find observations such as, ‘From the analysis of the literature it follows that there is no consensus about the practices that should be integrated into the human resource system’ (Martin-Alcazar, Romero-Fernandez, & Sánchez-Gardey, 2005, p. 648). Boon, Den Hartog, & Lepak (2019) lament the lack of understanding of ‘… how these systems are distinct in terms of the practices they include or exclude, how the selected practices help achieve the system's goal, and why these systems would have distinct effects on outcomes’ (p. 2499). Similarly, Laursen & Foss (2014) note, ‘In spite of the prominence in HRM and innovation streams of research of thinking on the clustering of practices (Ennen & Richter, 2010; Kaufman, 2008) there is still little theorising that predicts exactly which HRM practices bundle and why, and little empirical work that examines this issue’ (p. 522).
To address these issues theoretically, we develop a conceptual framework reconstructing a typology of work governance mechanisms. The main reason for constructing such an extended framework can be illustrated as follows. Recognising the criticality of the issue, a recent report commissioned by the International Labor Office (ILO) (Hardy & Ariyawansa, 2019) states, ‘Governance is central to the world of work. At the national level, it has traditionally been pursued through a combination of laws and regulations; voluntary agreements; labour market institutions; and the interaction of governments with employers’ and workers' organisations'. The report focused mainly on ‘public governance’ and urged a broadening of the view from ‘regulation’ to ‘governance’, acknowledging that ‘private governance’ in enterprises is an important component of overall governance. The present paper calls for a parallel broadening of the ‘private governance’ sphere from ‘the management of resources' intended as people to a broader governance of work relations. In fact, in the modern world (and in law), people are not exactly resources, but actors with rights over their human resources, with whom agreements and contracts must be established (Kaufman, 2010; Rousseau, 2004). It is the actors' relationship with the system in which they participate that can be governed or managed. In the spirit of bridging different streams of research on work and human resources, we will refer to all work-related practices and systems as ‘mechanisms for the governance of work relations'.
This extended governance approach also has the theoretical advantage of establishing a natural link with governance theory. To the extent that all the systems considered are governance mechanisms regulating labour relations, the available general typologies of governance mechanisms (e.g., Ouchi, 1980; Williamson, 1996) can be applied to reveal their complementarity or substitutability relations (Grandori & Furnari, 2008), as well as the complementarity between ‘private’ and ‘public’ governance (Hardy & Ariyawansa, 2019; Masten, 2022). The latter is represented in the empirical analysis proposed here by the available indices of the strength of labour-related institutions.
To explore these issues empirically, we use the CRANET database, the world's largest network of cross-country comparative data on human resource policies and practices, providing access to a sample of 6,688 firm-level observations on human resource practices in 20 industries in 35 countries. The database is also sufficiently broad in terms of the types of practices considered: it includes data not only on the classic set of systems considered in the HRM tradition (types of reward systems, evaluation and development, recruitment and promotion), but also the practices mostly considered in the IR tradition, such as unionisation, employment contracts, allocation of property rights, institutional democracy representation, as well as socially responsible practices aimed at all employees (such as work-life balance initiatives and welfare plans), and in particular, at vulnerable or disadvantaged groups. Thus, the empirical study offers greater breadth, both in terms of the practices and contexts considered, than most empirical analyses of HR configurations, including those using the CRANET database, that have typically selected particular subsets of systems (e.g., types of employment contracts), firms (e.g., MNC) or countries (e.g., European) (Gooderham & Nordhaug, 2011; Lazarova, Morley, & Tyson, 2008).
The remainder of the paper is divided into two main parts. The first part reconstructs an extended set of work-related practices, bridging different repertories and traditions, organised in a conceptual typology of work governance mechanisms. The second part presents an empirical analysis of a set of practices selected from the CRANET database as representative of the different classes of governance mechanisms, exploring how they combine into salient configurations, and how these configurations rank on various performance dimensions and vary across countries and industries.
WORK-RELATED SYSTEMS AS GOVERNANCE MECHANISMS
Attempts to reconstruct the set of systems comprised in the management domain that define HRM in the broadest, most inclusive way (e.g., Dyer & Reeves, 1995) have identified about 30 systems. They are usually grouped into functional classes such as staffing and performance appraisal (e.g., professional/formal practices for selection and promotion), reward systems (including pay for performance), work systems (including job design techniques and participatory devices such as quality control and teamwork), and employee relations (communication exchange, opinion surveys).
This type of classification is conducive to the evaluation and design of systems in terms of coverage, articulation, clarity and consistency (see Boon, Den Hartog, & Lepak, 2019; Delmotte, De Winne, & Sels, 2012), and the assessment of the complementarity of sub-functions, especially in the sense that they provide useful or even necessary inputs to each other. This type of synergy explains well the finding that jointly considering (and eventually modifying) a larger set of practices covering more functions is more effective than operating on smaller sets of practices (Ichniowski, Shaw, & Prennushi, 1997; Milgrom & Roberts, 1995). Utilising an insightful categorisation of types of complementarities by Horgan & Mühlau (2006), we can however observe that the above functional relations generate what they call flanking complementarities; but that there are two other possible major types of complementarities: reinforcing and compensatory. Systems or practices may be similar in kind, with complementarity arising from mutual reinforcement, resulting in ‘coherent’ sets of elements in the sense they are all of the same type. Alternatively, systems and practices may be dissimilar, in which case, complementarity may arise by compensating for the negative effects possibly generated by each of them. It has also been observed that the presence of compensatory complementarities increases the comparative effectiveness of mixed or hybrid combinations of mechanisms (Grandori & Furnari, 2008). On the basis of those arguments, we shall develop the conjecture that configurations including a more varied mix of mechanisms than usually envisaged may be effective and applied.
Table 1 outlines the conceptual framework and the three main steps leading to that core hypothesis. As a first step, Table 1 lists a wide set of mechanisms, including those mainly considered in management perspectives (left column) and in other traditions in the economics and sociology of work (right column). As a second step, the framework links the mechanisms to the main outcomes expected or pursued: firm performance through factor productivity, labour flexibility, worker motivation and effort in HRM systems; distributive and procedural justice, worker protection and worker voice and power in industrial relations and related perspectives. Third, Table 1 highlights which type of governance mechanisms the systems on the ‘two sides’ predominantly embody. The reasons are illustrated hereafter.
| Work governance mechanisms predominantly considered in human resource management | Work governance mechanisms predominantly considered in industrial relations and related perspectives |
|---|---|
- Pay for performance at different levels of individual, group and organisational outcomes (e.g., Ichniowski, Shaw, & Prennushi, 1997) - Flexible labour contracts and outsourcing (e.g., Sun, Aryee, & Law, 2007 ) - Transparent and technically advanced HRM systems in evaluation, reward, mobility and development (Koch & McGrath, 1996), professionally managed by dedicated and specialised HR units (Costa & Giannecchini, 2013; Lepak & Snell, 1999) - Horizontal communication, teamwork and team voice, knowledge sharing systems (e.g., Della Torre, 2018; Lepak & Snell, 2002) |
- Empowerment and job enrichment (Trist, 1981) - Employee ownership, employee representation on boards and committees (Dubb, 2016) - Organisational democracy ( Kaufman, 2000 , 2010 ; Harrison & Freeman, 2004) - Union voice and industrial relations systems (Della Torre, 2018) - Corporate social responsibility (Bansal, 2003) and organisational welfare practices ( Whitener, 2001 ) - Organisational justice procedures (e.g., voice-giving, right of appeal, transparent systems of rules and laws) applied to HR systems ( Folger & Greenberg, 1985 ) |
| Main expected/pursued outcomes | |
| Firm performance | Distributive and procedural justice |
| Flexibility of labour | Protection of labour |
| Worker motivation | Worker voice |
| Main kinds of mechanism | |
| Bureaucratic, communitarian, mercatistic | Democratic |
It has been noted that many classic appraisal and promotion systems are oriented toward increasing bureaucratic control, formalisation and professionalisation, while other practices are oriented toward team and community building (Budd & Colvin, 2008). It has also been noted that still other practices, such as pay for performance, are a form of priced exchange, and can be seen as practices that ‘infuse markets’ into organisations (Laursen & Mahnke, 2001; Zenger & Hesterly, 1997). Then, these mechanisms can be said to be ‘different in kind’ (Williamson, 2005) in their mode of governing work relations: some mechanisms are price-like, some are communitarian, some are bureaucratic (Ouchi, 1980). This reasoning would disentangle and considered ‘dissimilar’ - because they embody different governance mechanisms - practices that are usually grouped together and considered ‘similar’ - because they perform the same function or serve the same purpose. For example, the popular devices aimed at increasing ‘flexibility’ actually encompass systems that are different in terms of governance. Quantitative flexibility in the total workforce, through more external and temporary employment contracts (Berkery et al., 2017), are mercatistic work governance mechanisms. Systems aimed at achieving functional flexibility through administered task assignment (Piore, 2002) or flexible work schedules and locations (Spreitzer, Cameron, & Garrett, 2017) use hierarchical coordination and programs, hence pertaining to the ‘bureaucracy’ repertory. Other practices, also usually aimed at increasing flexibility, refer to the increased knowledge sharing, polyvalence and mutual support offered by clan-like and communitarian work organisations (Bal & Izak, 2021; Lindkvist, 2005). Another example is that of systems designed to induce ‘high commitment’ (Ceylan, 2013; Yang & Arthur, 2021). These range from mutualistic no-layoff and internal promotion policies, to teamwork and team-building practices, to high-powered incentives (Galia & Legros, 2005; Laursen & Mahnke, 2001).
The classic types of ‘mercatistic’, ‘bureaucratic’ and ‘communitarian’ mechanisms (Ouchi, 1980) therefore seem to be applicable to assess the similarity/dissimilarity between work governance mechanisms. Still, other relevant systems go beyond these three types, especially when the relations to be governed are work relations. In fact, Ouchi's (1980) threefold typology has already been extended to include democratic mechanisms that are able to govern different interests/goals under conditions where markets, bureaucracies and clans/communities cannot (Grandori, 2016). The issue of conflicting interests is arguably of paramount importance in the governance of work, but remarkably neglected in the HRM tradition; yet a core concern in other classic traditions, such as industrial relations, labour relations, organisational justice, economic and organisational democracy and social responsibility. The mechanisms or systems at the centre of these perspectives include procedurally fair voice-giving systems (Cropanzano, Bowen, & Gilliland, 2007; Folger & Greenberg, 1985), fair contract negotiation processes (Greenberg et al., 2004; Rousseau, 2004), protection of diversity and equal opportunity (Barrena-Martínez, López-Fernández, & Romero-Fernández, 2019), fair bargaining over surplus sharing and other aspects of the employment relation (Aoki, 1984; Kochan & Katz, 1988), stronger participatory devices, such as union and institutional democratic representation (Lavelle, Gunnigle, & McDonnell, 2010; Morley et al., 1996), and devices for diffusing economic property rights, such as employee ownership (Dubb, 2016). These mechanisms are democratic in kind because they derive from the rights of people at work, they recognise the diversity of interests and protect the diversity of knowledge and mindsets, thus distinguished from communitarian mechanisms that aim to build a homogeneous community.
Therefore, at least four general main types of governance mechanisms – mercatistic, bureaucratic, communitarian and democratic – are applicable and will be used in the following empirical analysis to classify the systems used to govern work relations in enterprises. Such a conceptual framework can and will be used to derive research questions and hypotheses about the possible governance configurations that we can empirically detect.
Are all mechanisms actually used and in what proportion? When/where? Are there any incompatibility/substitutability effects? Among which mechanisms? What are the characteristics of the combinations (configurations) of mechanisms used in a single enterprise? Are they ‘coherent’ packages of mechanisms, all of the same kind, supported by reinforcing complementarities, (i.e. system can either drive people's behaviours through incentives, or rely on rules and controls of conformity, or leverage identification and a sense of community), as often hypothesised? Or are there effective configurations that combine different types of mechanisms thanks to compensatory complementarities? How do different configurations of ‘private governance’ within enterprises combine with ‘public governance’ of work relations in different countries? Are strong labour-related institutions complementary to strong multi-mechanism private governance in enterprises? Or are they substitutes?
- Compensatory complementarities should be of particular importance in the governance of work relations. Indeed, although the problem of ‘industrial conflict’ in the field of HRM is rather remote, it is likely that different interests in the regulation of the relationship between persons and organisations will be pronounced, and that any given system or practice will have negative consequences for one party or the other. Therefore, compensatory complementarities are likely to be particularly important. For example, pay-for-performance systems have been shown to increase performance only if (or to a greater extent when) combined with high discretion, autonomy, job enrichment, quality circles and suggestion systems (Roberts, 2004). Similarly, the ‘flexibilization of internal labour markets’ (Piore, 2002) appears to be able to sustain performance when combined with democratic control over mobility and task assignments through collective/voice representation mechanisms (e.g., information disclosure, briefing groups, newsletters, trade union recognition and collective employee representation) (Della Torre, 2018).
- The strength of both ‘private’ and ‘public’ governance of work (Hardy & Ariyawansa, 2019) can therefore be conceptualised and measured along two characteristics: the multiplicity of mechanisms, and the presence of labour protection and representation systems, here defined as democratic mechanisms. Although they can in principle be functional substitutes, the presence of strong labour-related public institutions is likely to be an incentive, if not an obligation, to develop strong labour-related private institutions. After all, in governance theory, private agreements and the internal governance of economic relations within firms (‘private ordering’) have been seen as a ‘continuation’ of the ‘public ordering’ of laws and courts (Williamson, 1993). Hence, our conjecture is that multi-mechanism configurations, in the course of being high-performing for complementarities among the mechanisms, should be complementary to (sustained by) ‘evoluted’ public governance contexts.
- It has also been argued and found that more evolving, unstable, innovative and uncertain contexts/industries are populated by more knowledge-intensive workers and would therefore require more horizontal, inclusive and representative governance arrangements (Grandori, 2016; Martin-Alcazar, Romero-Fernandez, & Sánchez-Gardey, 2005). Hence, we also conjecture that innovative industries may call more intensively than other contexts for configurations with significant democratic and communitarian elements.
WORK GOVERNANCE AND PERFORMANCE AROUND THE WORLD: AN EMPIRICAL ANALYSIS OF SALIENT CONFIGURATIONS
Sample and measures
The database we used is CRANET, one of the most important in the field. The 2014 edition we used comprises 6,688 firm-level observations on human resource management systems in 20 industries and 35 countries, including private, public, mixed capital and non-profit organisations, also used in many other recent studies in this area (Gooderham & Nordhaug, 2011; Lazarova, Morley, & Tyson, 2008). The responses are provided by the organisations' directors of human resource management in an assisted mode and are always based on actual systems in place rather than opinions. The questionnaire also includes sections that provide elements on industry, performance and geographic location of the headquarters. The questionnaire is broad enough to find data on the adoption of practices belonging to all four main categories reconstructed in the previous section and to explore how they combine in each enterprise to relate the emerging configurations to performance. In addition, the information on industries and countries allows exploring how these configurations correlate with contextual factors that represent important theoretical antecedents of those arrangements, such as the uncertainty of the task-environment (Martin-Alcazar, Romero-Fernandez, & Sánchez-Gardey, 2005), and the strength of national labour-related institutions (García-Cabrera, Lucia-Casademunt, & Cuéllar-Molina, 2018; Lazarova, Morley, & Tyson, 2008).
The intensity of application of each type of governance mechanism is measured by an index that counts the number of individual practices in each category (0 if absent, 1 if present), then normalised to a range of values between 0 and 1, according to the formula (since each index is composed of a different number of practices).
The practices included in the questionnaire, which can be considered as proxies for the presence of the four main kinds of governance mechanisms, are summarised in Table A.1 in Appendix, and briefly illustrated here.
- The presence of pay-for-performance systems. The index is the presence of four types of incentives for each employee category (managerial, professional, clerical): individual performance-related pay, bonus based on individual, team and organisational goals/performance.
- The presence of flexible work contracts is a qualitatively different instrument, not necessarily co-varying, which also represents the ‘marketisation’ of labour relations. The index is the sum of the presence of different types of flexible work arrangements: annual hours contract, part-time work, flexitime, temporary/casual, fixed-term contracts, teleworking and compressed working week.
- Specialisation is measured by the presence of the following elements: the existence of an HR department with primary responsibility for major HR policy decisions, the use professional, specialised selection methods for each employee category (managerial, professional and clerical), the evaluation of training needs ex-ante and training effectiveness ex-post and the adoption of codified professional career management methods.
- Formalisation is measured by the existence of formal written rules and policies for the evaluation of the different categories of employees, and the existence of formal written strategies and policies for recruitment, training and development.
Communitarian governance is unfortunately not well captured by the questionnaire, as it does not include significant information on knowledge-related communitarian governance (nor on informal cultural community-building practices). Instead it monitors socially responsible firm-level practices; and can also be partitioned in two sub-classes: classic strategic CSR – whether or not there are provisions for social responsibility toward employees at the macro constitutional level (e.g., in the statutes and corporate charter); and more micro, targeted social inclusion programmes (for disadvantaged, vulnerable and minority groups such as ethnic minorities, older and younger workers, people with disabilities, women, low-skilled workers) and welfare programs (e.g., workplace childcare, career breaks, parental leave, pension schemes, education/training breaks, private health care). These systems are used as proxies for communitarian governance, and we take into account in interpreting the results that they measure only part of communitarian governance.
Democratic governance can be assessed through the questionnaire items in its two main classic components of institutional and union democracy. Institutional democracy is captured by the presence of employee representatives on the board of directors, the right of employees to be informed about corporate strategy, financial performance and work organisation, the use of profit sharing, stock options and other devices for sharing residual reward rights and the presence of peer and subordinate appraisal systems. Union democracy is measured by positive answers to the following questions: Are trade unions present? Do the trade unions have any influence in the organisation? Are trade unions recognised as parties in collective bargaining? Are there joint consultative committees or works councils?
The list of selected items in the questionnaire used as indicators of the presence of the mechanisms belonging to the eight subclasses is presented in Table A1 in the Appendix.
As for the contextual variables, each unit is classified by sector using 20 typical classes of the national statistical classification of economic activities, which exist in almost all countries and are matched to the country by the location of the headquarters. We divided industries into innovative vs mature, and countries into developed vs developing.
The allocation to innovative vs mature sectors is based on the OECD report ‘Science, Technology and Industry Scoreboard 2011’, which considers the top 10 innovative industries. The enterprises included in CRANET belong to seven innovative sectors (e.g., computer manufacturing, electronic products, electrical equipment, telecommunications, IT and other information services, etc.) and 13 mature sectors (e.g., agriculture, hunting, forestry, fishing, mining and quarrying, metal products, machinery and equipment, transport and public utilities). For the allocation to developed vs developing countries, we used the 2017 OECD income classification index. According to this criterion, 16 countries are classified as developed and 19 as developing. Examples of the former include Denmark, Finland and France, while the latter include the Philippines, South Africa and Turkey, among others. This operationalisation implies that, when interpreting the results, we have to take into account that developed economies, in the course of having generally more developed and articulated labour-related institutions, are also populated by firms of larger average size (3,273 employees vs 2,204 in developing countries).
All industries and countries have a sufficient number of observations, ranging from 1% to 7% of the total sample of 6,688 records. The average values of the eight governance variables in mature and innovative sectors and in developed/developing countries are shown in Table 2.
| Total | Developed countries | Developing countries | Innovative industries | Mature industries | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean | SD | N | Mean | SD | N | Mean | SD | N | Mean | SD | N | Mean | SD | N | |
| Market1_PayPerf | 0.46 | 0.33 | 6,284 | 0.46 | 0.32 | 3,165 | 0.45 | 0.33 | 3,119 | 0.56 | 0.31 | 1788 | 0.43 | 0.32 | 4,137 |
| Market2_FlexContra | 0.49 | 0.29 | 6,352 | 0.61 | 0.25 | 3,212 | 0.36 | 0.28 | 3,140 | 0.49 | 0.30 | 1804 | 0.48 | 0.29 | 4,175 |
| Burea1_HrSpec | 0.47 | 0.22 | 6,656 | 0.49 | 0.21 | 3,435 | 0.45 | 0.23 | 3,221 | 0.51 | 0.21 | 1831 | 0.46 | 0.21 | 4,267 |
| Burea2_HrFormal | 0.57 | 0.28 | 6,639 | 0.58 | 0.26 | 3,425 | 0.56 | 0.29 | 3,214 | 0.62 | 0.27 | 1830 | 0.55 | 0.28 | 4,267 |
| Social1_CSR | 0.62 | 0.36 | 6,136 | 0.69 | 0.33 | 3,044 | 0.54 | 0.37 | 3,092 | 0.66 | 0.35 | 1776 | 0.60 | 0.36 | 4,105 |
| Social2_Inclusionandwelfare | 0.44 | 0.26 | 6,678 | 0.46 | 0.26 | 3,454 | 0.41 | 0.25 | 3,224 | 0.47 | 0.26 | 1829 | 0.43 | 0.25 | 4,266 |
| Democ1_Institutional | 0.39 | 0.24 | 6,680 | 0.44 | 0.23 | 3,454 | 0.33 | 0.23 | 3,226 | 0.45 | 0.24 | 1831 | 0.38 | 0.23 | 4,266 |
| Democ2_Union | 0.51 | 0.35 | 6,220 | 0.57 | 0.34 | 3,091 | 0.45 | 0.34 | 3,129 | 0.47 | 0.35 | 1799 | 0.53 | 0.34 | 4,184 |
The questionnaire includes several performance measures, so that a multidimensional operationalisation of performance, which is desirable in principle (Barney, 2020), is possible. Unfortunately, only two parameters are measured objectively and independently, namely absenteeism and turnover. On the positive side, these are also the more job-specific performance parameters. The other parameters included in the questionnaire are service quality, productivity, profitability, innovation, stock market and environment, measured by comparative subjective judgments on a five-point Likert scale of the type: ’Compared to other organisations in your sector, how would you rate the performance of your organisation’. To control for possible common method biases, we examined the correlations between these two objective indicators and the others, which turned out to be quite high. More importantly, the emerging configurations rank approximately in the same order on all the performance parameters, both objectively and subjectively measured. This result strengthens confidence in the validity of both performance measures and the construction of configurations.
Analysis: Configurations, performances and contexts
Following an approach consistently found in the literature (Boon, Den Hartog, & Lepak, 2019), we applied a hierarchical cluster analysis using the Euclidean distance metric and the centroid linkage function with all the normalised indicators. We truncated the analysis at four clusters, since other analyses considering a larger number of clusters yielded less clear results. The adoption of four clusters generated the maximum internal homogeneity within groups and external heterogeneity between groups according to our data distribution. The profiles of these four clusters are reported and ‘labelled’ in Table 3. Two configurations have lower average intensity of all practices except one class of mechanisms: union negotiations in the configuration called ‘Simple Unionised’, and macro strategic statutory CSR in the configuration called ‘Responsible Bureaucracy’ (because it also ranks high in HR system formalisation). Configurations 3 and 4 are characterised by higher ‘multi-modality’ in governance, as they exhibit higher than average intensity of all practices: Configuration 3 is particularly high in the bureaucratic traits of formalised systems and specialised departments, coupled with significant ‘infusions’ of mercatistic governance – hence labelled ‘Market infused Bureaucracy’. ‘Pluralistic Governance’ is fully balanced across the four types of mechanisms, including institutional democratic governance devices.
| Cluster | ||||
|---|---|---|---|---|
| 1 | 2 | 3 | 4 | |
| ‘Simple unionised’ | ‘Responsible bureaucracy’ | ‘Market infused bureaucracy’ | ‘Pluralistic governance’ | |
| Market1_PayPerf | 0.25 | 0.48 | 0.49 | 0.66 |
| Market2_FlexContra | 0.35 | 0.55 | 0.45 | 0.58 |
| Burea1_HrSpec | 0.25 | 0.41 | 0.57 | 0.68 |
| Burea2_HrFormal | 0.29 | 0.62 | 0.60 | 0.85 |
| Social1_CSRmacro | 0.13 | 0.68 | 0.19 | 0.85 |
| Social2_CSRmicro | 0.26 | 0.45 | 0.40 | 0.62 |
| Democ1_Institutional | 0.20 | 0.42 | 0.38 | 0.58 |
| Democ2_Union | 0.41 | 0.46 | 0.50 | 0.58 |
| Total of observations | 1,096 | 1,058 | 1,275 | 1,611 |
Table 4 shows how the four configurations perform on the six indicators available in the questionnaire. The judgmental indicators of firm performance behave in the same way as the objective and work-specific indicators of performance (absenteeism and turnover) - with only one exception, discussed below, affecting only the relative positioning of the two low-performing configurations. The main result is clear: performance increases with the average intensity of use of all mechanisms, and a multi-modal, pluralistic governance model, in which all classes of mechanisms are significantly represented, is the best performing configuration.
| Configurations | Quality | Productivity | Profitability | Innovation | Stock evaluation | Environment | Absenteeism (days) | Annual turnover % (leave) |
|---|---|---|---|---|---|---|---|---|
| 1 – Simple Unionised | 3.87 (1,029) [0.214] |
3.56 (961) [0.206] |
3.28 (856) [0.198] |
3.24 (920) [0.201] |
2.72 (215) [0.128] |
3.32 (742) [0.183] |
10.38 (626) [0.2] |
11.19 (772) [0.198] |
| 2 - Responsible Bureaucracy | 3.95 (1,081) [0.225] |
3.67 (1,060) [0.227] |
3.45 (973) [0.225] |
3.42 (1,043) [0.228] |
3.13 (409) [0.244] |
3.53 (936) [0.231] |
12.46 (675) [0.216] |
12.66 (863) [0.222] |
| 3 – Market infused Bureaucracy | 4.10 (1,258) [0.262] |
3.77 (1,219) [0.261] |
3.49 (1,137) [0.263] |
3.56 (1,202) [0.262] |
3.26 (329) [0.196] |
3.59 (1,015) [0.251] |
9.63 (845) [0.270] |
11.01 (1046) [0.269] |
| 4 – Pluralistic Governance | 4.22 (1,433) [0.298] |
3.92 (1,424) [0.305] |
3.75 (1,352) [0.313] |
3.85 (1,409) [0.308] |
3.61 (720) [0.430] |
3.89 (1,345) [0.333] |
8.33 (974) [0.312] |
10.26 (1205) [0.31] |
- Notes: Weighted means on 5-point Likert scales; higher weighted means indicate better performance (the opposite for absenteeism and turnover); absolute number of observations in parentheses; column percentages in brackets; analysis considering all observations (total sample).
We also tested the relationships between the configurations and performance indices through regression analyses, with converging results (available upon request): the more pluralistic governance configurations (3 and 4) are associated with higher performance (higher subjective measures, and lower levels of absenteeism and turnover). The general implication of this pattern is that a balanced, diversified portfolio of systems in which practices derived from the classic HRM repertory are applied together with mechanisms infusing democratic representation and justice may be a general, universal trait of good work and human resource governance. As hypothesised, compensatory complementarities based on the dissimilarity of practices are important, and apparently more important than reinforcing complementarities among mechanisms of the same type.
However, some distinctions are in order. If the two lower performing configurations are compared among themselves, the ‘Responsible Bureaucracy’ can be ranked as superior to the ‘Simple Unionised’ on firm performance, but inferior on work efficiency and job satisfaction parameters (absenteeism and turnover) – not surprisingly. If the two lower-performing configurations are compared with high-performing configurations, their inferiority can be attributed to the application of fewer types of mechanisms, and in a partial way. In particular, in the ‘Simple Unionized’ case, there is a partial application of democratic governance, limited to collective bargaining, while all other mechanisms are not significantly present. In the Responsible Bureaucracy case, there is a partial application of communitarian governance limited at the macro-institutional level (CSR statutory practices) combined with a partial investment in bureaucratic governance through high system formalisation. Those results tentatively suggest that some combinations, limited to some subsets of mechanisms, may not produce either reinforcing or compensatory complementarities.
Other interesting qualifications can be added by also considering the frequency of each configuration in different contexts – developed vs developing countries and innovative vs mature sectors (Tables 5 and 6).
| Configurations | Developed countries | Developing countries | Total |
|---|---|---|---|
| 1- Simple Unionised | 305 [0.135] (0.27) |
823 [0.293] (0.729) |
1,128 [0.223] |
| 2- Responsible Bureaucracy | 624 [0.277] (0.549) |
512 [0.182] (0.45) |
1,136 [0.224] |
| 3- Market infused Bureaucracy | 521 [0.231] (0.398) |
786 [0.28] (0.601) |
1,307 [0.258] |
| 4- Pluralistic Governance | 798 [0.355] (0.539) |
682 [0.243] (0.460) |
1,480 [0.293] |
| Total | 2,248 (0.445) |
2,803 (0.554) |
5,051 |
| Chi-square (χ2) | 253.82**** | ||
- Notes: Percentages in brackets, total by column = 100; percentages in parentheses, total by row = 100;
- **** p < 0.001.
| Combinations | Innovative industries | Mature industries | Total |
|---|---|---|---|
| 1- Simple Unionised | 248 [0.164] (0.231) |
822 [0.244] (0.768) |
1,070 [0.219] |
| 2- Responsible Bureaucracy | 321 [0.213] (0.295) |
767 [0.228] (0.704) |
1,088 [0.223] |
| 3- Market infused Bureaucracy | 369 [0.245] (0.290) |
900 [0.267] (0.709) |
1,269 [0.260] |
| 4- Pluralistic Governance | 568 [0.377] (0.394) |
873 [0.259] (0.605) |
1,441 [0.296] |
| Total | 1,506 (0.309) |
3,362 (0.69) |
4,868 |
| Chi-square (χ2) | 81.75**** | ||
- Notes: Percentages in brackets, total by column = 100; percentages in parentheses, total by row = 100;
- **** p < 0.001.
The ‘Simple Unionised’ configuration is relatively more diffused in more mature industries and less developed countries, where, as we know from history, work relations used to be regulated by sheer bargaining power. Interpreted theoretically, this result signals that there may be some tensions, or a lack of reinforcing complementarity, between the two main types of democratic mechanisms considered - unionised bargaining and institutional representation - in the lack of flanking labour-related institutions in the public sphere. In other terms, a developed ‘public governance’ disciplining the matters to be regulated by bargaining, by institutional representation or by the law itself may be necessary for making the two types of democratic governance mechanisms more complementary.
The other configuration, which is relatively more common in developing economies, is the ‘Market infused Bureaucracy’, in which all bureaucratic mechanisms are used intensively and a significant dose of marketisation has been infused through incentives and atypical labour contracts. This pattern may suggest that HR-dominated governance and union-based governance may represent two possible alternative arrangements for governing relations with workers in a self-administered mode where the national economic and institutional context does not provide solid external support for the regulation of labour relations.
It is also noteworthy that ‘Market infused Bureaucracy’ performs relatively well, and is diffused in all types of industries, including innovative industries. These findings are consistent with configurational research on HR systems in innovative settings (Laursen & Mahnke, 2001) and support the thesis that market infusions via ‘highly powered incentives’ in the internal organisation is possible, but effective only when accompanied by professional measurement, control and performance evaluation systems of the bureaucratic type. In this case, complementarity seems to arise from a symbiotic ‘flanking’ relationship: one device (incentives) needs the inputs of a device that differs in kind (performance evaluation) to produce its output. In addition, compensatory complementarities can also play a role: whenever mercatistic practices are intensively adopted, they do so in conjunction with other ‘counterbalancing’ and compensatory systems, especially to increase transparency and control through the formalisation of HR and finally (in pluralistic models), also through the representation of work.
Finally, it is noteworthy that no configuration dominated by mercatistic mechanisms has emerged in our empirical analysis. This contrasts with the anecdotal observation that the contemporary governance of work is sailing towards generalised mercatisation in HRM. Our study shows that an intensive application of mercatistic mechanisms is found only in conjunction with other mechanisms that could mitigate/balance their effects (i.e., in market-infused bureaucracies, in pluralist governance configurations). Read prescriptively, our results suggest that general recommendations to adopt mercatistic devices, such as pay for performance or flexible work contracts, in isolation from other compensatory mechanisms may be particularly dangerous and may not lead to the intended performance gains.
The pattern of diffusion of the ‘multimodal’, ‘pluralistic’ governance configuration, which is relatively more represented in developed countries and in innovative industries, only confirms the promise of this model in times when these contexts are gaining ground in the general economic landscape.
CONCLUSIONS
The identified complementarities and relationships between configurations and performance have implications for better workplace design and for enriching the human resource system toolkit. The perspective offered here is complementary to and different from a (more common) predictive approach that asks what the trends will be, how technology or new practices will affect the nature of work, how motivation will vary, etc. (Barley, Bechky, & Milliken, 2017). The presented study focuses on ‘the governance side’ of the workplace, aiming to guide the design of work-related systems as governance systems, showing how to prescriptively interpret the identified variations of work governance configurations. Those conclusions can be expressed by articulating two relatively neglected principles for designing work systems: pluralism (to temper the disadvantages of coherence); and universality (to limit unfair contingency), as illustrated next.
Pluralism and coherence
A pluralistic governance model, characterised by a significant and balanced representation of all types of mechanisms (bureaucratic, mercatistic, communitarian and democratic), emerges as the best-performing configuration across countries and industries. It is therefore advisable to consider a set of mechanisms that is larger than the usual HRM set when designing the workplace of the future. In particular, we provided theoretical arguments and empirical evidence or claiming that systems drawn from HRM are compatible and can be usefully complemented by systems highlighted in the fields of industrial relations, economic democracy and social responsibility.
This tenet is different and refines former configurational studies on HR practices, mainly arguing that the larger the set, the better (e.g., Ichniowski, Shaw, & Prennushi, 1997). We suggest that effective configurations are not necessarily characterised by ‘many’ practices. What makes a difference in performance is, above all, the diversity in kind of the mechanisms in the mix. Therefore, specifying what the relevant ‘kinds’ or classes of elements is a crucial step for future studies – although the ‘kinds’ may be expanded. For example, technological coordination mechanisms and digital relations are becoming increasingly important and can be considered a further ‘kind’ of coordination mechanism (Chambefort & Chaudey, 2024).
The direct implication is to warn against the common approach to designing HR systems or other aspects of organisations according to a principle of ‘coherence’, if interpreted as ‘similarity in kind’ among the elements included in a configuration. Such an approach is likely to overlook the complementarities generated precisely by the different forces and counterbalances they bring to a system. Overall, our findings suggest that compensatory complementarities based on the dissimilarity of governance mechanisms appear more important than reinforcing complementarities between mechanisms of the same kind.
Recent studies on the workplace of the future support the claim that if the application of practices or systems of the same kind at high level of intensity may have negative consequences, even if the practice is intended to produce positive effects and it does so up to a point. Magrizos et al. (2022) demonstrate that excessive meaningfulness in work arrangements can negatively impact job stress, particularly for individuals with high levels of workaholism. Antonacopoulou & Georgiadou (2021) suggest that the work-from-anywhere approach may contribute to feelings of social isolation, where trust-based employment relationships and socially connected workforces (Kulik, 2022) alongside a caring organisational culture (Saks & Gruman, 2024), may serve as key barriers to the emergence of this issue. As in medicine or cooking, the main challenge is to find the right mix of ingredients, not to choose one type for the sake of coherence.
Universality and contingency
Contributing to the long-lasting and empirically inconclusive debate on the ‘convergence’ or ‘divergence’ of governance and organisation forms, and of HRM systems in particular (Gooderham & Nordhaug, 2011; Lazarova, Morley, & Tyson, 2008), our study reveals a kind of ‘universality’ that can be seen only by moving from the specific definition of practices to the higher level of the governance principles they embody. Then some unifying principles and homogeneity can be seen behind the diversity of solutions. Regardless of which specific practice is used to perform a governance function (incentivising, regulating, including, representing), our findings suggest that superior arrangements are characterised by a mix of these. Thus, there may be more homogeneity at a higher-order level than is usually recognised. In other terms, pluralism in kind of governance mechanisms can be considered a universally valid trait of work governance configuration, beyond the contingent difference of the specific practices applied, for responding to the diversity of relevant demands.
A second source of universality is that there are reasons to believe that when organising involves people and not just tasks, a higher dose of universal principles may be required (Pfeffer, 1994). Regardless of the context, things like human rights, non-discrimination policies and fairness criteria should be always respected, i.e., the possible consequences (even and especially when unintended) for those rights should be considered when designing any system. For example, Barber & Santuzzi (2015) found that ‘telepressure’ - defined as the preoccupation with and urgency to respond quickly to messages from clients, coworkers or supervisors - can be linked to burnout, both physical and cognitive, as well as absenteeism and diminished sleep quality among workers. Similarly, the implementation of surveillance software and always-on webcams to monitor employees working from home, clearly raises concerns linked to privacy laws (Harwell, 2020). Hence, no matter how different those practices are, they should all safeguard common universal principles.
Research tools deserve some reform for supporting the above two enrichments of perspective. First, surveys and data collection instruments intended to be comprehensive should be enlarged, both for not missing significant complementarities, and for being better equipped to solve labour problems, including strong participatory devises. In CRANET, in particular, the systems currently underrepresented in the questionnaire are those related to communitarian governance, including informal and cultural community building, as well as all knowledge-based components of communitarian governance (such as knowledge management systems). A new category of digital and technological mechanisms, as already observed, may be usefully added for capturing the recent transformations.
Second, fairness, so important in the governance of work relations, should be more present at various levels. Performance indicators can be enlarged toward including direct measures of fairness and equity outcomes. An even more important a challenging implication is to implement the universal human rights principle that people should not be framed as things, or objects to be managed, but as persons or actors holding human resources and providing labour services based on them or investing (part of) them under some agreement with organisations. Hence, a more radical implication of the approach advocated here is that the unit of analysis should be the ‘work relation’ rather than a set of ‘human resources’. If so, data should not come solely from employers' representatives, but utilise data from the different parties in the relation – in other terms, research procedures on work-related matters may also benefit from a higher dose of pluralism.
AUTHOR CONTRIBUTIONS
Gustavo Magalhães de Oliveira: Conceptualisation (equal), Formal Analysis, Writing – Original Draft Preparation (lead), Writing – Review & Editing (lead).
Maria Sylvia Macchione Saes: Conceptualisation (equal), Writing – Original Draft Preparation (supporting), Writing – Review & Editing (supporting).
Wilson Aparecido Costa de Amorim: Data Curation.
Anna Grandori: Conceptualisation (equal), Writing – Original Draft Preparation (lead), Writing – Review & Editing (lead).
ACKNOWLEDGMENTS
We are indebted and very grateful to Editor Anthony McDonnell and two EMR anonymous reviewers for their very helpful comments and precise suggestions during the review process, including some further works in the area of work governance and practice complementarities which have been key in strengthening the conceptual framework and connecting to pertinent important discourses. We are thankful to the CRANET Research Network for the use of the CRANET database obtained with the participation of the Brazilian partner (School of Economics, Business, Accounting and Actuarial Science at the University of São Paulo) in this article. We acknowledge that the manuscript has benefitted from professional language editing services offered by Jacqueline Fuchs. Maria Sylvia Saes acknowledges financial support from National Council for Scientific and Technological Development (CNPq) for the grants n. 304233/2023-4. Open Access funding enabled and organized by Projekt DEAL.
CONFLICT OF INTEREST STATEMENT
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
ETHICAL INFORMATION
This study analyses data from a survey conducted by the CRANET Research Network. Access to the CRANET database was facilitated by one of the co-authors, who is also a member of the network.
APPENDIX
| Type of governance mechanism | Sub-classes | CRANET database questions used to measure the mechanisms |
|---|---|---|
| Mercatistic | Pay-for-performance systems | Question: Do you offer any of the following (yes/no) for each category of employee (managerial, professional, and clerical)? Sum of yes responses:
|
| Flexible work contracts | Question: ‘Please indicate the approximate proportion of those employed by your organisation who are on the following work arrangements’. Sum of yes responses:
|
|
| Bureaucratic | Specialisation of HR functions | Question ‘Who has primary responsibility for major policy decisions on the following issues?’. Sum of responses identifying ‘the HR Department’ (from a list of 4) as the primary entity responsible:
Question ‘Please indicate which of the following selection methods are used in your organisation?’ Sum of ticked as present:
Question: ‘Do you systematically estimate the need for training of personnel in your organisation?’; ‘Do you systematically evaluate the effectiveness of training of personnel in your organisation?’. Sum of yes responses. |
| Formalisation of HR Systems | Question ‘Do you have a formal appraisal system for the following categories of employees?’ Sum of yes responses:
Question ‘Do you use the following to deliver HRM activities?’ Sum of yes responses:
Question ‘Does your organisation have a written…’ Sum of yes responses:
|
|
| Communitarian | Macro CSR: strategic statements | Question ‘Does your organisation have a written…’ Sum of yes responses:
|
| Micro CSR: social inclusion and welfare programs | Question ‘Do you offer any of the following schemes in excess of statutory requirements?’ Sum of yes responses:
Question for all types of employees (recruitment, training, career progression): ‘Does your organisation have action programs covering any of the following groups to improve their participation in the workforce…:’ Sum of yes responses:
Question ‘Do you use the following to deliver HRM activities?’ Sum of yes responses: - Human resource information system (HRIS) or electronic HRM system Question ‘Does your organisation have a written…’ Sum of yes responses:
|
|
| Democratic | Institutional democracy | Question ‘Does the person responsible for HR have a place on the board or equivalent top executive team? Sum of yes responses. Question ‘Which employee categories are formally briefed about the following issues?’ Sum of yes responses for the clerical category:
Question ‘Do you offer any of the following...’ Sum of yes responses for all types of employees - managers, professionals, clericals):
Question ‘If you have an appraisal system, who is formally expected to provide input data for the appraisal process?’. Sum of yes responses indicating employees/sub/peers as source of input (rather than immediate supervisor or supervisor's superior) for all types of employees - managers, professionals, clericals):
|
| Union democracy | Question ‘What proportion of the total number of employees in your organisation are members of a trade union? (Please round up to the nearest full percentage)’ (transformed into dummy variable: yes if a positive percentage is indicated). Question ‘To what extent (0 to 4 grades) do trade unions influence your organisation?’ (transformed into dummy variable: yes if a positive extent is indicated). Question ‘Do you recognise trade unions for the purpose of collective bargaining?’ Dummy variable (yes/no) Question ‘Do you use the following methods to communicate major issues to your employees?’ Dummy variable (yes/no)
|
Biographies
Gustavo Magalhães de Oliveira is Postdoctoral Researcher at the Institute for Food and Resource Economics (ILR), University of Bonn, Germany. He holds a Ph.D. in Management (Organizational Economics) from the University of São Paulo (USP), along with a M.Sc. from the same university. During his Ph.D., Gustavo worked as a visiting researcher at the University of Perugia, Italy. His research interests are focused on the application of quantitative methods to explore organisational and institutional issues in management and economics.
Maria Sylvia Macchione Saes is Professor of Economics and Business Strategy in the Department of Business at the University of Sao Paulo. She is also a director of the Center for Organization Studies. Additionally, she served as the editor-in-chief of the RAUSP - Management Journal. Her main areas of publication focus on the agri-food sector and value chain and firm strategy.
Wilson Aparecido Costa de Amorim is Professor of Human Resource Management in the Department of Business at the University of Sao Paulo. He is also the Main Manager of the Human Resource Management Department of the University of Sao Paulo. His main areas of publication focus are on the human resource management, labour relations and the labour market.
Anna Grandori is Professor of Business Organization at Bocconi University, Milan, Italy, an Honorary Member of EGOS and a EURAM Fellow. She has been an editor of various international journals in Organization and Management. Her approach integrates elements of organisational, economic, behavioural sciences and recently law to improve decision-making, organisation design and firm governance.
Open Research
DATA AVAILABILITY STATEMENT
Data not available due to privacy/ethical restrictions.
