Reconsidering Sex-Based Occupational Segregation and Agency Missions: An Analysis of U.S. State Bureaucracies (1987-2015) Using Two Different Thresholds

Abstract

We examine Equal Employment Opportunity Commission data (1987-2015) to determine whether sex-based occupational segregation among administrative and professional workforces is related to state agency policy missions. Based on two thresholds, the customary 30% benchmark and the 50% benchmark of parity, the findings indicate segregation is related to policy missions. The 30% benchmark suggests a story of widespread progress across state bureaucracies. The 50% benchmark suggests less progress, especially, in police, fire, corrections, utilities, natural resources, and highways. The authors argue it is important to use multiple indicators to assess the progress of women in the workplace.

Keywords

glass walls gender politics occupational segregation women’s employment sex-based parity state bureaucracies

Introduction

We’ve ceased making progress at the top in any industry anywhere in the world . . . In the United States, women have had 14 percent of the top corporate jobs and 17 percent of the board seats for 10 years. Ten years of no progress. In those same 10 years, women are getting more and more of the graduate degrees, more and more of the undergraduate degrees, and it’s translating into more women in entry-level jobs, even more women in lower-level management. But there’s absolutely been no progress at the top. You can’t explain away 10 years. Ten years of no progress is no progress.

—Sandberg (2013), NPR transcript

Glass walls and glass ceilings are metaphors used to describe the invisible barrier beside and above women seeking greater responsibilities and/or promotion within organizational hierarchies (Powell, 2012). Although researchers have conducted large-scale studies of the existence of both types of barriers in public sector employment in federal and municipal government workforces (among others, see Choi, 2018; Dolan, 2004; Kellough, 1989, 1990; Lewis & Emmert, 1986; Miller, Kerr, & Reid, 1999; Naff, 1994; Riccucci, 2009; Smith & Monaghan, 2013), there are fewer empirical studies examining the presence of glass walls and/or ceilings in state-level bureaucracies (Bullard & Wright, 1993; Kerr, Miller, & Reid, 2002).

We revisit and extend the analysis done by Kerr et al. (2002) by adding 18 years of data—for an overall total of 28 years, 1987-2015—to determine (a) the extent of sex-based occupational segregation among managerial-level workforces in U.S. state-level bureaucracies and (b) whether sex-based occupational segregation is related to agency policy missions. Researchers typically examine occupational segregation by employing a cut-off of 30%. If women hold less than 30% of managerial positions, the assumption is that glass walls exist (Kerr et al., 2002; Miller et al., 1999; Reid, Kerr, & Miller, 2003; Reid, Kerr, & Miller, 2000; Tomaskovic-Devey, 1993; Tomaskovic-Devey, Kalleberg, & Marsden, 1996). In addition to examining managerial-level occupational segregation using the customary threshold of 30% women, we also analyze the data using a 50% threshold. The latter threshold represents absolute or true sex-based parity.

Significance of the Research

Because we are interested in women’s representation in public sector ranks that are managerial, positions offering both employment security and policy-making influence/power, we limit our analysis to the top two (of eight) categories listed on the Equal Employment Opportunity Commission’s (EEOC’s) EEO-4 Form, officials/administrators and professionals. Table 1 provides descriptions of the responsibilities and duties of administrators and professionals. In state bureaucracies, administrator and professional positions offer women the greatest opportunity for financial rewards/security as well as access to policy-making influence and power. Because our research is an examination of the existence and extent of sex-based occupational segregation by agency type, we seek to determine whether female admininstrators or professionals are concentrated in agencies with distributive, regulatory, and/or redistributive policy missions (Lowi, 1985). We examine occupational categories and agency missions at two different thresholds, 30% women and 50% women.

Table 1.

Top Two Occupational Categories from the EEOC’s Instructional Booklet for the EEO-4 Form.

Officials/administrators: Occupations in which employees set broad policies, exercise overall responsibility for execution of these policies, or direct individual departments or special phases of the agency’s operations, or provide specialized consultation on a regional, district, or area basis and include department heads, bureau chiefs, directors, deputy directors, controllers, wardens, superintendents, sheriffs, police and fire chiefs and inspectors, examiners (bank, hearing, motor vehicle, warehouse), inspectors (construction, building, safety, rent and housing, fire, A.B.C. Board, license, dairy, livestock, transportation), assessors, tax appraisers and investigators, coroners, farm managers, and kindred workers.

Professionals: Occupations that require specialized and theoretical knowledge, which is usually acquired through college training or through work experience and other training that provides comparable knowledge and include personnel and labor relations workers, social workers, doctors, psychologists, registered nurses, economists, dieticians, lawyers, systems analysts, accountants, engineers, employment and vocational rehabilitation counselors, teachers or instructors, police and fire captains and lieutenants, librarians, management analysts, airplane pilots and navigators, surveyors and mapping scientists, and kindred workers.

Source. U.S. EEOC (1986); EEOC Form 164.

Note. EEOC = Equal Employment Opportunity Commission; EEO = Equal Employment Opportunity.

The argument for using the 30% threshold is that it represents a reasonable benchmark for examining and identifying positions and agencies where women are at or moving toward equal representation (Tomaskovic-Devey, 1993). This is essentially the critical mass argument. Agencies that achieve or exceed the 30% benchmark have in place a critical mass of female managers, which in turn increases “the likelihood that (1) female managers will be retained and promoted, and (2) women will have increased opportunities for influencing policy outputs” (Kerr et al., 2002, p. 415). By extending the data used by Kerr et al. (2002) by 18 additional years, we can determine whether the share of positions held by women continues to increase through 2015 using the customary 30% threshold. We do not intend to argue that the critical mass argument is unimportant. We do, however, isolate its primary assumption—that meeting the 30% threshold means that women have achieved or are in the process of achieving equal representation—and address the political implications of the assumption. Achieving the 30% benchmark in an agency will not be viewed the same way by all observers. For observers or policy makers less than fully committed to true gender equality, the argument might be that continued efforts to promote sex-based equality in the workplace are no longer necessary because at the threshold of 30%, women glass walls have eroded to the point where sex-based occupational segregation is no longer a problem.

By examining these same data at the 50% threshold, researchers can determine where and when gender equality or true parity is achieved in quality high-level management positions across different types of state agencies. There is an advantage to evaluating more than one measure of women’s progress. How performance is measured and how data are interpreted are frequently connected to political and economic interests (Levin-Waldman, 2004). Levin-Waldman (2004) argues that “there is no absolute model that can drive the policy process . . . policy ultimately boils down to how data are interpreted and how those interpretations are put in the service of those interests that have a particular agenda” (p. 139).

As we are well into the 21st century, the 50% benchmark appears to be a defensible normative and empirical goal. A marker of true parity can provide conceptual and theoretical insights not provided by a 30% threshold. The goal of parity in governmental bureaucracies is consistent with the proportional representation of women in the U.S. population, their participation in the U.S. labor force, and their employment in managerial jobs. Thus, the 50% marker serves as a straightforward measure of descriptive representation based on a standard of true equality. If high levels of descriptive representation are consistent with healthy, legitimate democracy, setting the bar at 30% may be unfair to women. Through equal opportunity and affirmative action policies and agencies such as the U.S. EEOC and its related statutes, diversity in the public sector workforce is valued and protected through direct prohibitions of employment discrimination, often including “best practices” or avenues for inclusion. Consequently, several agencies, statutes, and regulations inform and direct public sector managers in their efforts to hire, promote, or discipline employees (Battaglio, 2015). Management decisions made contrary to the legal and regulatory proscriptions and inclusion points will likely harm diversity efforts, rendering government less responsive to its diverse constituencies and their interests (Riccucci, 2009; Riccucci & Saldivar, 2014; Rice, 2007; Selden, 1997). Although descriptive representation does not always insure substantive policy representation and/or actual participation in decision making (Mosher, 1968), existing research indicates that women and minorities in public sector leadership positions are likely to promote diversity and produce more responsive and progressive policies (Gonzalez & Denisi, 2009; Riccucci, 2009). Evaluations of governmental performance based on the 50% benchmark of absolute parity make it easier to assess the relative value a gendered society places on certain types of policies or governmental functions as well as on the inclusion of women in high-level policy-making positions.

The critical importance of this research lies not only in the benefits to women who obtain highly desirable administrative and professional positions but also to the contributing organizations as they alter (a) traditional hiring/promotion processes and (b) management and leadership styles (Kottke & Agars, 2005; O’Neil & Hopkins, 2008; Tower & Alkadry, 2008; Weyer, 2007). Choi (2010) states that an evaluation of agency-level policy types or missions can provide important predictors of workforce diversity, although many researchers have found that an increase in women in management has a unique and positive effects on policy outputs and performance (Caceres-Rodriguez, 2013; D’Agostino, 2015; Fraga, Martinez-Ebers, Lopez, & Ramirez, 2005; Lawless, 2004; Sabharwal, 2015; but see Ford & Dolan, 1999; Hogue & Lord, 2007; Rosenthal, 2005). A strong normative case can be made for why women are needed in managerial-level positions in public bureaucracies, but barriers may keep this from happening, especially in certain types of agencies. As Guy (2017) observes, often the emotive demands of jobs are related to systematic gender segregation, as some jobs highlight empathy and nurturance, whereas other jobs place a premium on courage and expressions of confidence.

Why study state bureaucracies? Distinct constitutional and legal entities, states are critical governmental units in the U.S. federal system. States have responsibility for a wide range of policy areas, including, but not limited to, agriculture, banking, business regulation, courts, corrections, economic development, education, health, highways, hospitals, law enforcement, local government structures, natural resources, occupational regulation and licensing, parks/tourism, taxation, and welfare. Given the unique nature of state governments, Osborne (1988) and Van Horn (1989) argue that some state governments are more innovative, responsive, and effective but caution that these characteristics are not apparent in all states (Herbert, Brudney, & Wright, 1999; also in Bowling, Kelleher, Jones, & Wright, 2006; Jacobson, Palus, & Bowling, 2010; Kerr et al., 2002). Furthermore, Bullard and Wright (1993) and Kerr et al. (2002) posit that because women’s activism and state-level agency activism converge, there should be expanded opportunities for women interested in state-level employment in newly formed programs and agencies. These new agencies and programs are more likely to value diversity than older, established agencies (Bowling et al., 2006; Bullard & Wright, 1993). State bureaucracies have grown at a much faster rate than have federal or local bureaucracies over the last six decades (Meier & Bohte, 2007). Due in part to these expanding opportunities, we expect to observe growth in women’s shares of managerial positions in state government, especially, in agencies that emphasize so-called women’s work (Guy, 2017; Kerr et al., 2002).

More than 160 years removed from the beginning of the women’s rights movement, U.S. society has experienced many significant milestones and events such as ratification of the 19th Amendment in 1920, passage of the Equal Pay Act of 1963, passage of the Civil Rights Act of 1964, creation of the EEOC in 1965, the decision in Meritor Savings Bank v. Vinson (1986), passage of the Lilly Ledbetter Fair Pay Act of 2009, and more recently, the #MeToo movement.¹ Progress has been slow and not always steady; however, it may not be completely unreasonable today to expect women’s shares of managerial positions in many state bureaucracies to approach true parity, 50%. Released in 2017, a Bureau of Labor Statistics report found that in 2015, women comprised 56.7% of the U.S. labor force and accounted for 52% of all workers in management, professional, and related occupations. Furthermore, the 2017 report disclosed that

[t]he share of women in specific occupations within this large category varied. For example, 18 percent of software developers, 28 percent of chief executives, and 35 percent of lawyers were women, whereas 89 percent of registered nurses, 81 percent of elementary and middle school teachers, and 60 percent of accountants and auditors were women. (U.S. Bureau of Labor Statistics, 2017, p. 2)

Alkadry and Tower (2014) contend that proper representation of women in the workplace should be at 50%, including 50% in leadership ranks. Accordingly, our research strategy employs the customary 30% glass wall threshold, the basis for one story or narrative, as well as the 50% threshold of true parity, which is the basis for a different or second narrative. We evaluate each threshold over a 28-year period to examine the extent of women’s progress and lack of progress in administrative and professional positions in U.S. state-level bureaucracies.

Kerr et al. (2002) used restricted-access EEOC data that include EEO-4 reports for 49 states for 1987, 1989, 1991, 1993, 1995, and 1997.² Restricted access data are kept from the public domain and protected by an EEOC confidentiality provision prohibiting the release of any information that permits individual states to be identified. In addition to examining these same 10 years, we extend the data set by another 18. The data set covers 28 years, 1987 through 2015. Our primary objective is to examine the distribution of men and women administrative and professional positions over time in all 49 states by functional policy area (e.g., highways, public welfare, police, health, natural resources, etc.) to ascertain if agency policy commitments are associated with sex-based occupational segregation at the 30% and 50% benchmarks.

Research Questions

We examine the following research questions:

Research Question 1: What is the distribution of male and female administrators and professionals in various functional policy areas (e.g., corrections, police, highways, natural resources/parks, public welfare) in state bureaucracies?

Research Question 2: Is the overrepresentation and/or underrepresentation of male and female administrators and professionals in functional policy areas linked to agency missions?

Literature and Theory

Much of the literature examining the share of public sector management jobs demonstrates that where and when women enter these positions varies by agency mission and type (Bullard & Wright, 1993; Choi, 2018; Connell, 2006; Johnson & Crum-Cano, 2011; Lewis & Nice, 1994; Miller et al., 1999; Newman, 1994; Reid, Kerr, & Miller, 2000; Smith & Monaghan, 2013; Sneed, 2007). This generalization is supported by analysis on all levels of government, federal, state, and local. Riccucci’s (2009) examination of federal government employment indicates that women consistently hold lower ranks than men, suggesting that women are concentrated in the lower pay grades and in certain types of agencies. Smith and Monaghan (2013) find underrepresentation of women in federal regulatory agency leadership posts and that women are expected to work in leadership positions in so-called feminine policy areas. These general patterns of representation repeat themselves in local governments, the level of government, which by far employs the greatest number of public servants. Researchers consistently find evidence for gender-based occupational segregation and pay disparities at the local level (Lewis & Nice, 1994; Reid et al., 2003).

At the state level, Bullard and Wright (1993) find that women are absent or extremely underrepresented at the highest administrative levels in oil and gas, agriculture, labor, water pollution, economic development, corrections, and law enforcement agencies. Similarly, in their study of men’s and women’s shares of positions in state and local government agencies, Lewis and Nice (1994) find that men are overrepresented in police, streets and highways, and fire departments, whereas women were heavily overrepresented in housing, health, and public welfare agencies. Choi (2018) finds that although redistributive agencies employ more women in traditionally male-dominated occupational categories, those women tend to be paid less than in other agencies. Likewise, Guy (2017) argues that the monetary penalty for holding positions associated with characteristics such as care and nurturing lowers wages for both women and men, but women bear most of the burden.

Of particular theoretical relevance to our research questions is the work of Lowi (1985), Newman (1994), and Johnson and Crum-Cano (2011). Although researchers agree that no one theory can fully explain sex-based occupational segregation and all the contributing factors (Johnson & Crum-Cano, 2011; Sneed, 2007), many researchers have used Lowi’s (1985) policy typology. Lowi (1972) posits that policy causes politics—and as such, he was able to demonstrate how power relationships can vary depending upon the type of policy under consideration. Moreover, Lowi’s (1985) overall contention is that occupational segregation within managerial levels occurs unevenly, largely dependent upon whether the agency has an overarching regulatory, distributive, or redistributive policy mission. In Newman’s (1994) study of Florida state-level bureaucracies, she argues that most women work in regulatory and redistributive agencies and that women experience the harshest discrimination in hiring and promotion practices in distributive agencies. In their work testing Lowi (1985) and Newman’s (1994) theoretical positions, Sneed’s (2007) analysis of Michigan state-level data found sex-based occupational segregation in regulatory and distributive departments but not in redistributive departments. Later, Johnson and Crum-Cano (2011), using survey data from the American Planning Association, found gender imbalance/discrimination in both distributive and regulatory agencies, but no imbalance in redistributive agencies. Kerr et al. (2002) found that high levels of occupational segregation existed among administrative-level employees in agencies with regulatory and distributive policy missions, but the professional ranks in these same agencies became less segregated over time. Little to no occupational segregation was found at either the administrative or professional levels in agencies with redistributive policy commitments (Kerr et al., 2002).

We build our hypotheses from the characteristics and functions associated with Lowi’s (1985) three policy types (see Table 2). Because agencies with distributive policy missions rely heavily on occupational and professional norms, promote specialists rather than generalists, possess relatively high levels of discretion, and are not particularly sensitive to discriminatory practices, we believe, among the managerial ranks, males will be heavily overrepresented, and there will be little if any advancement toward gender balance over time in these agencies. By contrast, agencies with redistributive policy commitments tend to hire and promote women, and as such, we expect to find, as did Johnson and Crum-Cano (2011) and Sneed (2007), that redistributive agencies are characterized by little to no evidence of glass walls among their administrative and professional workforces.

Table 2.

Policy Type by Characteristics and Functions.

Policy type	Policy type characteristics	Policy type functions^a	EEOC agency types
Distributive	Resistant to change due to mutually reinforcing relationships between clientele and their agencies; reliance on professional and occupational norms; highly specialized. As such, diversity is not valued^b	Development and management of water resources; construction and management of streets, highways and bridges; provision and operation of historical sites, parks, and recreation; and community development (see Appendix for a description of EEO-4 form functions)^c	• Highways • Natural resources • Community development
Regulatory	Impose obligations and sanctions to control individual and collective actions and behavior^d	Corrections, fire, police, labor relations, securities, utilities and transportation, energy, oil and gas, and so on (see Appendix for a description of EEO-4 form functions)^c	• Police • Fire • Utilities/transportation • Corrections
Redistributive	Shift wealth and/or rights between groups and classes^c/^d often hiring from within the agency^c	Public welfare systems, employment security, hospitals and sanitariums, provision of public health services, programs for the aging, and vocational rehabilitation^c	• Welfare • Hospitals • Health • Housing • Employment security

Note. EEOC = Equal Employment Opportunity Commission; EEO = Equal Employment Opportunity.

All functions are performed by state governments.

Corson and Paul (1966); Lowi (1985); Lewis and Emmert (1986); and Newman (1994).

Kerr, Miller, and Reid (2002) and/or EEO-4 Form 164.

Lowi (1964, 1972).

Agencies with regulatory policy missions include the regulation of criminal behavior or policing, corrections, and fire fighting. In these agencies, we expect to find an overwhelming representation of men in managerial ranks. Thus, we expect to observe high levels of occupational segregation in these agencies. This expectation comes from studies finding that within traditional policing agencies, there are a number of exclusionary practices that keep women from high-level ranks (Mladenka, 1991; Newman, 1994; Stein, 1986; Warner, Steel, & Lovrich, 1989) and as such, gender balance is not promoted in these agencies (Cayer & Sigelman, 1980; Riccucci, 1986; Warner, Steel, & Lovrich, 1989). Johnson and Crum-Cano (2011) also found sex-based occupational segregation in the regulatory and distributive agencies. In regulatory agencies other than policing, corrections, and fire, such as utilities and transportation, we do not expect to find evidence of occupational segregation. We expect greater representation of women in administrative and professional ranks in utilities and transportation because, as Kerr at al. (2002) stated: “(1) these agencies are likely to hire attorneys (see Newman, 1994); and (2) women’s enrollment in law schools has increased over the last two decades” (p. 414). More recently, a December 2016 New York Times article reports:

For the first time, women make up a majority of law students, holding just over 50 percent of the seats at accredited law schools in the United States. The number of men and women enrolled in juris doctorate programs has been nearly equal for a number of years, but this is the first time women have moved past the 50 percent mark, according to data released Thursday by the American Bar Association. Currently, 55,766 women nationwide are studying for a juris doctor degree, compared with 55,059 men, according to the bar association. First-year students are more than 51 percent women, or 19,032, and 48.6 percent men, or 18,058. (https://www.nytimes.com/2016/12/16/business/dealbook/women-majority-of-us-law-students-first-time.html)

This updated information brings us to our final queries. Will patterns of sex-based occupational segregation follow for the post-1997 period the same pattern found by Kerr et al (2002)? Furthermore, will these patterns persist if we change the benchmark from 30% to 50%, so that the benchmark is consistent with true gender parity?

Hypotheses

Our hypotheses based on the 30% threshold are as follows:

Hypothesis 1: Women will be underrepresented in administrative and professional positions in distributive agencies (i.e., women will not reach the 30% threshold).

Hypothesis 2: In administrative and professional positions in redistributive agencies, women will reach or surpass the 30% threshold.

Hypothesis 3: In regulatory agencies, we expect the highest level of occupational segregation in policing and fire.

Likewise, for the analysis employing the 50% benchmark (i.e., gender parity), we expect to find gender parity in agencies with redistributive policy commitments, but we expect to find widespread patterns of female underrepresentation in distributive agencies and regulatory agencies, with the highest levels of underrepresentation in police and fire.

Research Design: Data, Variables, and Method

Data

The EEOC provided the data for this analysis.³ The data include all EEO-4 reports from states from 1987 to 2015 and represent the most comprehensive source of U.S. state-level employment data available. With the exception of state employees in public schools (EEO-5 reports) and institutions of higher education (EEO-6 reports), EEO-4 reports include all state employees and are collected for every odd-numbered year. From these data, we are able to analyze sex-based occupational segregation by position (administrators and professionals) and by functional policy area (e.g., health services, community development, police, etc.).

Variables and Method

The purpose of our study is to determine whether functional policy areas at the state level show evidence of sex-based occupational segregation. Toward this purpose, we evaluate our hypotheses by comparing the distribution of men and women in professional and administrative positions by functional policy area for the period 1987 through 2015. Following previous research, we first employ the cut-off point of 30%. If 30% or more of the particular position’s workforce in a given functional policy area is made up of women, this is interpreted as the absence of glass walls (Miller et al., 1999; Tomaskovic-Devey, 1993; Tomaskovic-Devey et al., 1996). Then, based on the premise that equity means true or absolute parity, we conduct additional analysis using the threshold of 50% women.

Table 3 reports the number (and percentages) of administrators and professionals for 1987 and 2015 by functional area. For 2015, the most recent year for which we have data, the EEOC’s EEO-4 reports include 108,374 administrators and 721,210 professionals for a total of 829,584 management-level state employees. The number of administrative positions increased by 72% between 1987 and 2015, whereas the number of professional positions increased by 41%. Growth in the number of managerial level positions in state bureaucracy has been uneven across functional areas. Much of the increase may be explained by rapid growth in the number of state government employees as well as the concomitant devolution of programs and power from the federal government to the states.

Table 3.

Total Number and Percentages of Male and Female Administrators and Professionals for 1987 and 2015, and by Functional Area for All 50 States.

Functional area	1987					2015
	Male		Female		Total	Male		Female		Total
	n	%	n	%	n	n	%	n	%	n
Administrators
Distributive	11,147	89	1,429	11	12,576	16,399	71	6,685	29	23,084
Regulatory	7,488	83	1,510	17	8,998	10,229	61	6,559	39	16,788
Redistributive	14,674	65	8,013	35	22,687	12,115	37	20,997	63	33,112
Indeterminate	13,995	75	4,764	25	18,759	17,678	50	17,712	50	35,390
Total	47,304	75	15,716	25	63,020	56,421	52	51,953	48	108,374
Professionals
Distributive	69,710	83	14,063	17	83,773	79,912	63	46,702	37	126,614
Regulatory	44,111	69	19,954	31	64,065	54,100	46	64,263	54	118,363
Redistributive	95,732	38	157,611	62	253,343	79,634	27	218,448	73	298,082
Indeterminate	63,622	57	48,440	43	112,062	74,764	42	103,387	58	178,151
Total	273,175	53	240,068	47	513,243	288,410	40	432,800	60	721,210

Following Lowi (1964), our functional policy categories are distributive, redistributive, and regulatory. This scheme is inelegant as Lowi’s (1964) categories were initially developed for and applied to federal government policies rather than to the state and local government functional categories developed by the Equal Employment Opportunity Commission. Appendix lists responsibilities for each functional area (EEOC Form 164, 1997)

Empirical Results for the 30% Threshold

Occupational Segregation Among Administrators by Functional Policy Area and Year

With our interest in the performance of states in the area sex-based occupational segregation, first we examine the percentage of state administrative workforces in each functional area at the threshold of 30% women. Table 4 reports the percentage (and number) of states that meet or exceed the threshold by functional area and year.

Table 4.

Empirical Results for the 30% Threshold: Occupational Segregation Among Administrators by Functional Policy Area and Year.

	States meeting the 30% threshold for administrators by functional area
	1987		1991		1995		1999		2003		2007		2011		2015
Distributive
Highways—total states		47		48		48		46		47		42		44		47
Number/percent wall goal	2	4%	2	4%	1	2%	4	9%	4	9%	7	17%	13	30%	20	43%
Natural resource—total states		48		48		49		49		49		44		47		48
Number/percent wall goal	2	4%	1	2%	0	0%	3	6%	9	18%	13	30%	20	43%	26	54%
Community development—total states		32		34		32		34		33		30		35		38
Number/percent wall goal	18	56%	18	53%	21	66%	25	74%	27	82%	24	80%	31	89%	37	97%
Regulatory
Police—total states		45		47		47		48		48		44		46		47
Number/percent wall goal	4	9%	6	13%	6	13%	9	19%	17	35%	17	39%	19	41%	24	51%
Fire—total states		13		12		9		10		12		11		10		9
Number/percent wall goal	2	15%	2	17%	0	0%	1	10%	3	25%	3	27%	2	20%	3	33%
Utility/transportation—total states		31		33		37		37		35		29		34		33
Number/percent wall goal	6	19%	9	27%	9	24%	15	41%	17	49%	14	48%	21	62%	21	64%
Corrections—total states		48		49		49		48		48		44		47		48
Number/percent wall goal	4	8%	7	14%	10	20%	14	29%	23	48%	27	61%	34	72%	41	85%
Redistributive
Welfare—total states		45		46		48		49		49		43		46		48
Number/percent wall goal	33	73%	41	89%	44	92%	48	98%	48	98%	40	93%	44	96%	48	100%
Hospitals—total states		42		40		38		39		39		34		39		38
Number/percent wall goal	25	60%	29	73%	32	84%	35	90%	38	97%	33	97%	39	100%	38	100%
Health—total states		43		48		49		47		47		41		43		45
Number/percent wall goal	25	58%	38	79%	44	90%	43	91%	47	100%	41	100%	43	100%	45	100%
Housing—total states		19		18		21		20		23		18		21		21
Number/percent wall goal	7	37%	9	50%	9	43%	11	55%	15	65%	14	78%	17	81%	18	86%
Employment security—total states		47		47		46		46		47		40		43		44
Number/percent wall goal	14	30%	23	49%	30	65%	36	78%	38	81%	37	93%	41	95%	42	95%
Indeterminant
Financial administration and general control—total states		47		49		48		49		49		46		48		48
Number/percent wall goal	4	26%	21	43%	32	67%	41	84%	47	96%	45	98%	47	98%	48	100%

Note. The wall goal is a workforce that is at least 30% female. Percent wall goal is the percentage of states that employ at least 30% female administrators in each functional area. Shown here is the number and percentage of states meeting the wall goal. The values presented here represent only the numbers and percentages of females. The Equal Employment Opportunity Commission data are limited to binary male/female gender categories.

Distributive functions

Although we observe progress toward gender balance between 1987 and 2015, occupational segregation among administrative cadres is still prevalent in highways and natural resources, where only 43% and 54% of the states, respectively, met the goal in 2015. However, in community development, 97% of reporting states met the 30% threshold by 2015, suggesting the absence of glass walls. In sum, in both highways and natural resources, we find evidence of glass walls for roughly half of reporting states, but in community development, we do not find evidence of glass walls. This is probably because community development agencies have policy missions serving both business interests, a distributive component, as well as redevelopment interests in low income areas, a redistributive element (Kerr et al., 2002).

Regulatory functions

We hypothesize that we will observe the highest level of sex-based occupational segregation in policing and fire. The data indicate over-time increases in all regulatory functions in the percentage of states reaching the 30% benchmark. The most dramatic growth is in corrections which increased from 8% in 1987 to 85% in 2015. As recently as 2015, only 33% and 51% of the states, respectively, reach 30% women among administrative workforces in fire and police. These results suggest the presence of glass walls in police and fire but not in corrections. In utilities/transportation about two thirds of reporting states reach the 30 percent benchmark, whereas slightly more than one third of reporting states indicate evidence of glass walls.

Redistributive functions

Following previous researchers, we predict that women will be well represented in administrative ranks in redistributive agencies. The results reported in Table 4, indicate minimal levels of occupational segregation among redistributive agency administrative cadres. As late as 2015, 100% of the states met or surpassed the 30 benchmark in welfare, hospitals, and health; 86% met the goal in housing; and 95% met the goal in employment security. Although these gains are impressive, suggesting the erosion or absence of glass walls, we emphasize that salaries are lower in redistributive agencies than in other types of agencies (Choi, 2018; Guy, 2017). Furthermore, redistributive agencies are rule-bound and centralized, have limited administrative discretion, and often service a client base of women (Lowi, 1985; Newman, 1994).

Financial administration and general control

We label this category “indeterminant” because it includes agencies with a wide range of missions (e.g., judicial, treasurer, comptroller, auditor, etc.). We observe substantial gains in this category, as 100% of states reached the 30% goal for administrators in 2015. We turn now to analysis of occupational segregation in the professional ranks in all functional areas for 1987-2015, again based on the 30% benchmark.

Occupational Segregation Among Professionals by Functional Policy Area and Year

Distributive functions

Although sex-based occupational segregation is not as prevalent in the professional ranks of distributive functions as in the administrative ranks, we do still observe some evidence of segregation (see Table 5). In highways, for example, in 2015, 43% of states met the goal of 30% women in administrative positions, whereas 57% of states met the goal in professional positions. By contrast, as of 2015, natural resources and community development agencies made substantial over-time gains in meeting the 30 benchmark for administrative positions (4%-54% and 56%-97%, respectively) and professional positions (6%-79% and 79%-100%, respectively). In sum, although states are showing greater progress toward gender balance among professionals as compared with administrators in distributive agencies, some states still show evidence of glass walls.

Table 5.

Empirical Results for the 30% Threshold: Occupational Segregation Among Professionals by Functional Policy Area and Year.

	States meeting the 30% threshold for professionals by functional area
	1987		1991		1995		1999		2003		2007		2011		2015
Distributive
Highways—total states		47		48		48		47		47		42		44		47
Number/percent wall goal	0	0%	4	8%	3	6%	6	13%	11	23%	19	45%	25	57%	27	57%
Natural resources—total states		48		49		49		49		49		44		47		48
Number/percent wall goal	3	6%	7	14%	16	33%	18	37%	27	55%	27	61%	33	70%	38	79%
Community development—total states		33		36		33		35		36		32		36		39
Number/percent wall goal	26	79%	32	89%	30	91%	34	97%	35	97%	31	97%	35	97%	39	100%
Regulatory
Police—total states		45		47		48		48		48		44		46		47
Number/percent wall goal	8	18%	18	38%	28	58%	37	77%	43	90%	40	91%	46	100%	45	96%
Fire—total states		14		13		13		11		13		12		12		13
Number/percent wall goal	1	7%	3	23%	4	31%	3	27%	4	31%	5	42%	6	50%	6	46%
Utility/transportation—total states		33		35		37		37		35		31		34		32
Number/percent wall goal	15	45%	16	46%	21	57%	24	65%	25	71%	29	94%	28	82%	30	94%
Corrections—total states		48		49		49		48		48		44		47		48
Number/percent wall goal	35	73%	43	88%	47	91%	48	100%	47	98%	44	100%	47	100%	48	100%
Redistributive
Welfare—total states		46		47		48		49		49		43		46		48
Number/percent wall goal	46	100%	47	100%	48	100%	49	100%	49	100%	42	98%	46	100%	48	100%
Hospitals—total states		43		41		38		39		39		33		39		38
Number/percent wall goal	43	100%	41	100%	38	100%	39	100%	39	100%	33	100%	39	100%	38	100%
Health—total states		44		48		49		47		47		41		43		45
Number/percent wall goal	42	95%	48	100%	49	100%	47	100%	47	100%	41	100%	43	100%	45	100%
Housing—total states		21		20		21		21		23		19		21		22
Number/percent wall goal	18	86%	18	90%	19	90%	20	95%	22	96%	19	100%	20	95%	20	91%
Employment security—total states		47		48		46		46		47		41		44		45
Number/percent wall goal	43	91%	47	98%	46	100%	46	100%	47	100%	40	98%	44	100%	45	100%
Indeterminant
Financial administration and general control—total states		47		49		48		49		49		46		48		48
Number/percent wall goal	42	89%	48	98%	48	100%	49	100%	49	100%	46	100%	48	100%	48	100%

Note. The wall goal is 30% or more female. Percent wall goal is the percentage of states that employ at least 30% female professionals in each functional area. Shown here is the number and percentage of states meeting the wall goal. The values presented here represent only the numbers and percentages of females. The Equal Employment Opportunity Commission data are limited to binary male/female gender categories.

Regulatory functions

As reported in Table 5, almost all states met the 30% goal among professional workforces in police, corrections, and utilities/transportation. Only 6 out of 13 states achieved the 30% threshold in fire. State police forces experienced the most dramatic increase among professionals, from 18% in 1987 to 96% in 2015. Among state professional workforces, we find evidence for the persistence of glass walls in fire but find evidence for the erosion of glass walls in other regulatory agencies.

Redistributive functions

We find the weakest support for occupational segregation in the professional ranks of redistributive agencies. Since 1997, with the exception of housing (which vacillates between 91% and 100%), all redistributive functions report 100% of states either meeting or surpassing the 30% goal. The gains in this functional area are very impressive, but we reiterate that these agencies are characterized by lower salaries and limited discretion.

Financial administration and general control

Consistent with the findings of Kerr et al. (2002), professional workforces in this indeterminant category are every bit as gender balanced as those in redistributive agencies. From 1997 through 2015, 100% of reporting states reached or surpassed the 30% goal.

In summary, we observe substantial progress for women with increases in all functional areas. Many of the patterns we observe are consistent with Lowi’s (1985) and Newman’s (1994) theoretical arguments. As we argue above, the 30% benchmark for occupational segregation is considerably lower than the standard for true proportional representation of women. To provide a more complete picture of women’s progress in the upper echelons of state bureaucracies, we supplement the 30% threshold analysis with a companion analysis of gender parity, based on the 50% threshold. If glass walls have truly been broken, we should observe large percentages of states achieving the goal of true parity. We turn now to our second analysis and narrative, examination of sex-based occupational segregation using the 50% benchmark first for administrators, and then, for professionals.

Empirical Results for the 50% Threshold

Parity Among Administrators by Functional Policy Area and Year

Gender parity is achieved if a state reaches or exceeds 50% women among its administrative (or professional) workforce in a given functional area. This benchmark paints a different picture than does the 30% benchmark. Table 6 shows the progress made by states for administrators in each functional area for a selection of years between 1987 and 2015.

Table 6.

Empirical Results for the 50% Threshold: Gender Parity Among Administrators by Functional Policy Area and Year.

	States meeting the 50% threshold for administrators by functional area
	1987		1991		1995		1999		2003		2007		2011		2015
Distributive
Highways—total states		47		48		48		46		47		42		44		47
Number/percent parity	0	0%	1	2%	0	0%	0	0%	0	0%	0	0%	1	2%	0	0%
Natural resources—total states		48		48		49		49		49		44		47		48
Number/percent parity	0	0%	0	0%	0	0%	0	0%	0	0%	0	0%	0	0%	2	4%
Community development—total states		32		34		32		34		33		30		35		38
Number/percent parity	6	19%	5	15%	4	13%	9	26%	12	37%	7	23%	15	43%	25	66%
Regulatory
Police—total states		45		47		47		48		48		44		46		47
Number/percent parity	0	0%	0	0%	1	2%	1	2%	3	6%	4	9%	3	7%	6	13%
Fire—total states		13		12		9		10		12		11		10		9
Number/percent parity	2	15%	1	8%	0	0%	1	10%	1	8%	0	0%	0	0%	2	22%
Utility/transportation—total states		31		33		37		37		35		29		34		33
Number/percent parity	0	0%	0	0%	2	5%	4	11%	3	9%	6	21%	5	15%	5	15%
Corrections—total states		48		49		49		48		48		44		47		48
Number/percent parity	1	2%	0	0%	1	2%	0	0%	0	0%	2	5%	5	11%	7	15%
Redistributive
Welfare—total states		45		46		48		49		49		43		46		48
Number/percent parity	12	27%	19	41%	19	40%	29	59%	36	73%	37	86%	41	89%	46	96%
Hospitals—total states		42		40		38		39		39		34		39		38
Number/percent parity	7	17%	13	33%	11	29%	16	41%	23	59%	27	79%	32	82%	32	84%
Health—total states		43		48		49		47		47		41		43		45
Number/percent parity	2	5%	9	19%	13	27%	14	30%	30	64%	31	76%	36	84%	42	93%
Housing—total states		19		18		21		20		23		18		21		21
Number/percent parity	3	16%	3	17%	6	29%	8	40%	9	36%	10	56%	13	62%	11	52%
Employment security—total states		47		47		46		46		47		40		43		44
Number/percent parity	1	2%	3	6%	4	9%	4	9%	12	26%	15	38%	24	56%	28	64%
Indeterminant
Financial administration and general control—total states		47		49		48		49		49		46		48		48
Number/percent parity	0	0%	3	6%	0	0%	3	6%	5	10%	8	17%	13	27%	19	40%

Note. Parity is a workforce that is at least 50% female. Percent parity is the percentage of states that employ at least 50% female administrators in each functional area. Shown here is the number and percentage of states reaching parity. The values presented here represent only the numbers and percentages of females. The Equal Employment Opportunity Commission data are limited to binary male/female gender categories.

Distributive functions

The steady erosion of glass walls observed between 2001 and 2015, using the 30% benchmark, does not appear to be present for all distributive functions using a marker of 50%. Among highway administrators, no states achieve sex-based parity in 2015, and only two states achieve parity in natural resources in 2015. The percentage of states achieving parity in community development increased from 19% to 66% over the 28-year period. We find support for widespread patterns of female underrepresentation in distributive agencies. Roughly two thirds of states reach parity in community development, but we find almost no evidence for parity among administrative ranks in other distributive agencies.

Regulatory functions

Although the percentage of states meeting the 30% benchmark for administrators is low, at 30%, we observe a steady increase for all regulatory agencies from 1987 to 2015. This does not hold true at the 50% benchmark. Based on the parity standard, all regulatory agencies experienced increases, but these were minimal, reaching only 15% (police), 15% (corrections), and 22% (fire) by 2015. We hypothesized and find most states do not achieve parity among administrative workforces in regulatory agencies.

Redistributive functions

Similar to the progress made from 1987 to 2015 in meeting the 30% benchmark, we predicted that most states would meet parity in redistributive agencies. Most states achieve parity and experienced increases from 1987 to 2015; however, these numbers are not consistently high across agencies. In 2015, well more than 80% of states achieved parity in welfare, health, and hospitals, but in housing, only 52% of states reach parity, and in employment security, only 64% of states do so. We hypothesized that parity would be present among administrative cadres in the redistributive functions. Although gender parity was generally achieved for welfare, hospitals, and health, large numbers of states do not achieve parity in housing and employment security. The percentage of states meeting parity is higher for redistributive agencies than for other types of agencies.

Financial administration and general control

We observe large over-time gains for states in financial administration/general control using the 30% benchmark. By contrast, using the 50% threshold, we observe 0% to 6% of states achieving parity from 1987 through 1999 with more substantial increases beginning in 2007 (see Table 6). By 2015, 40% of states reach the 50% benchmark. Occupational sex-based segregation, while not an issue in the indeterminant function at 30%, is fairly extensive under the 50% parity benchmark. We now turn to findings on states meeting parity for professional workforces by functional area from 1987 to 2015.

Parity Among Professionals by Functional Policy Area and Year

Distributive functions

Unlike the findings based on the 30% benchmark, Table 7 does not show uniform increases in states reaching the 50% benchmark in all distributive agencies between 1987 and 2015. Natural resources fluctuates between 0% and 2%, with only one state meeting parity by 2015. Moreover, highways increases from 0% to 6% of states achieving parity over the 28-year period. Table 7 presents a different picture for community development with an increase in the percentage of states reaching parity from 21% to 90% over 28 years. These data show limited progress among professional cadres in distributive agencies. We hypothesized female underrepresentation among professional workforces in the distributive function. Our expectations are confirmed for highways and natural resources, but almost all states achieve parity in community development by 2015.

Table 7.

Empirical Results for the 50% Threshold: Gender Parity Among Professionals by Functional Policy Area and Year.

	States meeting the 50% threshold for professionals by functional area
	1987		1991		1995		1999		2003		2007		2011		2015
Distributive
Highways—total states		47		48		48		47		47		42		44		47
Number/percent parity	0	0%	0	0%	0	0%	0	0%	0	0%	1	2%	1	2%	3	6%
Natural resources—total states		48		49		49		49		49		44		47		48
Number/percent parity	1	2%	0	0%	0	0%	0	0%	1	2%	0	0%	0	0%	1	2%
Community development—total states		33		36		33		35		36		32		36		39
Number/percent parity	7	21%	9	25%	12	36%	17	49%	20	56%	23	72%	29	81%	35	90%
Regulatory
Police—total states		45		47		48		48		48		44		46		47
Number/percent parity	1	2%	5	11%	6	13%	7	15%	8	17%	14	32%	26	56%	26	55%
Fire—total states		14		13		13		11		13		12		12		13
Number/percent parity	1	7%	2	15%	2	15%	2	18%	1	8%	2	17%	2	17%	4	31%
Utility/transportation—total states		33		35		37		37		35		31		34		32
Number/percent parity	1	3%	3	9%	6	16%	7	19%	6	17%	8	26%	8	24%	9	28%
Corrections—total states		48		49		49		48		48		44		47		48
Number/percent parity	3	6%	5	10%	5	10%	8	17%	19	40%	22	50%	26	55%	31	65%
Redistributive
Welfare—total states		46		47		48		49		49		43		46		48
Number/percent parity	41	89%	46	98%	46	96%	48	98%	49	100%	42	98%	46	100%	48	100%
Hospitals—total states		43		41		38		39		39		33		39		38
Number/percent parity	42	98%	41	100%	38	100%	39	100%	39	100%	32	97%	39	100%	38	100%
Health—total states		44		48		49		47		47		41		43		45
Number/percent parity	29	66%	38	79%	39	80%	40	85%	43	91%	38	93%	41	95%	44	98%
Housing—total states		21		20		21		21		23		19		21		22
Number/percent parity	12	57%	11	55%	15	71%	17	81%	20	87%	16	84%	15	71%	17	77%
Employment security—total states		47		48		46		46		47		41		44		45
Number/percent parity	25	53%	30	63%	31	67%	43	93%	42	89%	39	95%	42	95%	44	98%
Indeterminant
Financial administration and general control—total states		47		49		48		49		49		46		48		48
Number/percent parity	3	6%	6	12%	14	29%	29	59%	33	67%	34	74%	39	81%	43	90%

Note. Parity is a workforce that is at least 50% female. Percent parity is the percentage of states that employ at least 50% female professionals in each functional area. Shown here is the number and percentage of states reaching parity. The values presented here represent only the numbers and percentages of females. The Equal Employment Opportunity Commission data are limited to binary male/female gender categories.

Regulatory functions

In 2015, only about 30% of states reached the 50% benchmark in fire and utilities/transportation. Although large gains were observed for police and corrections, only 55% of states met parity in police and only 65% met parity in corrections by 2015. We hypothesized that there would be widespread female underrepresentation among state professional workforces in regulatory agencies. Although sex-based occupational segregation among professionals did not appear to be widespread in the regulatory function based on the 30% benchmark (with the exception of fire), when the benchmark is set at true parity a different story emerges, many states fall short of the goal.

Redistributive functions

The trends observed using the 30% benchmark for redistributive agencies are roughly consistent with those observed based on the parity benchmark. By 2015, all states achieved parity in welfare and hospitals, and all but one achieved parity in health, and all but one state did so in employment security. Only 77% of states achieve parity in housing; however, this percentage is high compared with percentages for most distributive and regulatory agencies. Consistent with our expectations, we find that states achieve parity among professional workforces in the redistributive function. This finding provides a basis optimism for it supports the proposition that glass walls have eroded. This optimism needs to be tempered. As noted previously, redistributive agencies pay relatively low salaries and are characterized by relatively low amounts of policy and program discretion, once again reinforcing the adage that women’s work is less valued than men’s work.

Financial administration and general control

Similar to the results based on the 30% benchmark in financial administration and general control, 43 out of 48 states (90%) reached the parity goal among professional workforces in 2015.

The empirical results for the 50% threshold (i.e., gender parity) suggest a different story than the one based on results from the application of the customary 30% threshold. In this section, we present selected findings in three separate line graphs designed to help illustrate the narrative relating to sex-based progress (and lack of progress) in state-level administrative and professional workforces based on true parity, the threshold of 50%. We hypothesized that there would be sex-based parity in the redistributive function. The hypothesis is supported for professionals, but for administrative workforces, several states do not achieve parity in housing and employment security, even as recently as 2015 (see Figure 1).

Figure 1.

Gender parity by year in redistributive functions: Administrators.

We hypothesized that there would be underrepresentation of females (i.e., many states would not achieve the 50% benchmark) in distributive and regulatory agencies. This hypothesis is supported for both administrators and professionals in regulatory agencies. In the distributive function, we find that an overwhelming majority of state professional workforces fail to reach parity in highways and natural resources. The percentage of states reaching parity in professional cadres in community development agencies increases from about 20% to roughly 90% between 1987 and 2015 (see Figure 2).

Figure 2.

Gender parity by year in the distributive function: Professionals.

Although growth in the percentage of states reaching the 30% threshold appears to be impressive among administrative workforces in regulatory agencies, application of the parity standard points to a lack of progress. Based on the 50% threshold, little to no improvement is observed among administrators in highways and natural resources (see Table 6). Similarly, the percentage of states achieving parity is extremely low among administrative workforces in regulatory agencies (see Figure 3). Clearly, setting workforce goals at 30% female managers suggests an overly optimistic picture of gender balance when, in fact, many states are falling short of parity.

Figure 3.

Gender parity by year in the regulatory function: Administrators.

Summary and Discussion

Extending, through 2015, the data employed by Kerr et al. (2002) and by replicating the use of the customary 30% benchmark, we find a fairly optimistic picture of women’s public sector employment progress. Overall, occupational segregation has become less frequent across functional areas. Consistent with Lowi (1985) and Newman (1994), Kerr et al. (2002) found occupational sex-based segregation in distributive and regulatory functions. Our data show some improvement over time in highways and natural resources as well as evidence for the erosion of glass walls in community development. We find sex-based occupational segregation to be most prevalent in regulatory agency administrative workforces where relatively few states reach the 30% benchmark. For professionals, however, sex-based occupational segregation based on the 30% benchmark appears to be improving. Similar to Kerr et al. (2002), we find overall support for the erosion of glass walls in redistributive agencies. We also find relatively little evidence for the presence of glass walls in financial administration and general control. To recap, the narrative at the 30% threshold is one of steady, over-time progress toward gender balance in almost all types of state agencies, suggesting cause for optimism, and suggesting that before too long numbers of women and men in state-level administrative and professional positions across different agencies will be roughly equal.

Based on the standard of proportional representation for women, we raised the benchmark to absolute gender parity (a 50% threshold) and conducted a second analysis. The second analysis is reflective of a different narrative. It is no surprise that occupational segregation in distributive agencies appears to be more prevalent when the benchmark is set at 50%. Most states fail to achieve parity in highways and natural resources, even as late as 2015. Although gains are observed in regulatory agencies for administrators and professionals, many states fall far short of achieving gender parity, especially, among administrative workforces. Although Kerr et al. (2002) found no evidence of glass walls in the redistributive function, we find a different story employing the 50% benchmark. Parity is the norm for professional workforces, but for administrators, this is true for only welfare, hospitals, and health. In 2015, many states did not achieve gender parity among administrators in housing and employment security. In financial administration and general control, we observe widespread gender parity for professionals, but 60 percent of reporting states do not reach parity for their administrative workforces.

Raising the benchmark to parity allows us to evaluate equity in leadership posts. In doing so, we find that far fewer states meet the goal. Parity does not appear to exist in most states in regulatory agencies or in the majority of distributive agencies for either administrators or professionals. Likewise, redistributive agencies that indicate overwhelmingly high rates of gender balance at 30% are less likely to meet the goal of absolute parity. Although analysis based on the 30% marker tells a story of significant progress and promise for women in state-level bureaucracies (with a few exceptions), this story is misleading and in need of qualification. Setting the bar at true parity tells a much different, less optimistic story. If Levin-Waldman’s (2004) observation about the political uses of data is accurate, less progressive stakeholders will emphasize findings from the analysis based on a 30% threshold and perhaps ignore or discredit findings based on higher thresholds such as 50%. However, progressively minded stakeholders sympathetic to feminist values will emphasize analysis based on the standard of absolute parity. Those favoring the status quo might prefer the 30% standard; those favoring more than incremental change might prefer the 50% standard. In any case, there is value to conducting analyses at 30% and 50% as well as to the simultaneous consideration of these analyses. The narrative based on the 30% marker has for long been the dominant story, perhaps explaining its widespread acceptance. When Levin-Waldman (2004) observes that no one model can drive the policy process, his words seem to reinforce Chimamanda Ngozi Adichie’s (2013) warning about reliance on the single narrative. We do not judge people based on a single story, nor should we judge or make policy based on one story or based on the dominant narrative. Clearly, raising the bar to 50% makes the narrative less forgiving. Based on the second narrative, states still have a long way to go before achieving equality among managers in state government. We argue that in future research, the 30% benchmark should not be used as evidence of women “breaking” the glass wall as we move further into the 21st century. Furthermore, any analysis based on the 30% benchmark should be supplemented with other types of analysis, analysis based on standards more relevant to the contemporary period or on standards that are less conservative/more progressive. Absolute parity is a progressive standard consistent with a commitment to addressing the subordination of women in the workplace as well as their autonomy to influence and make policies within their respective agencies and for their constituencies. What we find exemplifies and extends Sandberg’s (2013) statement to NPR: “Ten years of no progress is no progress” to 18 years of no progress.

Footnotes

Appendix

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Larra Rucker

Notes

Author Biographies

Valerie H. Hunt is an associate professor of political science and associate director of the public policy PhD. program at the University of Arkansas. Her research focuses on social inequality, bureaucratic representation, and intersectionality. Her work has appeared in such journals as Race, Class, and Gender; American Indian Quarterly; Journal of Women, Politics, and Policy; Journal of Sociology & Welfare; and Sociology Spectrum.

Larra Rucker is a public policy PhD candidate at the University of Arkansas. Her research focuses on child care policy, issues specific to the early childhood education workforce, and gender-based occupational segregation.

Brinck Kerr is a professor of political science and director of the public policy PhD program at the University of Arkansas. His research focuses on representation, race/ethnicity, and gender. His work has appeared in many journals including American Journal of Political Science, Public Administration Review, Administration & Society, American Review of Public Administration, Urban Affairs Review, and Social Science Quarterly.

References

Adichie

C. N.

(2013, November 18). Africa: The danger of the single story [Amizade blog post]. Retrieved from https://amizade.org/danger-of-a-single-story

Alkadry

M. G.

Tower

L. E.

(2014). Women and public service: Barriers, challenge and opportunities. New York, NY: Routledge.

Battaglio

R. P.

Jr. (2015). Public human resource management: Strategies and practices in the 21st century. Los Angeles, CA: Sage.

Bowling

C. J.

Kelleher

C. A.

Jones

Wright

D. S.

(2006). Cracked ceilings, firmer floors, and weakening walls: Trends and patterns in gender representation among executives leading American state agencies, 1970–2000. Public Administration Review, 66, 823-836.

Bullard

Wright

(1993). Circumventing the glass ceiling: Women executives in American state governments. Public Administration Review, 53, 189-202.

Caceres-Rodriguez

(2013). The glass ceiling revisited: Moving beyond discrimination in the study of gender in public organizations. Administration & Society, 45, 674-709.

Cayer

N. J.

Sigelman

(1980). Minorities and women in state and local government: 1973-1975. Public Administration Review, 40, 443-450.

Choi

(2010). Diversity in the U.S. federal government: Antecedent and correlates of diversity in federal agencies. Review of Public Personnel Administration, 30, 301-321.

Choi

(2018). Pay equity in government: Analyzing determinants of gender pay disparity in the U.S. federal government. Administration & Society, 50, 346-371.

10.

Connell

(2006). Glass ceilings or gendered institutions? Mapping the gender regimes of public sector worksites. Public Administration Review, 66, 837-849.

11.

Corson

J. J.

Paul

R. S.

(1966). Men near the top. Baltimore, MD: Johns Hopkins University Press.

12.

D’Agostino

M. J.

(2015). The difference that women make: Government performance and women-led agencies. Administration & Society, 47, 532-548.

13.

Dolan

(2004). Gender equity: Illusion or reality for women in the federal executive service Public Administration Review, 64, 299-308.

14.

Ford

Dolan

(1999). Contemporary women state legislators: A diverse group with diverse agendas. In Duke

L. L.

(Ed.), Women in politics: Outsiders or insiders? (pp. 157-172). Upper Saddle River, NJ: Prentice Hall.

15.

Fraga

L. R.

Martinez-Ebers

Lopez

Ramirez

(2005, August-September). Strategic intersectionality: Gender, ethnicity, and political incorporation. Paper presented at the annual meeting of the Western Political Science Association, Oakland, CA.

16.

Gonzalez

Denisi

(2009). Cross-level effects of demography and diversity climate on organizational attachment and firm effectiveness. Journal of Organizational Behavior, 30, 21-40.

17.

Guy

M. E.

(2017). Mom work versus dad work in local government. Administration & Society, 49, 48-64.

18.

Herbert

F. T.

Brudney

J. L.

Wright

D. S.

(1999). Reinventing government in the American states: Measuring and explaining administration reform. Public Administration Review, 59, 19-30.

19.

Hogue

Lord

R. G.

(2007). A multilevel, complexity theory approach to understanding gender bias in leadership. The Leadership Quarterly, 18, 370-390.

20.

Jacobson

Palus

Bowling

(2010). A woman’s touch? Gendered management and performance in state administration. Journal of Public Administration Research and Theory, 20, 477-504.

21.

Johnson

B. J.

Crum-Cano

B. D.

(2011). Glass walls in urban planning: An examination of policy type and gender segregation within a profession. Review of Public Personnel Administration, 31, 386-409.

22.

Kellough

J. E.

(1989). Federal equal employment opportunity policy and numerical goals and timetables: An impact assessment. New York, NY: Praeger.

23.

Kellough

J. E.

(1990). Integration in the public workplace. Determinants of minority and female employment in federal agencies. Public Administration Review, 50, 557-566.

24.

Kerr

Miller

Reid

(2002). Sex-based occupational segregation in U.S. state bureaucracies, 1987-97. Public Administration Review, 62, 412-423.

25.

Kottke

J. L.

Agars

M. D.

(2005). Understanding the processes that facilitate and hinder efforts to advance women in organizations. Career Development International, 10, 190-202.

26.

Lawless

J. L.

(2004). Politics of presence? Congresswomen and symbolic representation. Political Research Quarterly, 57, 81-99.

27.

Levin-Waldman

O. M.

(2004). Policy orthodoxies, the minimum wage, and the challenge of social science. Journal of Economic Issues, 38, 139-154.

28.

Lewis

G. B.

Emmert

M. A.

(1986). The sexual division of labor in federal employment. Social Science Quarterly, 67, 143-156.

29.

Lewis

G. B.

Nice

(1994). Race, sex, and occupational segregation in state and local governments. American Review of Public Administration, 24, 393-410.

30.

Lowi

T. J.

(1964). American business, public policy, case studies, and political theory. World Politics, 16, 677-715.

31.

Lowi

T. J.

(1972). Four systems of policy, politics, and choice. Public Administration Review, 32, 298-310.

32.

Lowi

T. J.

(1985). The state in politics: The relation between policy and administration. In Noll

(Ed.), Regulatory policy and the Social Sciences (pp. 67-104). Berkeley: University of California Press.

33.

Meier

Bohte

(2007). Politics and the bureaucracy: Policymaking in the fourth branch of government. Belmont, CA: Thomson Wadsworth.

34.

Miller

Kerr

Reid

(1999). A national study of gender-based occupational segregation in municipal bureaucracies: Persistence of glass walls? Public Administration Review, 59, 218-230.

35.

Mladenka

(1991). Public employee unions, reformism, and Black employment in 1200 American cities. Urban Affairs Quarterly, 26, 532-548.

36.

Mosher

(1968). Democracy and the public service. New York, NY: Oxford University Press.

37.

Naff

K. C.

(1994). Through the glass ceiling: Prospects for the advancement of women in the federal civil service. Public Administration Review, 54, 507-514.

38.

Newman

M. A.

(1994). Gender and Lowi’s thesis: Implications for career advancement. Public Administration Review, 54, 277-284.

39.

O’Neil

D. A.

Hopkins

M. M.

(2008). Women’s careers at the start of the 21st century: Patterns and paradoxes. Journal of Business Ethics, 80, 727-743.

40.

Osborne

(1988). Laboratories of democracy. Boston, MA: Harvard Business School Press.

41.

Powell

G. N.

(2012). Six ways of seeing the elephant: The intersection of sex, gender, and leadership. Gender in Management: An International Journal, 27, 119-141.

42.

Reid

M. F.

Kerr

J. B.

Miller

(2000). A study of the advancement of women in municipal government bureaucracies: Persistence of glass ceilings? Women & Politics, 21, 35-53.

43.

Reid

M. F.

Kerr

J. B.

Miller

(2003). Glass walls and glass ceilings: Women’s representation in state and municipal bureaucracies. New York, NY: Praeger.

44.

Riccucci

N. M.

(1986). Female and minority employment in city government: The role of unions. Policy Studies Journal, 15, 3-16.

45.

Riccucci

N. M.

(2009). The pursuit of social equity in the federal government: A road less traveled? Public Administration Review, 69, 373-382.

46.

Riccucci

N. M.

Saldivar

(2014). The status of employment discrimination suits in police and fire departments across the United States. Review of Public Personnel Administration, 34, 263-288.

47.

Rice

M. F.

(2007). A post-modern cultural competency framework for public administration and public service delivery. International Journal of Public Sector Management, 20, 622-637.

48.

Rosenthal

C. S.

(2005). Women leading legislatures: Getting there and getting things done. In Thomas

Wilcox

(Eds.), Women and elective office: Past, present, and future, 2nd ed. (pp. 197-212). New York, NY: Oxford University Press.

49.

Sabharwal

(2015). From glass ceiling to glass cliff: Women in senior executive service. Journal of Public Administration Research and Theory, 25, 399-426.

50.

Sandberg

(2013, March 11). “Lean in”: Facebook’s Sheryl Sandberg explains what’s holding women back [NPR interview]. npr. Retrieved from http://www.npr.org/2013/03/11/173740524/lean-in-facebooks-sheryl-sandberg-explains-whats-holding-women-back

51.

Selden

S. C.

(1997). The promise of representative bureaucracy: Diversity and responsiveness in a government agency. Armonk, NY: M. E. Sharpe.

52.

Sapiro

(2018). Sexual harassment: Performances of gender, sexuality, and power. Perspectives on Politics, 16, 1053-1066.

53.

Smith

A. E.

Monaghan

K. R.

(2013). Some ceilings have more cracks: Representative bureaucracy in federal regulatory agencies. The American Review of Public Administration, 43, 50-71. doi:10.1177/0275074011420997

54.

Sneed

B. G.

(2007). Glass walls in state bureaucracies: Examining the difference departmental function can make. Public Administration Review, 67, 880-891.

55.

Stein

(1986). Representative local government: Minorities in the municipal workforce. Journal of Politics, 48, 694-713.

56.

Tomaskovic-Devey

(1993). Gender and racial inequality at work: The sources and consequences of job segregation. Ithaca, NY: ILR Press.

57.

Tomaskovic-Devey

Kalleberg

A. L.

Marsden

P. V.

(1996). Organizational patterns of gender segregation. In Kalleberg

A. L.

Knoke

Marsden

P. V.

Spaeth

J. L.

(Eds.), Organizations in America (pp. 276-301). Thousand Oaks, CA: Sage.

58.

Tower

L. E.

Alkadry

M. G.

(2008). The social costs of career success for women. Review of Public Personnel Administration, 28, 144-159.

59.

U.S. Bureau of Labor Statistics. (2017). Women in the labor force: A databook (BLS Report No. 1065). Retrieved from https://www.bls.gov/opub/reports/womens-databook/2016/pdf/home.pdf

60.

U.S. Equal Employment Opportunity Commission. (1986). EEOC form 164, state and local government information (EEO-4). Washington, DC: U.S. Government Printing Office.

61.

Van Horn

C. E.

(1989). The state of the states. Washington, DC: CQ Press.

62.

Warner

Steel

Lovrich

(1989). Conditions associated with the advent of representative bureaucracy: The case of policing. Social Science Quarterly, 70, 562-578.

63.

Weyer

(2007). Twenty years later: Explaining the persistence of the glass ceiling for women leaders. Women in Management Review, 22, 482-496.