From the Politics of Bureaucracy to the politics of representative bureaucracy

Introduction: Bringing society back in

The history of the Politics of Bureaucracy is deeply rooted in an era of profound social and political transformation. Guy Peters vividly describes how the initial ideas for the book were conceived amid heightened social tensions, which eventually manifested themselves in urban riots in the United States and beyond. Much of this outburst of political energy was also translated into social movements (promoting emancipatory causes such as civil liberties and women’s lib movements or rallying behind peace and environmental protection) that left their distinct marks on the political development of Western democracies – and apparently on the deep insights of the Politics of Bureaucracy, too. It flows from this that the nexus between state and society is a core element of Guy Peters’ breakthrough book. Taking this view to its logical conclusion, we cannot understand public organisations unless we understand how much they are embedded in their social and political context, including their historical backgrounds and cultural traditions. Consequently, it counts among the greatest achievements of the Politics of Bureaucracy to have successfully shifted the ‘publicness’ of public administration to centre stage – against strong competition from other lines of thought and practice (ranging from the scientific management tradition to the rational-planning discourse and the new public management movement).

Against this background, the concept of representative bureaucracy can be interpreted as an effort to bring society back into the study of public administration. This strand of research is primarily interested in the social composition of the public sector workforce, how it relates to the overall make-up of society and what difference possible disparities make for the quality and legitimacy of public service delivery (cf. Andrews et al., 2005; Dolan and Rosenbloom, 2003; Groeneveld and Walle, 2010; Peters et al., 2015c; Selden et al., 1998; Selden and Selden, 2001; Von Maravic et al., 2013). The term had first been coined in critical discussions of class-based dominance in British politics and administration (Kingsley, 1944, 2003), but it is no coincidence that the academic discussion finally took root in the US context of the late 1960s and early 1970s (Krislov, 1974; Meier, 1975; Meier and Nigro, 1976; Mosher, 1968; Subramaniam, 1967) – and has been covered in the Politics of Bureaucracy since its first edition. With increasing social disparities and growing ethnic and cultural diversity characterising most contemporary societies, the interaction between public administration and society goes through a major transformation. There is a growing concern that established response patterns of public sector reform (such as the fine-tuning of bureaucratic rules and structures to adjust to external changes or efficiency-driven reforms to address austerity problems) cannot adequately cope with those new societal challenges. This situation gives rise to the question if and how public bureaucracies can serve as (politically and socially) representative institutions in their own right by promoting the ‘representativeness’ of their workforce. Indeed, we may have arrived at a point in social and political developments (not unlike the situation in which the plan of the breakthrough book was conceived) at which ‘technical’ or ‘managerial’ interventions with the administrative machinery tend to miss the essence of the problem load. Rather, the intensity of social tensions and political polarisation requires – well aligned with Guy Peters’ approach – a much broader understanding of public organisations as creatures of society and politics.

In what follows, the argument unfolds in three steps. First, the article takes a closer look at the concept of representative bureaucracy as a paradigm for public sector reform, before the attention will shift to the desired outcomes of representativeness in public organisations as well as to potential drawbacks and pitfalls of this reform approach. Finally, the rationale of the Politics of Bureaucracy will be utilised to draft a research agenda for this important sub-field of comparative public administration. The following reflections on representative bureaucracy are deeply informed by close collaboration with the author of the breakthrough book (Peters et al., 2015a, 2015b, 2015c; Peters and Schröter, 2018; Von Maravic et al., 2013). It goes without saying that all omissions, deficits and misunderstandings are entirely my own.

Representative bureaucracy: Definition and theoretical background

The term ‘representative bureaucracy’ refers to the relationship between the demographic composition of the public sector workforce (particularly with regard to gender, ethnicity and cultural traits such as language and religion) and the society it is supposed to serve (Van der Meer and Roborgh, 1996). In very general terms, ‘representative bureaucracy’ can be described as the ‘body of thought and research examining the potential for government agencies to act as representative political institutions if their personnel are drawn from all sectors of society’ (Dolan and Rosenbloom, 2003: xi). To operationalise this definition, however, the social roots and context of representation have to be recognised. Both in theory and practice, it is not ‘all sectors of society’ that receive academic and/or political attention. Rather, the concept of representation tends to focus on specific cleavage lines in society, the saliency of which is likely to shift over time and differ across national boundaries, depending on the social, political and economic context. In addition, the representation of social groups can take (at least) two different forms: passive and active. While passive representation is rather a formal and descriptive measure, active representation refers to the capacity of individual group members to act willingly and purposefully as representatives, trustees or advocates of the collective interest of their social group.

Just as much as the Politics of Bureaucracy revolves around questions of public sector accountability as well as social and political means of controlling public bureaucracies, the concept of representative bureaucracy can be couched in terms of a search for control over administrative power. Sparked by the concern to keep public bureaucrats responsive to the public at large (Subramaniam, 1967), the design of representative bureaucracy was meant to serve as an effective internal measure of controlling government officials’ behaviour. In this context, however, control over bureaucrats is not being exerted by contracts, performance measures or individual incentives (as advocated by the new public management) or hierarchical lines of command and legal constraints (as is the case with the Weberian bureaucratic machine). Rather, control and accountability are primarily understood as functions of group identities that are represented in public bureaucracies.

It flows from this that a strong collectivist undercurrent runs through the representative bureaucracy approach – not unlike the dominant strand of the participatory democracy literature (Pateman, 1970). This group-centred approach (based on a strong sense of belonging, identity and solidarity among group members) corresponds with an anthropological picture of (wo)man that conceptualises human nature as that of a social animal or zoon politicon, defined by a strong sense of group identity and driven by a strong impetus for communication and interaction with other members of their community. Based on this assumption, a model of representative bureaucracy will flourish best on the fertile grounds of a communitarian society – a social theory that places particularly high emphasis on the right of groups to be equally recognised and represented in society and politics. Seen from this angle, the concept of representativeness is also driven forward by the ‘politics of recognition’ and trends towards identity politics in contemporary societies (Taylor and Gutman, 1994).

Inside (public) organisations, representativeness relies heavily on high degrees of bureaucratic discretion as a precondition for ‘active representation’. In addition, this strategy for administrative change shows a close normative proximity to the findings from implementation studies, which highlight the importance of ‘street-level bureaucrats’ and support a ‘bottom-up’ perspective on public policy delivery (Lipsky, 1980). It is in this intellectual environment in which the concept of representative bureaucracy can thrive best, because it focuses on the direct service delivery by teachers, social workers or police officers (to name but a few prominent types of ‘street-level bureaucrats’) to specific communities. Therefore, empowerment appears in this perspective as conducive to foster the connection between passive and active representation as employees enjoy greater discretion and flexibility in dealing with clientele groups and the public at large. Looking at the broader picture, a distinctly more consociational, consensus-oriented and power-sharing style of political and administrative decision-making also seems to be in order for advocates of representativeness (Lijphart, 1999). Necessarily, any heroic assumptions of strong leadership; ‘experts-know-best’ attitudes; and quick, uniform and efficient (in a nutshell: non-participatory) decision-making procedures are sources of potential clashes with the worldviews of ‘representativeness’ protagonists.

In this pure, ideal-typical form, representative bureaucracy is bound to clash with rival established principles of public sector reform (cf. Pollitt and Bouckaert, 2011). At this level of theoretical abstraction, reform paradigms of administrative change can be traced back to fundamental mechanisms of social coordination: clans, markets and hierarchies (Ouchi, 1980; Williamson, 1975; cf. Hood, 2000, for an approach based on cultural theory). As Peters (2019) puts it, there is a certain logic of appropriateness (March and Olsen, 1989) for bureaucratic behaviour that is based on connections to ‘family’ or ‘clan’ – and in a broader sense, representative bureaucracy as a group- and solidarity-based approach falls into this category. In principle, this logic of representativeness clearly contrasts with both the logic of appropriateness of Weberian merit-based civil services in well-ordered hierarchies and with the logic of consequentiality of markets and individualist utility functions.

To begin with, the drift towards communitarian thinking – inherent in the concept of representative bureaucracy – does not resonate well with principles of universalism on which most of the bureaucratic institutions in liberal democracies are built. Rather than constituting specific group rights in society and guarding their equal recognition, the universalist tradition focuses on individual rights and duties that are defined regardless of any membership to social groups. In terms of bureaucratic behaviour, this fundamental perspective translates into the right of universal access to public services and the entitlement to uniform service standards. This ‘impersonality’ of Weberian bureaucracies stands in contrast to the image of active representation, which not only risks but also encourages distinct behavioural patterns relative to specific social groups. For the same reason, the principle of merit-based recruitment may potentially conflict with strategies towards representativeness if group representation is seen as paramount to formal qualifications.

In a similar vein, protagonists of market-driven social coordination (cf. Lindblom, 2002; cf. Self, 1993 for a critical review) tend to be just as far away from the core values of the representativeness school of thought. The ideal-typical market model of public management – based on the rational actor or homo oeconomicus – only accepts the methodological individualism as a legitimate yardstick to evaluate human and organisation behaviour. As a result, there will be inherent tensions between the in-built quest for efficiency and any requirement to represent social groups – and even their activists – in the organisation. Rather, the logic of consequentiality (March and Olsen, 1989) of rational actors with externally pre-set utility functions and cost-benefit calculations to guide their behaviour will take its course and potentially collide with the group-based call for representation.

Although ideal-typically defined reform paradigms appear to be incompatible or even mutually exclusive if taken to their logical conclusions, their own real-types in administrative practice may even be complementary to some extent. As Peters points out, real-life public bureaucracies in liberal democracies have found their own ways to accommodate the representation (if not incorporation) of political parties and major interest groups in their organisational structures and decision-making procedures. At the same time, we can also learn from the Politics of Bureaucracy that seemingly neutral or objective measures for recruitment, selection and career advancement can also be interpreted as examples of the mobilisation of bias to favour certain candidates at the expense of others (cf. Peters, 2015). In a similar vein, efficiency-driven organisations in practice may quickly take sides with approaches fostering social representation and equity if they realise that this reform strategy improves their performance.

Concluding from this theoretical reflection, representativeness – as a reform paradigm – is more likely to augment rather than to replace rivalling approaches if high standards of meritocratic civil services, rule of law and efficient public management are to be safeguarded. However, representative bureaucracy stands out of competing reform strategies as no other paradigm has developed any antennae sensitive to the quest for group recognition and social inclusion.

Desired outcomes and potential pitfalls of representative bureaucracy

Much of the academic interest in representative bureaucracy is propelled forward by the normative belief in the value of having public bureaucracies that mirror in their own ranks and files the salient group identities of the society they are supposed to serve. Given this set of underlying values and normative expectations, advocates of representative bureaucracy seem to be predisposed to expect diversity strategies to translate into a broad range of consequences positively associated with organisational performance (cf., for example, Bradbury and Kellough, 2008; Hindera, 1993; Keiser et al., 2002; Meier and Nicholson-Crotty, 2006; Meier and Stewart, 1992; Pitts, 2005, 2009; Riccucci, 2002; Wilkins and Keiser, 2006). From a more analytical perspective, however, a complete picture of this reform potential also includes the potential drawbacks and pitfalls of a strategy for change that promotes representativeness in the public sector (see also Schröter and Von Maravic, 2015).

Before delving into the subject matter, one has to recognise that employment in the public sector means different things to different people in different institutional contexts. For a start, public employment has a strong signalling effect as to who owns the state (Wimmer, 1997). For many public employees and civil servants, their employment is also code for upward social mobility, because public employers are often also providers of training and education for their employees. In any event, a public sector job (or even career) is not only a secure source of income but signifies also a process of social inclusion and political socialisation into the machinery of the government executive or the ‘state at work’. For public employers, to change perspectives, recruitment and selection policies for public sector jobs are not the only means to find the best match for an open position or a career entry. Rather, policy decisions about public sector hiring and employment standards are at the same time macro-economic instruments in the hands of government to pursue labour market or regional development policies and to address wider economic and social problems.

It seems to be fair to suggest that much of the academic and political debate on representativeness has been propelled forward by what can be labelled the ‘performance claim’ of representative bureaucracy. The suggested virtues of representative bureaucracy seem to fall into different, though partially overlapping, categories. To capture this multi-faceted nature of desired outcomes of representativeness, three different dimensions suggest themselves: (1) first, the organisational function relates to the degree of mission accomplishment (output) as well as to the internal workings of the organisation (throughput); (2) second, representativeness may fulfil a (macro-)economic function as part of the wider implications of public employment; and (3) third, the (party-)political and social (social-structural and socio-cultural) function of group representation in the public sector has to be recognised. To be sure, each of those functional dimensions may land on the good or bad side of social group representation in the public workforce. Accordingly, the following sections briefly review the intended and unintended consequences of representativeness.

The politics of representative bureaucracy: Towards a research agenda

The Politics of Bureaucracy not only provides an important impetus for the research of representative bureaucracy but also serves as an analytical frame for a research agenda in this field of inquiry. The major impetus comes from one of the book’s core messages that public administration is tightly interwoven with politics and society. Administrative organisations are shaped by societal and political developments. Conversely, the institutional set-up of public administration, including the structure and culture of their workforce, significantly affects the way we lead our political and social life. As a key contribution to the study of comparative public administration, the Politics of Bureaucracy suggests a genuinely comparative perspective for future research on representative bureaucracies. In this context, comparisons are important along at least two dimensions. First, the organisational and task-specific context seems to matter for the functioning of representative bureaucracy. The existing body of research still very much focuses on a small segment of the total universe of public sector organisations. Mostly, local service organisations with close proximity to social communities and relatively high degrees of professional discretion seem to fall into this well-researched category. In sharp contrast, public organisations at different levels of government or with different mission statements have received much less scholarly attention. This omission also includes non-state providers (such as charitable organisations, voluntary welfare associations and self-help groups, and also for-profit providers) of public services, which play an increasingly important role in service provision (Peters and Schröter, 2018). Second, distinct national institutional environments, shaped by state traditions and administrative cultures, will leave their marks on emerging patterns of representative bureaucracy. Representative bureaucracy is not just the result of normative commitment or of existing social conditions. Rather, it is embedded (just like any other public sector paradigm) in complex social, political and administrative systems – a key insight that we owe in no small measure to the author of Politics of Bureaucracy.