Women and Public Administration in a Comparative Perspective

Abstract

This article explores two questions related to whether passive representation leads to active representation using Brazilian municipal data: Does electing women to public office increase the proportion of women in public administration? Does the representation of women in elected office and public administration lead to better representation of women’s interests? Results suggest that women elected leaders increase the probability that women will be appointed to head public agencies, and through these agency heads they indirectly affect representation in other administrative positions. In addition, women elected officials and public administrators are also associated with the adoption of more women-friendly policies.

Keywords

gender local government representation Latin America women

One perennial question for the study of women in public administration is that of active representation: Do bureaucracies that have more women employees generate policies or processes that differ from those produced by less representative bureaucracies? Although the representation of women in government bureaucracies varies a great deal (see Peters, von Maravic, & Schroeder, 2013) and merits study in its own right, the process of active representation is seen as contributing to a more democratic polity. The relatively robust literature on active representation of women in U.S. bureaucracies stands in stark contrast to the lack of studies in other countries (but see Andrews, Ashworth, & Meier, 2014; Park, 2013). This absence is surprising because there is an extensive comparative literature on active representation by women in legislative bodies (e.g., Escobar-Lemmon & Taylor-Robinson, 2014; Heath, Schwindt-Bayer, & Taylor-Robinson, 2005; Matland, 1998; Tripp & Kang, 2008).

This article has two objectives. First, using the concept of gender identity, it builds on the representative bureaucracy literature and offers hypotheses about the conditions that are likely to foster the active representation of women. The argument follows the logic of top-down representation and suggests that increasing the presence of women in elected offices will increase the presence of women in top-level public management positions, which will then increase the overall number of women in public administration. As a result, this cascade of numerical representation (in both elected and administrative positions) will lead to the active representation of women’s interests. This task is complicated by the fluid nature of gender, gender identity, and intersecting identities, which vary across time and space. The process of identity salience is both highly political, as government institutions and policy generate outcomes that have disparate impacts on women and men, and highly social with influences from religion, economics, and historical patterns.

Second, we examine these theoretical expectations using an extensive database of more than 5,000 Brazilian local governments. This application of representative bureaucracy theory contributes to the literature by exploring gender representation in the context of a developing country that has a “friendly government, cruel society” (Gómez, 2010) orientation toward issues of gender and sexuality. Despite the presence of a legislated gender quota and the election of a woman president, Brazil has some of the lowest levels of women’s representation in the world. In 2014, women comprised only 10% of the lower chamber of the legislature and 14% of the senate,¹ and only one woman—Suely Campos—was elected governor (out of 27 states) in the 2014 elections. At the local level, approximately 10% of mayors are women. The representation of women and women’s interests is a crucial concern in Brazil given the vast economic and social inequalities that exist between men and women and the prevailing occurrence of violence against women throughout the country (Waiselfisz, 2012). Furthermore, exploring gender representation at the local level is imperative because municipalities are responsible for administering important social services and providing resources for women in need.

Venues for the Active Representation of Women

The literature on representative bureaucracy distinguishes between passive and active representation. A bureaucracy is representative in the passive sense if the bureaucracy has the same characteristics as the population on variables of interest (e.g., race, ethnicity, social class, gender, etc.; see Mosher, 1982). In fact, passive representation is sometimes called symbolic representation because it defines representation as “standing for” (Pitkin, 1967). Although much literature is focused on passive representation and its determinants, the concern of this article is when, and under what conditions, passive representation generates active representation.

Active representation is defined as Person A (the representative) actively seeking to pursue actions that are in the interests of Person B (the represented). In short, if A acts as B would act if B were in the same position with the same information that A possess, then A represents B. This logic indicates that passive representation is a characteristic, but active representation is a process. If A is an active representative of B, it does not guarantee that B’s interests will be served; only that A will attempt to do so. A could fail in the effort to attain the outcomes of representation but still partake in the process of active representation.

Existing studies of passive and active representation have not actually documented that active representation exists (that is they have not detailed the process), but rather only that passive representation is correlated with some outputs or outcomes that benefit the group to be represented. Such a linkage has been taken as evidence that representation exists (see Lim, 2006). The basic logic of how passive representation might generate active representation is relatively straightforward. The literature assumes that all bureaucrats exercise discretion and that in the exercise of discretion the values of the bureaucrat will come into play. These values include the bureaucrats’ various identities; the most important for the purposes of this study is gender identity. Gender identity refers to “one’s sense of oneself as male, female, or transgender” (American Psychological Association, 2006) and may be correlated with one’s biological sex, but does not have to be.

When the sphere of discretion exercised by the bureaucrat overlaps with the bureaucrat’s gender identity, that is, when the values and preferences the bureaucrat holds as a result of his or her gender identity correspond to some discretionary action where gender identity is relevant, active representation is possible. In more specific terms, Keiser, Wilkins, Meier, & Holland (2002, p. 556) contend that women in the bureaucracy are likely to represent women when a policy issue is gendered, which occurs if (a) the policy benefits women as a class, (b) the policy issue has been defined as a women’s issue by the political process, or (c) if the gender of the bureaucrat changes the relationship between the bureaucrat and the client.

The threefold definition of Keiser et al. (2002) clearly allows for what is a gendered issue to vary across countries and different time periods within a country. A country such as Sweden with elaborate family leave policies de-genders some employment issues in contrast to countries such as Ecuador or Germany, which provide limited governmental protections for family leave, or countries that have no legally mandated protections such as the United States or Papua New Guinea (Addati, Cassirer & Gilchrist, 2014). On the second criteria for a gendered issue, the vast variation in political processes across the world will place some gender-related issues on the political agenda and exclude others. In short, the variation in national politics will put different gender-related issues on the agenda in different countries. Finally, whether or not the gender of the bureaucrat changes the relationship between the bureaucrat and the client will depend on the degree of gender consciousness of both the bureaucrat and the client, and gender consciousness clearly varies across time, countries, and individuals (Block, 1973; Pfau-Effinger, 1998).

All individuals—and public administrators are no exception—have multiple identities and the intersection of these various identities greatly complicates the translation of passive representation into active representation. A woman public administrator has a variety of identities that compete with her gender identity and knowing when an administrator views an issue as relevant to her identity as woman versus her identity as an engineer, a racial minority, a public employee, or countless other identities is very difficult. This competition among identities can serve to suppress gender representation just as institutional rules or formal structures can limit representation.

Prior to the publication of Keiser et al.’s (2002) work and their specification of the theory of representative bureaucracy and gender, studies had found that there were value differences between female and male bureaucrats (Dolan, 2000) and that greater passive representation could be linked to organizational concerns such as flex time or family-friendly policies (Duerst-Lahti & Kelly, 1995; Guy, 1992), but not that the bureaucracy changed its policy outputs. Subsequent work has shown that greater representation of women in the bureaucracy produces different outcomes in education (Keiser et al., 2002), child support enforcement (Wilkins & Keiser, 2004), law enforcement (Meier & Nicholson-Crotty, 2006), the effectiveness of fire and rescue service authorities (Andrews et al., 2014), U.S. state governments (Jacobson, Palus, & Bowling 2010), and media coverage of women’s issues, social welfare spending, and government performance in South Korea (Park, 2013).

Park’s (2013) study of gender representation in Korean bureaucracy is especially relevant to the current study in that it examines not just bureaucratic outcomes or outputs, but the influence of women bureaucrats on policy decisions made by other institutions. As one moves from more developed countries, with long-standing civil service systems based on merit and norms of political neutrality, to less developed countries, the distinction between bureaucrats and politicians becomes less clear. Elected mayors often have extensive administrative duties (see Avellaneda, 2009, 2013), and bureaucratic expertise might be tapped for roles that are considered more political in developed contexts. In Latin America, it is common for administrative positions to be appointed by elected officials. In addition, the historical legacy of bureaucratic authoritarianism in some South American countries (O’Donnell, 1979), such as Brazil, contributes to the intertwining of politics and the bureaucracy today. Because top-level administrative appointees will also exercise significant discretion and are likely to interact with women at lower levels of the bureaucracy to push policy issues (in the U.S. context, see Riccucci & Saidel, 1997), gender representation in public administration should consider the roles of all three of these types of public administrators. These literatures indicate that a full understanding of women in public administration needs to include not just the traditional bureaucrats but also elected chief executives and top-level administrative appointees.

Bureaucratic representation also does not occur in a vacuum. One of the consistent findings in the literature is that there is a positive association between representation in electoral institutions and representation in the bureaucracy on identities such as race, ethnicity, and gender (studies include Eisinger, 1982; Mladenka, 1989; Meier & O’Toole, 2006). The predominant argument in this literature is one of top-down representation: Electoral institutions become more representative and they in turn put pressure on bureaucratic institutions to also be more representative (see Eisinger, 1982; Mladenka, 1989).² Women elected representatives can realize two benefits from fostering the bureaucratic representation of women.³ First, such actions constitute what Eulau and Karps (1977) call allocative responsiveness, the provision of visible pork barrel or patronage benefits to a group. Whether such bureaucrats, in turn, actually do anything to benefit the represented does not detract from their symbolic value in this regard. Second, all politicians face uncertainty; the next election could remove them from office. Increasing the number of bureaucrats who share the politician’s values, as a result, is a way to lock in policy decisions beyond the next election. Such a process might also reduce the salience of an issue where there is strong opposition to representation simply because the bureaucratic process is less visible to the public than a political process. The street-level bureaucracy literature (May & Winter, 2009) indicates that a bureaucratic strategy of representation is likely to be more successful to the degree it penetrates the entire bureaucracy rather than just the top level of public managers.

The theoretical logic indicates that to get a full view of the representation of women, examining the levels of representation in electoral institutions, in top-level public management positions, and in the lower levels of the bureaucracy is necessary. This study includes all three levels in its examination of gender representation in Brazil and tests both the linkages between representation in different institutions and the overall impact on policies that benefit women. The first set of hypotheses (Hypotheses 1, 2, and 3) suggest that having more women in electoral institutions leads to more representative bureaucracies, both in terms of top-level public managers and lower level bureaucrats:

Hypothesis 1: Increasing women’s presence in elected office will increase the presence of women in top-level public management positions.

Hypothesis 2: Increasing women’s presence in elected office will increase the overall presence of women in public administration.

Hypothesis 3: Increasing the presence of women in top-level public management positions will increase the overall presence of women in public administration.

The second series of hypotheses (Hypotheses 4, 5, and 6) suggest that having more women in electoral institutions and in public bureaucracies will lead to the active representation of women’s interests, conceptualized in this study as the adoption and implementation of policies that are beneficial to specific subgroups of women or to women in general:

Hypothesis 4: Increasing women’s presence in elected office will increase the representation of women’s interests.

Hypothesis 5: Increasing women’s presence in top-level public management positions will increase the representation of women’s interests.

Hypothesis 6: Increasing women’s presence in public administration will increase the representation of women’s interests.

Exploring Gender Representation in Brazilian Local Governments

These hypotheses are tested within the context of Brazilian local governments. Brazil presents an interesting paradox when it comes to issues of gender and sexuality. Due to the continued electoral success of leftist parties in Brazil, the federal government has been characterized as generally friendly toward women’s and LGBT (lesbian, gay, bisexual, and transgender) interests. The government legalized same-sex marriage in 2013 and was praised for its anti-discriminatory policies and responsiveness to the AIDS epidemic (Gómez, 2010). A 30% gender quota law for all levels of government was passed in 1997 and an amendment to the law was passed in 2009 requiring political parties to allocate at least 10% of their free broadcasting time and 5% of their public campaign funds to women candidates.⁴ In addition, a comprehensive law to combat violence against women (Lei Maria da Penha) was passed in 2006, which enacted harsher punishments for perpetrators of domestic abuse and led to the creation of various resources for women, such as shelters for battered women.

At the same time, the general public, as well as law enforcement agencies, has been characterized as extremely socially conservative and “deeply affected by sexism, homophobia, and racism” (Vianna & Carrara, 2010). A recent study of violence against women in Brazil found that a woman is assaulted every 15 seconds, one is murdered every 2 hours, and nearly 100,000 women have been killed over the past 30 years, many at the hands of their intimate partner (Waiselfisz, 2012). Countless news articles have been published detailing these atrocities against women and LGBT people, with some articles pointing out the paradox between the progressive steps taken by the government and the social norms that foster tolerance of women’s rights violations (“Brazil’s Tough Laws on Violence,” 2015; “Violence and Women in Brazil,” 2013). Yet, protesters and women’s rights advocates are often critical of the government for not doing enough.

Given the prevalence of domestic abuse and hyper-sexualized images of women in the media (Vianna & Carrara, 2010) that likely contribute to the suppression and objectification of women, it is crucial that the government be on the forefront of these issues to ensure that the rights of women (and other identity groups) are protected. The seemingly most logical way to ensure that the interests of women are represented is by increasing the number of women in elected offices and public administration. However, previous research on Brazilian municipalities finds that women elected officials might not behave differently from their male counterparts and both men and women are strategic in their decisions to represent gendered interests (Funk, 2015). Thus, whether women in elected and administrative positions actually do represent women’s interests is an empirical question—one that we address below.

Brazil is a federal country with three tiers of government (union, states, municipalities) and is one of the most decentralized countries in the world (Falleti, 2010; Samuels, 2004). Brazilian municipalities provide a useful case for testing the question of gender representation for several reasons. First, there are a large number of municipalities—around 5,570—in Brazil, facilitating the use of large-N statistical analyses while holding other variables (e.g., electoral laws, language, institutional arrangements) constant at the national level. Second, the representation of women in elected office and in public administration differs across units, providing sufficient variation on the variables of interest. In addition, municipalities vary greatly in their size, level of development, and availability of resources, which makes this case an interesting one to explore gender identity and the representation of gendered interests. Third, many political and budgetary powers are decentralized to municipal governments, so municipal public officials have the power to create and implement policy in a number of areas, including policies related to women and gender issues. Municipal governments control nearly 20% of total government expenditures (Varsano & Mora, 2001), indicating that Brazilian local governments provide not only an interesting case to study gender representation but also an important one.

Table 1 presents results from a public opinion survey that was conducted in Brazil in 2014 (World Values Survey, Wave 6). The survey included a variety of questions about opinions of women, including views of women in the workforce and whether men make better political or business leaders than women. The survey indicates that while the majority of both men and women tend to have favorable views of women as leaders and women in the workforce, a notable percentage of men and women agree with the statement “when a mother works for pay, the children suffer” indicating that Brazilians, on the whole, might have more traditional or conservative views of women’s roles as mothers and homemakers than residents of countries where women are more prevalent in the workforce or where child care duties are more equally divided between men and women. Furthermore, difference of means tests indicate that there is a statistically significant difference between men and women, with women respondents having more equitable views toward women in general.⁵ If women elected officials and women bureaucrats hold views similar to women in the general public, then they should be better representatives of women’s interests and be more in favor of (and more likely to advocate for) policies that are beneficial to women. In other words, women elected officials and public administrators are likely to hold values that are associated with their gender identity and to engage these values when making policy decisions, especially when the decision concerns an issue that is of interest to women.⁶

Table 1.

Public Opinion About Women in Brazil (N = 1,486).

	% respondents agree		% respondents disagree
	Women	Men	Women	Men
When jobs are scarce, men should have more right to a job than women	13.59	20.21	77.89	66.73
When a mother works for pay, the children suffer	64.40	56.00	34.74	43.11
On the whole, men make better political leaders than women do	22.22	35.06	72.60	62.61
On the whole, men make better business executives than women do	21.36	36.67	76.70	61.00
If a woman earns more money than her husband, it is almost certain to cause problems	37.32	29.34	45.52	51.88

Source. 2014 World Values Survey (Wave 6).

Note. The totals do not add to 100% because some respondents reported “don’t know” or “neither” to the survey questions.

The Brazilian case is also useful for research because the Brazilian government collects and disseminates data on a variety of topics, providing indicators of electoral and bureaucratic representation, municipal policy outcomes, and relevant municipal-level control variables. Three electoral institutions are relevant for the representation of women at the municipal level: mayor, vice-mayor, and local legislators (akin to city councilors in the United States). Elections for these three elected offices are held simultaneously (most recently October 7, 2012), and candidates for mayor and vice-mayor run on the same party ticket. The size of the city council varies depending on the size of the municipality in concordance with Article 29 of the 1988 Constitution and can range anywhere between 9 and 55 members. During the time period used for this study (2013), women held about 12% of mayoralties, 15% of vice-mayoralties, and composed between 0% and 67% of city councils, with an average of 14% women.

In addition to these elected positions, women can also be selected to head different municipal public agencies (often called municipal secretaries). In this study, we examine women’s representation in two of these agencies: the social assistance agency and the public health agency. In 2013, women headed about 82% of municipal social assistance agencies and 51% of municipal health agencies. We also explore women’s representation in the municipal executive administration as a whole, as municipal elected officials can appoint people to these positions and these bureaucrats may exert influence over the policy decisions made by elected officials.⁷ Finally, we use a survey conducted by the Brazilian federal government to measure municipal policy outputs related to women and women’s interests. In 2013, the government included several questions on its annual survey of municipalities (Pesquisa de Informações Básicas Municipais) that asked about the management of gender policy at the municipal level, including whether the municipality provides various services for women or has adopted certain women-friendly policies (see Table 3).⁸

In addition to our main variables of interest, we include controls for the mayor’s partisanship (classified as left, center, right, or unclassified based on the works of Mainwaring [1999], Power and Zucco [2009, 2012], and Rodrigues [2002])⁹ and whether the mayor is serving a second term in office.¹⁰ Brazil has a highly fragmented multiparty system (Mainwaring, 1999) and Brazilian politicians are known to switch parties frequently (Desposato, 2006). Unlike the two-party system in the United States, multiple political parties compete in Brazilian elections and many of these parties succeed in winning elections at various levels of government. At the time of this study, nearly 30 political parties were represented in executive offices at the municipal level (see Table A1 in the appendix). For this reason, it is more useful to classify mayors’ partisanship based on ideological dispositions, rather than actual party attachments.

We also control for other variables that might affect the representation of women in public administration or the adoption/implementation of gender policy, including the percentage of workers in the formal sector who are women,¹¹ the percent women in the electorate, the wage gap between men and women (calculated as the average income of men minus the average income of women in the municipality), the municipal population, average age of the electorate, percentage of the electorate with at least a high school level education, the average income of municipal residents, and the percentage of municipal residents who live in an urban area.

Empirical Analyses and Results

Table 2 presents the results of three models that test whether having more women in elected political positions increases the number of women in top-level management positions and in the municipal executive administration overall. Two executive positions are selected for analysis—social assistance and health care—because these agencies are likely to deal with women’s issues. The dependent variable in the first model is coded 1 if a woman is head of the social assistance agency and 0 if the head is a man. Model 1 suggests that having a woman vice-mayor and having a larger percentage of women on the city council increases the probability that the head of the municipal social assistance agency will be a woman. However, having a woman mayor does not have a statistically significant effect. For Model 2, the dependent variable is coded 1 if the head of the municipal health agency is a woman and 0 if the head is a man. Results from Model 2 suggest that having women in elected offices increases the probability that the health agency head will be a woman. This effect is statistically significant for all three elected positions: the mayor, the vice-mayor, and the percent women on the city council.

Table 2.

Does the Presence of Women in Elected Office Increase the Presence of Women Public Administrators?

	Model 1 (logit): Woman social assistance agency head	Model 2 (logit): Women health agency head	Model 3 (OLS): % women in municipal executive administration
Woman Mayor	−0.157 (0.108)	0.162^† (0.084)	−0.931** (0.321)
Woman Vice Mayor	0.205^† (0.106)	0.154* (0.076)	−0.362 (0.291)
% Women City Council	0.007* (0.003)	0.005* (0.002)	0.009 (0.010)
Mayor Left Party	−0.239** (0.092)	0.011 (0.070)	−0.004 (0.268)
Mayor Right Party	−0.143 (0.091)	0.102 (0.067)	−0.132 (0.258)
Mayor Unclassified Party	−0.309^† (0.187)	0.080 (0.144)	0.242 (0.554)
Mayor Second Term	−0.003 (0.080)	0.048 (0.061)	−0.295 (0.233)
% Women Formal Workers	−0.005 (0.004)	0.003 (0.003)	0.278*** (0.011)
% Women Electorate	−0.012 (0.026)	0.049** (0.019)	−0.039 (0.073)
Wage Gap (Men − Women)	0.001 (0.000)	0.000 (0.000)	0.002^† (0.001)
Population (logged)	−0.170*** (0.044)	−0.176*** (0.034)	2.012*** (0.130)
Average Age of Electorate	−1.085*** (0.206)	0.105 (0.154)	−1.396* (0.591)
% Electorate High School Educated	−0.004 (0.006)	−0.012* (0.005)	−0.038* (0.018)
Average Income	−0.002*** (0.000)	−0.000 (0.000)	0.012*** (0.001)
% Urban Population	0.007** (0.002)	0.002 (0.002)	0.041*** (0.007)
Constant	8.718*** (1.258)	−1.053 (0.913)	31.496*** (3.502)

Pseudo-R²	.023	.012	.203
Observations	5,540	5,541	5,461

Note. Standard errors are in parentheses. OLS = ordinary least squares.

†

p = .10. *p = .05. **p = .01. ***p = .001.

In Model 3, the dependent variable is a continuous measure of the percent women in the municipal executive administration. Inconsistent with results from Models 1 and 2, we find that having more women in elected offices does not increase the percent women in municipal public administration. Rather, results suggest that having a woman mayor actually decreases the proportion of women in the executive administration. These results are interesting and suggest that the number of women in elected office may not always increase the number of women in the bureaucracy. However, the finding that women elected officials increase the likelihood that women will be appointed to top-level administrative positions is encouraging and supports the top-down representation argument that has been offered by the literature.

Although there is no direct positive influence of elected officials on the number of women in the bureaucracy, women politicians might have an indirect influence through their appointment of top managers (Models 1 and 2 in Table 2). To determine if women top managers are associated with greater increases in the number of women in the bureaucracy, we add dummy variables to Model 4 for both the gender of the executive in charge of social assistance and the executive in charge of health. The results in Table 3 show that the percentage of women in public administration increases when the agency head for health is a woman. The relationship between the gender of the social assistance executive and the percent women in the executive bureaucracy is positive but not statistically significant at conventional levels (p = .146). However, this lack of significance might be due to a lack of variation in the explanatory variable (82% of the social assistance executives are women). Overall Tables 2 and 3 provide evidence of an indirect influence of women politicians on the number of women in public administration via the appointment of women agency heads.

Table 3.

Do Women Top-Level Managers Increase Women’s Presence in the Municipal Executive Bureaucracy?

	Model 4 (OLS): % women in municipal executive administration
Woman Mayor	−0.936** (0.322)
Woman Vice Mayor	−0.404 (0.291)
% Women City Council	0.008 (0.010)
Woman Soc. Asst. Head	0.399 (0.275)
Woman Health Head	0.530* (0.210)
Mayor Left Party	0.010 (0.268)
Mayor Right Party	−0.138 (0.259)
Mayor Unclassified Party	0.249 (0.554)
Mayor Second Term	−0.294 (0.233)
% Women Formal Workers	0.277*** (0.011)
% Women Electorate	−0.045 (0.073)
Wage Gap (Men − Women)	0.002^† (0.001)
Population (logged)	2.042*** (0.130)
Average Age of Electorate	−1.358* (0.592)
% Electorate High School Educated	−0.036* (0.018)
Average Income	0.012*** (0.001)
% Urban Population	0.040*** (0.007)
Constant	30.689*** (3.537)

R ²	.204
Observations	5,456

Note. Results are from an OLS model. OLS = ordinary least squares.

†

p = .10. *p = .05. **p = .01. ***p = .001.

Next, we move to an analysis that tests whether having more women in elected and administrative positions increases the adoption and implementation of policies that are beneficial to women. Table 4 lists a number of women-friendly policies along with the number of municipalities in Brazil that have adopted such policies. We combine these policies using an additive scale to measure whether increasing women’s presence in electoral institutions and in public administration increases the representation of women’s interests in the municipality. All of the policies used in the additive scale load positively onto a single factor and are also positively correlated. Furthermore, Cronbach’s alpha, which is a measure of the internal consistency of a number of items included in a summative rating scale, provides a reliability coefficient of .798.¹² This suggests that the linear combination of these dichotomous policy variables can provide a useful measure of an underlying latent variable, such as the overall “woman friendliness” of the municipality.

Table 4.

Number of Brazilian Municipalities That Have Adopted Various Women-Friendly Policies.

Women-friendly policies	No. of adoptions
Municipality provides general services for women victims of violence	1,209
Municipality provides health services for women victims of violence	397
Municipality provides shelters for women victims of violence	141
Municipality has a participatory council for women’s rights	975
Municipal agency responsible for women’s policy promotes gender training in other areas of municipal government	626
Municipal agency responsible for women’s policy takes responsibility for certain subgroups of women	935
Municipality has a plan for developing women’s policy	250
Municipality develops policies for women in coordination with other agencies	2,138
Municipal agency responsible for women’s policy coordinates with other municipal bodies to implement policy	1,286
Municipality provides daycare services	3,767
Municipality provides birthing centers	428
Municipality executes policies for gender equality and women’s autonomy	1,030
Municipality has a budget devoted to implementing women’s policy	228
No. of municipalities in the dataset	5,563

Table 5 presents the results of this analysis. The dependent variable is a count of the number of women-friendly policies that have been adopted by the municipality (ranges from 0 to 13; mean = 2.41), so a negative binomial regression is used. Results from this analysis suggest that increasing women’s presence in both electoral institutions and in public administration may increase the overall woman friendliness of the municipality—or the municipality’s efforts to represent the interests of women. The analysis suggests that having a larger percentage of women on the city council and in the municipal executive administration, and having a woman social assistance agency head, increases the number of women-friendly policies that are adopted in the municipality.

Table 5.

Does the Presence of Women in Elected Offices and Public Administration Lead to the Adoption of Women-Friendly Policies?

	Model 5: No. of women-friendly policies adopted
Woman Mayor	0.038 (0.036)
Woman Vice Mayor	0.011 (0.032)
% Women City Council	0.002^† (0.001)
% Women in Municipal Executive Administration	0.004** (0.002)
Woman Social Assistance Agency Head	0.062* (0.031)
Woman Health Head	−0.018 (0.023)
Mayor Left Party	0.074* (0.030)
Mayor Right Party	0.004 (0.029)
Mayor Unclassified Party	0.053 (0.061)
Mayor Second Term	0.095*** (0.026)
% Women Formal Workers	−0.003* (0.001)
% Women Electorate	0.058*** (0.008)
Wage Gap (Men − Women)	−0.000 (0.000)
Population (logged)	0.330*** (0.014)
Average Age of Electorate	−0.373*** (0.066)
% Electorate High School Educated	0.007*** (0.002)
Average Income	0.000 (0.000)
% Urban Population	−0.001^† (0.001)
Constant	−4.111*** (0.408)

α	.271
Pseudo-R²	.074
Observations	5,456

Note. Results are from a negative binomial model.

†

p = .10. *p = .05. **p = .01. ***p = .001.

A glance at the control variables indicates the complexity of gender policy in Brazil; such policies are influenced by a wide range of factors other than gender representation including political ideology, the proportion of women voters, population size, whether the mayor is serving for a second term, and the average age and education level of municipal residents. Despite this wide range of additional factors, gender representation in public administration is positively related to policies that benefit women.

Discussion and Conclusion

Whether passive (i.e., symbolic, descriptive) representation leads to active (i.e., substantive) representation is central to the study of identity groups in politics and public administration. This article explores the case of gender identity within the context of Brazilian local governments, drawing on the theory of representative bureaucracy. Findings presented in this article suggest that the presence of women in elected offices can increase the numbers of women in public administration; however, this relationship may not hold for all administrative positions. We find that having a woman mayor, woman vice-mayor, or a large percentage of women on the city council increases the presence of women in top-level administrative positions, but does not directly increase the overall percentage of women in the municipal executive administration. Greater numerical representation of women in elected offices does increase gender diversity at the top of the bureaucracy, and that in turn influences the proportion of women in public administration. The influence is indirect rather than direct. Furthermore, we find that the presence of women in electoral institutions and in public administration increases the representation of women’s interests or the overall “woman friendliness” of the locale. In particular, having a larger percentage of women city councilors and public administrators, or having a woman as a top-level public manager, increases the adoption of policies that are beneficial to women.

Although the current study examines gender representation in public administration in a single country, it does bring some innovations to the literature that imply that representative bureaucracy can be a generalizable phenomenon. Unlike past studies of gender representation and public administration, Brazil is a developing country that is not known to be on the forefront of gender-related issues. The finding that women in public administration are associated with policies that benefit women in a context that is not an economically well-developed nation suggests that such studies might be fruitful in similar countries in Latin America, Africa, or Asia. The present study also stresses that women in public administration do not necessarily operate in a vacuum: The representation of women in elected offices is closely linked to bureaucratic representation in Brazil. This finding suggests that future research should probe the symbiotic relationship between elected and bureaucratic representation. Are these forms of representation substitutes for each other? Do they have a multiplicative effect where the combination is greater than the sum of its parts or is there a division of duties whereby the bureaucracy focuses on certain issues and permits the political branches to deal with other concerns? Future research should explore these questions as well as the applicability of representative bureaucracy theory to various country and cultural contexts.

Footnotes

Appendix

Table A2.

Summary Statistics and Sources of Data.

Variable	n	mean	SD	Minimum	Maximum	Source
No. Women’s Policies	5,565	2.410	2.452	0	13	IBGE, MUNIC 2013
% Women in Municipal Executive Administration	5,478	63.109	8.591	0	100	RAIS 2013
Woman Mayor	5,559	0.121	0.326	0	1	IBGE, MUNIC 2013
Woman Vice Mayor	5,565	0.153	0.360	0	1	TSE 2012
% Women City Council	5,564	13.564	11.147	0	66.667	TSE 2012
Woman Social Assistance Agency Head	5,557	0.822	0.382	0	1	IBGE, PMBAS 2013
Woman Health Head	5,557	0.512	0.500	0	1	IBGE, MUNIC 2013
Mayor Left Party	5,561	0.307	0.461	0	1	Authors; IBGE, MUNIC 2013
Mayor Right Party	5,561	0.345	0.475	0	1	Authors; IBGE, MUNIC 2013
Mayor Unclassified Party	5,561	0.041	0.199	0	1	Authors; IBGE, MUNIC 2013
Mayor Second Term	5,559	0.276	0.447	0	1	IBGE, MUNIC 2013
% Women Formal Workers	5,561	43.977	10.815	1.374	100	RAIS 2013
% Women Electorate	5,565	49.710	1.967	41.061	56.931	TSE 2012
Wage Gap (Men - Women)	5,546	120.179	128.742	−220	780	IBGE, Census 2010
Population (logged)	5,565	9.193	1.071	6.829	15.969	TSE 2012
Population	5,565	24891	149638	924	8,619,170	TSE 2012
Average Age Group of Electorate	5,565	3.934	0.202	3.236	4.706	TSE 2012
% Electorate High School Educated	5,565	30.245	9.785	5.904	73.324	TSE 2012
Average Income	5,546	557.146	121.479	200	1500	IBGE, Census 2010
% Urban Population	5,565	63.826	22.041	4.179	100	IBGE, Census 2010

Note. All data are from official government sources. IBGE = Instituto Brasileiro de Geographia e Estatística; MUNIC = Pesquisa de Informações Básicas Municipais; RAIS = Relação Anual de Informações; TSE = Tribunal Eleitoral Superior; PMBAS = Perfil dos Municípios Brasileiros Assistência Social.

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.

Notes

Author Biographies

Kenneth J. Meier is the Charles H. Gregory Chair in Liberal Arts and Distinguished Professor of Political Science at Texas A&M Univerisity. He is also a Professor of Public Management in the Cardiff School of Business, Cardiff University (Wales). His research interests include public management, representation, race and politics, and public policy.

Kendall D. Funk is a doctoral candidate in the Political Science department at Texas A&M University. Her research interests include gender and politics, local governments, decentralization, Latin America, and government performance.

References

Addati

Cassirer

Gilchrist

(2014). Maternity and paternity at work: Law and practice across the world. Geneva: International Labour Office.

American Psychological Association. (2006). Answers to your questions about transgender individuals and gender identity. Retrieved from http://www.apa.org/topics/sexuality/transgender.aspx

Andrews

Ashworth

Meier

K. J.

(2014). Representative bureaucracy and fire service performance. International Public Management Journal, 17, 1-24.

Avellaneda

C. N.

(2009). Municipal performance: Does mayoral quality matter? Journal of Public Administration Research and Theory, 19, 285-312.

Avellaneda

C. N.

(2013). Mayoral decision-making: Issue salience, decision context, and choice constraint? An experimental study with 120 Latin American mayors. Journal of Public Administration Research and Theory, 23, 631-661.

Block

J. H.

(1973). Conceptions of sex role: Some cross cultural and longitudinal perspectives. American Psychologist, 28, 512-526.

Brazil’s tough laws on violence against women stymied by social norms. (2015, May). The Guardian. Retrieved from http://www.theguardian.com/global-development/2015/may/12/brazil-tough-laws-violence-against-women-stymied-by-social-norms

Desposato

S. W.

(2006). Parties for rent? Ambition, ideology, and party switching in Brazil’s Chamber of Deputies. American Journal of Political Science, 50, 62-80.

Dolan

(2000). The senior executive service: Gender, attitudes and representative bureaucracy. Journal of Public Administration Research and Theory, 10, 513-529.

10.

Duerst-Lahti

Kelly

R. M.

(1995). Gender, leadership and development. Ann Arbor: University of Michigan Press.

11.

Eisinger

P. K.

(1982). Black employment in municipal jobs: The impact of black political power. American Political Science Review, 76, 380-392.

12.

Escobar-Lemmon

Taylor-Robinson

M. M.

(2014). Representation: The case of women. New York, NY: Oxford University Press.

13.

Eulau

Karps

P. D.

(1977). The puzzle of representation: Specifying components of responsiveness. Legislative Studies Quarterly, 2, 233-254.

14.

Falleti

(2010). Decentralization and subnational politics in Latin America. Cambridge, UK: Cambridge University Press.

15.

Funk

K. D.

(2015). Gendered governing? Women’s leadership styles and participatory institutions in Brazil. Political Research Quarterly, 68, 564-578.

16.

Gómez

E. J.

(2010). Friendly government, cruel society: AIDS and the politics of homosexual strategic mobilization in Brazil. In Corrales

Pecheny

(Eds.), The politics of sexuality in Latin America: A reader on lesbian, gay, bisexual and transgender rights (pp. 233-250). Pittsburgh, PA: University of Pittsburgh Press.

17.

Guy

M. E.

(1992). Women and men of the states: Public administrators at the state level. Armonk, NY: M.E. Sharpe.

18.

Heath

M. R.

Schwindt-Bayer

L. A.

Taylor-Robinson

M. M.

(2005). Women on the sidelines: Women’s representation on committees in Latin American legislatures. American Journal of Political Science, 49, 420-436.

19.

Henley

Arabsheibani

G. R.

Carneiro

F. G.

(2008). On defining and measuring the informal sector: Evidence from Brazil. World Development, 37, 992-1003.

20.

Htun

Power

T. J.

(2006). Gender, parties, and support for equal rights in the Brazilian Congress. Latin American Politics and Society, 48(4), 83-104.

21.

Jacobson

Palus

C. K.

Bowling

C. J.

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

22.

Jones-Correa

(2000). Bringing outsiders in: Questions of immigrant incorporation. In Wolbrecht

Hero

R. E.

(Eds.), The politics of democratic inclusion (pp. 75-103). Philadelphia, PA: Temple University Press.

23.

Keiser

L. R.

Wilkins

V. M.

Meier

K. J.

Holland

C. A.

(2002). Lipstick and logarithms: Gender, institutional context, and representative bureaucracy. American Political Science Review, 96, 553-564.

24.

Lewis

P. G.

Ramakrishnan

S. K.

(2007). Police practices in immigrant-destination cities: Political control or bureaucratic professionalism? Urban Affairs Review, 42, 874-900.

25.

Lim

(2006). Representative bureaucracy: Rethinking substantive effects and active representation. Public Administration Review, 66, 193-204.

26.

Mainwaring

(1999). Rethinking party systems in the Third Wave of Democratization: The case of Brazil. Stanford, CA: Stanford University Press.

27.

Marrow

H. B.

(2009). Immigrant bureaucratic incorporation: The dual roles of professional missions and government policies. American Sociological Review, 74(5), 756-776.

28.

Matland

R. E.

(1998). Women’s representation in national legislatures: Developed and developing countries. Legislative Studies Quarterly, 23, 109-125.

29.

May

P. J.

Winter

S. C.

(2009). Politicians, managers, and street-level bureaucrats: Influences on policy implementation. Journal of Public Administration Research and Theory, 19, 453-476.

30.

Meier

K. J.

Nicholson-Crotty

(2006). Gender, representative bureaucracy, and law enforcement: The case of sexual assault. Public Administration Review, 66, 850-860.

31.

Meier

K. J.

O’Toole

L. J.

(2006). Bureaucracy in a democratic state: A governance perspective. Baltimore, MD: Johns Hopkins University Press.

32.

Meier

K. J.

Smith

K. B.

(1994). Representative democracy and representative bureaucracy: Examining the top down and the bottom up linkages. Social Science Quarterly, 75, 790-803.

33.

Mladenka

K. R.

(1989). Blacks and Hispanics in urban politics. The American Political Science Review, 83, 165-191.

34.

Mosher

F. C.

(1982). Democracy and the public service. New York, NY: Oxford University Press.

35.

O’Donnell

G. A.

(1979). Modernization and bureaucratic-authoritarianism: Studies in South American politics. Berkeley: Institute of International Studies, University of California.

36.

Park

(2013). Does gender matter? The effect of gender representation of public bureaucracy on governmental performance. American Review of Public Administration, 43, 221-242.

37.

Peters

B. G.

von Maravic

Schroeder

(2013). Representative bureaucracy in action. North Hampton, MA: Edward Elgar.

38.

Pfau-Effinger

(1998). Gender cultures and the gender arrangement—A theoretical framework for cross-national gender research. Innovation: The European Journal of Social Science Research, 11, 147-166.

39.

Pitkin

H. F.

(1967). The concept of representation. Berkeley: University of California Press.

40.

Power

T. J.

Zucco

(2009). Estimating ideology of Brazilian legislative parties, 1990-2005: A research communication. Latin American Research Review, 44, 218-246.

41.

Power

T. J.

Zucco

(2012). Elite preferences in a consolidating democracy: The Brazilian legislative surveys, 1990–2009. Latin American Politics and Society, 54(4), 1-27.

42.

Riccucci

N. M.

Saidel

J. R.

(1997). The representativeness of state-level bureaucratic leaders: A missing piece of the representative bureaucracy puzzle. Public Administration Review, 57, 423-430.

43.

Rodrigues

L. M.

(2002). Partidos, ideologia e composição social [Parties, ideology and social composition]. Revista Brasileira de Ciências Sociais, 17(48), 31-47.

44.

Samuels

(2004). The political logic of decentralization in Brazil. In Montero

A. P.

Samuels

(Eds.), Decentralization and democracy in Latin America (pp. 67-93). Notre Dame, IN: University of Notre Dame Press.

45.

Tripp

A. M.

Kang

(2008). The global impact of quotas on the fast track to increased female legislative representation. Comparative Political Studies, 41(3), 338-361.

46.

Varsano

Mora

(2001). Fiscal decentralization and subnational fiscal autonomy in Brazil: Some facts of the nineties (IPEA Working Paper, 854). Rio de Janeiro, Brazil: IPEA.

47.

Vianna

A. R. B.

Carrara

(2010). Sexual politics and sexual rights in Brazil: A case study. In Corrales

Pecheny

(Eds.), The politics of sexuality in Latin America: A reader on lesbian, gay, bisexual and transgender rights (pp. 122-134). Pittsburgh, PA: University of Pittsburgh Press.

48.

Violence and women in Brazil: What happens indoors stays indoors. (2013, September 24). The Independent. Retrieved from http://www.independent.co.uk/voices/comment/violence-and-women-in-brazil-what-happens-indoors-stays-indoors-8809947.html

49.

Waiselfisz

J. J.

(2012). Mapa da violência 2012: Atualização: Homicídios de mulheres no Brasil [Map of violence 2012: Update: Homicide of women in Brazil]. Brasilia, Brazil: Centro Brasileiro de Estudos Latino-Americanos/Faculdade Latino-Americana de Ciências Sociais Brasil.

50.

Wilkins

V. M.

Keiser

L. R.

(2004). Linking passive and active representation by gender: The case of child support agencies. Journal of Public Administration Research and Theory, 16, 87-102.