We must consider how our definition of 'representative' has broadened and the system has evolved. While the founders of America defined 'representative' very narrowly--including primarily the white male landowners or property owners--our understanding of the term today encompasses all people, including ethnic minorities and women.--Harry Kranz 1976, xiii
The United States is becoming increasingly racially and ethnically diverse (see Appendix A). From 2000 to 2010, the total U.S. population grew by 9.7 percent, from 281.4 million in 2000 to 308.7 million in 2010. The "Asian alone" census group experienced the fastest rate of growth, while the "Black" population increased at a faster rate than the total population. More than half of the growth in the total population of the United States was due to the increase in the Hispanic population. Conversely, the "White alone" population experienced the slowest rate of growth; yet, the proportion of Whites who reported more than one race did grow by 37 percent (2010 Census Quickfacts on Race). In comparison, the racial demographics of the federal workforce are not proportionally representative of the U.S. general population (see Appendix B). This disparity raises several important normative and practical issues surrounding representation for diversity management and governance, more generally.
With these demographic changes, attitudes and approaches toward representation are likewise shifting. Public administration scholarship and practice have the ability to contribute to this dynamic process of defining representation, diversity, and crafting diversity initiatives to meet the needs of the public. As such, President Obama issued Executive Order 13583: Establishing a Coordinated Government-wide Initiative to Promote Diversity and Inclusion in the Federal Workforce on August 18, 2011. This research seeks to answer the following question: how is diversity being defined and what are the implications for a representative bureaucracy in practice? To answer these questions, a critical discourse analysis of Executive Orders 13078, 13163, 13171, 13518, 13548, and 13585 is performed. Prime emphasis is devoted to the most recent and comprehensive documents, Executive Order 13583 and the subsequent Government-wide Diversity and Inclusion Strategic Plan.
A fundamental reason for promoting diverse representation in public organizations is that a representative workforce is more reflective of the society it serves, making representation an intrinsic good. Additionally, a view of representative bureaucracy as an instrumental good may promote representation in the form of positive policy outcomes when active representation benefits the population served. For representative bureaucracy to be meaningful in both scholarship and practice, a clearer understanding of diversity in the context of bureaucratic representation and its potential for fulfilling public purposes must be undertaken. A deeper understanding of the value of diversity and bureaucratic representation would shape both the intrinsic and the instrumental justifications for bureaucratic representation. The research reported here uncovers difficulties inherent in moving toward a more meaningful treatment of "diversity" by demonstrating the reliance on vague and contentious rationales for promoting diversity. Ambiguous articulations of what both representation and diversity is and why diverse representation is purported to achieve public purposes pose challenges for management and governance.
This article first provides a review of relevant representative bureaucracy scholarship. Next, an explanation of the research design, focusing specifically on the approach of critical discourse analysis is presented. Then, the central findings of the analysis are detailed. Finally, the implications for management and governance, specifically the challenges and new avenues for promoting diverse representation are highlighted.
REVIEW OF REPRESENTATIVE BUREAUCRACY LITERATURE
This section presents some of the foundational works and concepts that contribute to the representative bureaucracy literature in relation to normative ends of managing diversity and diverse governance. From Kingsley's (1944) original articulation of a normative framework for understanding changing demographics in the British Civil Service, discussions of representative bureaucracy over the last several decades have fluctuated from a set of normative arguments and recommendations for how bureaucrats could or should represent the populations they serve to empirical investigations of passive and active representation and bureaucratic role perceptions. Most early scholarship on representative bureaucracy focused on the interorganizational characteristics of representation (Kingsley 1944; Long 1952; Van Riper 1958). More recently, scholars have begun to acknowledge the impact of socialization and organizational culture on representative bureaucracy (Meier 1975, 1993; Meier and Stewart 1992; Meier & Nigro 1976; Saidel & Loscocco 2005; Selden 1997, 2006; Sowa & Selden 2003; Selden, Brudney & Kellough 1998).
The term "representative bureaucracy" originated with J. Donald Kingsley (1944) in Representative Bureaucracy: An Interpretation of the British Civil Service. He examines how the British Civil Service was becoming more consistent with changing socioeconomic trends to reflect the "dominant forces" of society. It is important to emphasize that Kingsley (1944) did not advocate a bureaucracy that reflected all of society, but only of the most prominent groups in terms of socioeconomic status. Like later scholars (Long 1952; Van Riper 1958; Mosher 1968; Meier 1975; Sowa & Selden 2003), Kingsley (1944) believed representative bureaucracy serves as a stabilizing source of discretionary control. He explained, "The degree to which all democratic institutions are representative is a matter of prime significance. No group can safely be entrusted with power who do not themselves mirror the dominant forces in society; for they will then act in an irresponsible manner or will be liable to corruption" (1944, pp. 282-283). Kingsley's (1944) initial analysis laid the groundwork for generating other normative approaches to understanding and analyzing representative bureaucracy.
Norton Long (1952), in "Bureaucracy and Constitutionalism," built on Kingsley's initial analysis, claiming that representative bureaucracy is desirable to promote important interests that are not embodied by elected office holders (members of Congress and the president of the United States). Long (1952) states: "given the seemingly inevitable growth in the power of bureaucracy through administrative discretion and administrative law, it is of critical importance that the bureaucracy be both representative and democratic in composition and ethos" (p. 813). Paul Van Riper (1958) takes the theory even further in a normative direction by identifying different characteristics in his framework for a representative bureaucracy that are meant to reflect the citizenry, including social traits and values such as occupation, class, geography, ethos, and attitudes (p. 527). He argues, "There is a minimal distinction between the bureaucrats as a group and their administrative behavior and practices on the one hand, and the community or societal membership and its administrative behavior, practices, and expectations of government on the other" (Van Riper 1958, 552). In what Meier (1975) calls "the most comprehensive analysis of representative bureaucracy," Mosher (1968) makes a fundamental distinction that has dominated the scholarship; namely, the identification of two types of representation: "active" and "passive." He (1968) argues, "administrative decisions are a function of administrative capabilities, orientations, and values, which in turn depend on bureaucrats' backgrounds, training, education, and current association" (as cited in Meier, 1975, p. 527).
Active and Passive Representation
Mosher (1968) questions how to create a bureaucratic system that is consistent with democratic principles, and in doing so he formulates these two distinct types of representation. Passive representation takes place when the demographics of a bureaucracy are similar to those of the citizens they serve. Alternatively, active representation occurs when bureaucrats act on behalf of the interests of the groups who share their own demographic characteristics. Mosher (1968) originally thought of active representation as a negative consequence and advocated for a solely passive model of bureaucracy. The desirability of active representation has shifted with more recent normative theories and empirical studies proposing that passive representation in the form of a diverse bureaucracy should lead to active representation, or policy outcomes reflecting the interests of a diverse group of citizens (Krislov 1974; Meier and Stewart 1992; Meier 1993; Selden & Selden 2001). Samuel Krislov (1974) emphasizes the value of active representation: "Human potentialities brought by bureaucrats to their jobs are inevitable and advantageous . . . The qualities of judgment, information, and fervor that bureaucrats do bring as they aid decision-makers are in fact resources of immense social advantage" (p. 81).
If active representation is a desirable end, the question of what characteristic should be represented remains. Beginning in the 1970s, theorists informed by the changing political climate in the United States began to focus on dimensions of representation beyond socioeconomic status. Scholarship on representative bureaucracy turned to race, sex, and ethnicity as more pertinent measures of "just" representation and positive contributors of new forms of diversity in public agencies (Krislov 1974; Meier 1975; Meier and Nigro 1976; Kranz 1976). Centered on these indicators of representation, promoting active representation by emphasizing...