Representative Budgeting: Women Mayors and the Composition of Spending in Local Governments

Abstract

One potential consequence of increasing women’s numeric representation is that women elected officials will behave differently than their men counterparts and improve women’s substantive representation. This study examines whether electing women to local offices changes how local government expenditures are allocated in ways that benefit women. Using compositional expenditure data from more than 5,400 Brazilian municipalities over eight years, we find significant differences in the ways men and women mayors allocate government expenditures. Our findings indicate that women mayors spend more on traditionally feminine issues, and less on traditionally masculine issues, relative to men mayors. In regard to specific policy areas, we find that women spend more on women’s issues, including education, health care, and social assistance, and less on masculine issues, including transportation and urban development, relative to men mayors. We further find that women’s legislative representation significantly influences the allocation of expenditures as a larger percentage of women councilors increases spending on traditionally feminine issues, as well as education, health care, and social assistance, relative to other policy issues. These findings indicate that women local elected officials improve women’s substantive representation by allocating a larger percentage of expenditures to issues that have historically and continue to concern women in Brazil.

Keywords

women’s representation gender expenditures Brazil municipal finance

Women remain underrepresented in nearly all countries and all levels of government around the world.¹ However, women’s representation has increased in recent years. Over the past two decades, the percentage of women in national legislatures has nearly doubled, from an average of 12 percent in 1997 to 23 percent in 2017 (Inter-Parliamentary Union 2017). Women’s representation as presidents and prime ministers is improving as well (Jalalzai 2013), as is women’s representation in subnational governments, though the evidence is more limited (Escobar-Lemmon and Funk 2018; Sundström and Stockemer 2015; Vengroff, Nyiri, and Fugiero 2003). Women’s presence in these political institutions is vital for achieving the democratic ideals of fairness and equality. Even if women do not behave differently than men, their presence matters. Women’s presence can have symbolic consequences and shape important perceptions, such as views of women (Alexander 2012; Kerevel and Atkeson 2015), satisfaction with democracy (Schwindt-Bayer 2010), or trust in government (Hinojosa, Fridkin, and Kittilson 2017). Yet recent increases in women’s political presence beg the question of whether increasing numbers of women in political institutions has consequences for both policy and women’s substantive representation. If women do behave differently than men, the case for improving women’s numeric representation only strengthens, as electing women will result in different outcomes than electing men.

The implications of increasing women’s numeric representation extend beyond the national level of government. At the local level, increasing women’s representation could fundamentally reshape important outcomes including, but not limited to, the quality of women’s substantive representation. Local governments, especially those in decentralized countries, are responsible for important functions such as allocating government expenditures, collecting taxes, delivering public services, managing land usage, maintaining city infrastructure, and creating and implementing public policies. Given the role of local governments in overseeing these important functions and their potential to shape policies that matter to women, it is important to study the consequences of women’s representation within the local context.

This study offers several new innovations. First, we contribute to the growing body of knowledge about the consequences of women’s representation by examining whether increasing women’s local representation changes the spending priorities of local governments. We use panel data from Brazilian municipalities to examine whether women mayors allocate government expenditures differently than men mayors, and specifically whether women mayors spend more in policy areas that concern women citizens. Second, we argue that budget allocations reflect policy priorities. Given the constraints of local government budgets, allocating more money to one area often requires a decrease in another. If women mayors allocate more expenditures to policy areas that disproportionately affect women, this indicates that they are prioritizing the needs of women above other issues, and also engaging in substantive representation.

Using recent innovations in modeling compositional variables, we analyze expenditures in multiple policy areas simultaneously. The findings indicate that women mayors allocate a larger proportion of expenditures to traditionally feminine issues and a smaller proportion to traditionally masculine issues overall, compared with men mayors. Furthermore, these differences persist when analyzing more specific policy areas. Women mayors allocate more to education, health care, and social assistance, compared with men mayors, by decreasing the proportion of spending in areas such as transportation and urban development. We also find that the percent women city councilors impacts expenditure decisions despite the fact that the council has little direct influence over municipal budgets. A larger percentage of women councilors results in greater expenditures in traditionally feminine areas overall, as well as increases in education, health care, and social assistance spending at the expense of spending on administration and other areas. These findings suggest that increasing women’s representation as mayors and city councilors in Brazil will likely improve women’s substantive representation as women local officials are more likely to prioritize spending in areas that continue to disproportionately concern Brazilian women.

The Consequences of Women’s Representation

Whether electing women results in different consequences than electing men is the subject of a growing body of research. On the one hand, there might not be any significant consequences. The equality view of women’s representation suggests that if women are equal participants in the political process, there will not be observable differences in the policies enacted by men and women (Lovenduski 2005). Under conditions of full equality, women would not be sidelined to working in certain areas (Heath, Schwindt-Bayer, and Taylor-Robinson 2005), but instead able to focus on a diverse set of policy issues, as do men representatives, and participate equally in the governing process. This type of equality would result in no observable differences in the types of policies adopted by men and women policy makers.

On the other hand, research shows that political institutions are gendered institutions (Acker 1992), meaning that women elected officials often do not operate under conditions of full gender equality. Furthermore, a large body of evidence shows that women typically have different preferences and behaviors than men, and that women elected officials are more likely to act in the interests of women citizens.² In terms of preferences, women state legislators in the United States are more likely to prioritize issues that have traditionally concerned women, such as health care, social services, family and children’s issues, and the environment (Little, Dunn, and Deen 2001). Studies of Latin American legislators also indicate that women place greater priority on issues related to women, children, and families (Schwindt-Bayer 2010).

In terms of behaviors, research finds that women legislators participate more in debates about policy issues that concern women (Funk, Morales, and Taylor-Robinson 2017; Osborn and Mendez 2010; Taylor-Robinson and Heath 2003). They are also more likely to initiate bills related to women’s rights, children, and families (Jones 1997; Schwindt-Bayer 2010; Taylor-Robinson and Heath 2003) and social issues, including education, health care, and welfare (Schwindt-Bayer 2010; Swers 2014). Women chief executives are more likely to appoint women to top bureaucratic positions, such as cabinet ministers at the national level (Reyes-Housholder 2016) and public managers at the local level (Funk, Silva, and Escobar-Lemmon, 2017). A number of additional studies find that increasing women’s representation results in the adoption of policy programs that favor women’s substantive interests. Examples include the provision of child care in Norwegian municipalities (Bratton and Ray 2002), the implementation of women-friendly policies in Brazil (Meier and Funk 2017), greater gender equality in Sweden (Wängnerud and Sundell 2012), and the adoption of family leave policies in developed democracies (Kittilson 2008). All these studies provide evidence that women represent women.

Consequences for Government Spending

Gender differences in policy priorities and behaviors can also result in significant differences in the ways that women and men policy makers choose to allocate government resources. Indeed, there is evidence that women’s representation influences the distribution of government expenditures. Cross-national studies find that the percent women in national legislatures is associated with increases in social spending (Bolzendahl 2009), spending on family benefits (Bolzendahl 2011; Ennser-Jedenastik 2017), and spending on health and social welfare (Chen 2010), but associated with decreases in defense spending (Koch and Fulton 2011).³

Women’s representation has been shown to shape subnational government expenditures as well. There is evidence that increasing women’s presence in U.S. state legislatures results in more spending on health care (Courtemanche and Green 2017; Rehavi 2007) and less on corrections (Rehavi 2007). U.S. cities with women mayors are more likely to fund social welfare programs (Holman 2014) and allocate federal grants to issues like child care centers and services for children, youth, and abused spouses (Smith 2014). In India, increasing the representation of women from lower castes leads state legislatures to invest more in health and early education (Clots-Figueras 2011; Halim et al. 2016), and increasing women’s representation as village council heads results in greater investments in public goods that matter to women (Chattopadhyay and Duflo 2004). Further studies indicate that increasing women’s representation in local councils increases spending on child care and education in Sweden (Svaleryd 2009), social welfare spending in Taiwan (Chen 2013), and health care spending in Switzerland (Braendle and Colombier 2016).⁴

Yet findings from other studies suggest that gender representation does not significantly influence government spending. Studies of local governments in the United States (Ferreira and Gyourko 2014), Spain (Campa 2011), Italy (Rigon and Tanzi 2012), and Bavaria (Schild 2013) find that women’s representation in local governments has no impact on the composition of expenditures. Moreover, many studies that do find significant gender effects highlight the importance of factors that condition the relationship between gender representation and spending; for example, the power of representatives to act unilaterally (Holman 2014; Koch and Fulton 2011; Smith 2014), citizen spending demands (Courtemanche and Green 2017), intersecting identities such as social class (Clots-Figueras 2011; Halim et al. 2016), or the extent of women’s representation in other political offices (Holman 2014; Koch and Fulton 2011).

In sum, the literature suggests that women’s representation may significantly impact the way governments allocate resources, with greater representation yielding more spending in favor of women’s interests. However, the relationship between women’s representation and government spending may be conditional on other factors. Below, we develop theoretical expectations about the relationship between women’s representation in executive offices and expenditure allocations within the context of Brazilian municipalities. We explain why women mayors may—or may not—be more inclined than men mayors to allocate expenditures to certain policy functions, particularly those that benefit women as a group.

Representative Budgeting Theory

Government expenditures reflect policy priorities. If women’s representation significantly alters how expenditures are allocated, this indicates that women also change the priority that governments place on certain policy issues. In addition, modifying the allocation of expenditures often requires trade-offs as governments do not have unlimited funds to distribute. Thus, increasing expenditures in one area nearly always requires a decrease (at least as a proportion of the budget) in another area.

The distribution of government spending is also a form of representation. As opposed to other forms of representation that are rarer—for example, bill initiatives, political speeches, and policy adoptions—spending is an essential government function that occurs regularly. Government expenditures also matter for policy outcomes. Whereas some forms of representation may be purely symbolic or contingent on the availability of resources for proper implementation and enforcement, expenditures are likely to have a substantive impact across policy areas. Furthermore, small changes in allocations can have large effects over time, as government budgets often change incrementally (Wildavsky 1986). If women policy makers shift expenditures in one year, this change is likely to persist over time. Thus, the allocation of expenditures can provide a meaningful form of long-term representation for various groups in society, including women. This leads back to the question: when women are in power, do they allocate expenditures differently than their men counterparts? Do they prioritize issues that women care about? Prior research on gender representation and government spending offers mixed findings, with most studies concluding that the relationship is context dependent. In the context of Brazil, there are competing expectations about whether women policy makers will act differently than men.

First, less than 10 percent of mayoralties in Brazil are held by women, and the average city council has just 12 percent women, meaning that Brazilian local politics are still dominated by men. Thus, local governments may very well be “gendered institutions” (Acker 1992). Under conditions of gendered institutions, women representatives may desire to represent women, but may be constrained by the gendered rules and norms of the political office. The gendered nature of executive offices, in particular, might incentivize women executives to act “like men.” Whereas men are assumed to possess the qualities traditionally associated with a chief executive, such as assertiveness and competence (Eagly and Karau 2002; Stivers 2002), women must work to counteract stereotypes and combat credibility challenges associated with their gender (Funk 2015; Koch and Fulton 2011). Thus, the gendered nature of Brazilian local governments, and the local executive office in particular, may cause women mayors to behave similarly to men, resulting in no differences in expenditure allocations.

Second, there may be no observable differences in the distribution of expenditures because men and women mayors actually have similar spending preferences. The equality view of women’s representation posits that if women are truly equal participants in politics, they will not be sidelined to particular policy areas, but able to initiate changes across diverse areas (Lovenduski 2005). The few women who are successfully elected as mayors in Brazil—having won the most powerful position in local government—may have achieved equality and thus are able to work on a diverse set of issues similar to men. Women mayors might value women’s issues, but also aspire to represent all constituents, so they prioritize spending in both women’s and non-women’s policy issues (Schwindt-Bayer 2010), resulting in no notable differences in the ways men and women mayors allocate expenditures.

Null Hypothesis: There is no difference in the way men and women mayors allocate expenditures.

However, research still frequently finds significant gender differences in representatives’ attitudes toward women’s rights and equality issues (Schwindt-Bayer 2010; Taylor-Robinson and Heath 2003) and that women representatives are more likely to view women as an important constituency in need of representation (Childs 2004; Reingold 1992). Further studies find evidence that women prioritize issues that have traditionally concerned women, such as family and children’s issues, health care, social services, and the environment (Jones 1997; Little, Dunn, and Deen 2001; Poggione 2004; Swers 2014). Women representatives are also more active in traditionally feminine issues (e.g., health care, education, welfare), for example, by participating in debates (Funk, Morales, and Taylor-Robinson 2017; Osborn and Mendez 2010), initiating legislation (Schwindt-Bayer 2010), serving on committees (Baekgaard and Kjaer 2012; Barnes 2016; Heath, Schwindt-Bayer, and Taylor-Robinson 2005), and initiating policy programs (Bratton and Ray 2002; Meier and Funk 2017) in these areas.

Based on these findings from previous research, we might expect women mayors in Brazil to change the way government expenditures are allocated by providing greater funds to policy areas that disproportionately affect women. Women elected officials share the same gender identity as women in the general population; thus, they are likely to share similar socialization experiences, interests, and policy concerns as other women. Women mayors, like other women in Brazil, have experienced living in a gender-structured society that disadvantages women. As a consequence, women mayors may be more likely than men mayors to identify with women constituents as an underrepresented group and desire to represent their interests. One way they might do so is by allocating a greater proportion of expenditures to areas that concern women.

Hypothesis 1: Women mayors allocate more expenditures to women’s policy issues, relative to men mayors.

Yet previous research also suggests that women executives’ ability to represent women’s interests through the allocation of government spending may be conditional on the representation of women in the legislative branch (Holman 2014; Koch and Fulton 2011). In other words, women mayors might need a “critical mass” (Childs and Krook 2008; Kanter 1977) of women on the city council to enact any meaningful changes to local expenditures. As discussed in greater detail below, the city council is responsible for approving the mayor’s proposed budget and can directly and indirectly influence the mayor’s decisions (within certain limitations). If women mayors are considered to be “tokens,” and lack the support of the city council, they may find it difficult to initiate budgetary changes. However, if there are many women in government, it may be easier for women mayors to propose changes that favor women’s policy interests. Thus, we add nuance to our theory in the second hypothesis by proposing that the effect of women mayors may be conditional on the presence of women city councilors.

Hypothesis 2: Women mayors allocate more expenditures to women’s policy issues, relative to men mayors, when the percent women on the city council is sufficiently large.

Measuring Women’s Policy Issues

Research has defined “women’s issues” in various ways. The narrowest version defines women’s issues as those related to women’s rights and equality (Baldez 2011). Others use a broader definition that incorporates children and family issues, or pro-poor policies (Escobar-Lemmon, Schwindt-Bayer, and Taylor-Robinson 2014; Funk, Morales, and Taylor-Robinson 2017). The broadest definition incorporates issues that correspond to women’s traditional social roles and has been used to study gender differences in cabinet portfolio allocations (Escobar-Lemmon and Taylor-Robinson 2009; Krook and O’Brien 2012), legislative activities (Reingold 2000; Schwindt-Bayer 2010; Swers 2002), and committee assignments (Baekgaard and Kjaer 2012; Barnes 2016; Heath, Schwindt-Bayer, and Taylor-Robinson 2005).

There is a growing consensus that what counts as a women’s issue is context dependent and varies across time and space. Policy areas can be considered gendered if they disproportionately affect (i.e., benefit or harm) women as a group or can become gendered through the political process (Holman 2014; Keiser et al. 2002). We measure women’s issues in two ways and divide our analyses accordingly. First, we classify expenditure areas based on whether they correspond to women’s or men’s traditional domains. Second, we highlight three areas that continue to disproportionately affect Brazilian women: education, health care, and social assistance.

Traditional Feminine and Masculine Issues

We first classify issues according to whether they correspond to women’s or men’s domain, or neither. These classifications are rooted in traditional gender roles, with the Brazilian context in mind. Like elsewhere, Brazilian women have been primarily responsible for homemaking and child care, whereas men have been responsible for breadwinning and protecting the family. The Brazilian Treasury reports municipal government expenditures in twenty-eight categories. We classify education, health care, social assistance, culture, housing, environment, and citizenship rights as feminine issues as these correspond to women’s traditional roles in the domestic sphere as caregivers and nurturers.⁵ We classify urban development, transportation, agriculture, social security, sanitation, sports, energy, commerce, public safety, employment, industry, defense, science and technology, agrarian economy, and international relations as masculine issues as these correspond to men’s traditional roles in the public sphere as economic providers and defenders. The remaining categories are “unclassified” as they do not clearly correspond to either domain. For robustness, we checked these classifications against the existing literature (e.g., Escobar-Lemmon and Taylor-Robinson 2009; Krook and O’Brien 2012) and asked two scholars of gender in Latin America to independently classify each area, and then compared classifications and reconciled any discrepancies.⁶

It is important to note what we term feminine policies should not be mistaken for feminist policies, that is, those aimed at improving women’s rights and equality. Spending on feminine (or masculine) issues may incorporate feminist policies, but we cannot make these connections given the way expenditures data are reported.⁷ For example, we do know that spending on public daycares is included under education. Municipal child care provision provides women with greater opportunity to pursue full-time employment and better balance family and career; thus, spending on education does include feminist issues. However, we do not know whether other feminist concerns are incorporated in certain expenditure areas. Health care might include services for domestic violence victims or birthing centers. Housing might cover women’s shelters. Employment might include municipal initiatives aimed at diminishing wage disparities or sexual harassment. Yet we are unable to determine this given the limitations of the data. Although these classifications are imperfect, they do allow us to test whether women mayors prioritize spending in areas that have traditionally concerned women.

Table 1 presents these classifications, the average percent spent in each area, and the difference between men and women mayors. Spending on education, health care, administration, urban development, transportation, and social assistance constitutes the majority (more than 84%) of expenditures. Difference of means tests suggest there are significant gender differences in four of these areas: education, health care, social assistance, and transportation. Furthermore, women mayors spend around 2.4 percentage points more on feminine issues overall and 2.5 percentage points less on masculine issues overall, compared with men mayors. However, these significant differences may not hold after controlling for other variables. We address this further below.

Table 1.

Average Expenditures in Each Policy Area and Difference of Means by Mayoral Gender (N = 43,318).

Expenditure function	Traditionally	Average %	Women Mayors	Men Mayors	Difference
Aggregate feminine issues		57.122	59.328	56.914	2.414**
Education	Feminine	29.399	30.852	29.261	1.591**
Health care	Feminine	21.820	22.164	21.787	0.377**
Social assistance	Feminine	3.883	4.127	3.860	0.267**
Culture	Feminine	1.124	1.222	1.114	0.108**
Housing	Feminine	0.462	0.491	0.459	0.032
Environmental management	Feminine	0.396	0.432	0.393	0.040**
Citizenship rights	Feminine	0.039	0.0392	0.0394	−0.0002
Aggregate masculine issues		20.888	18.592	21.105	−2.513**
Urban development	Masculine	8.718	8.818	8.709	0.109
Transportation	Masculine	4.475	3.389	4.577	−1.189**
Agriculture	Masculine	1.966	1.601	2.001	−0.400**
Social security	Masculine	1.865	1.625	1.887	−0.263**
Sanitation	Masculine	1.600	1.254	1.632	−0.379**
Sports & leisure	Masculine	0.996	0.892	1.005	−0.113**
Energy & natural resources	Masculine	0.375	0.293	0.382	−0.089**
Commerce & services	Masculine	0.337	0.258	0.345	−0.087**
Public safety	Masculine	0.213	0.166	0.218	−0.052**
Employment	Masculine	0.161	0.144	0.163	−0.019
Industry	Masculine	0.157	0.126	0.159	−0.033**
National defense	Masculine	0.010	0.008	0.010	−0.002
Science & technology	Masculine	0.009	0.014	0.009	0.005**
Agrarian economy	Masculine	0.005	0.004	0.005	−0.001
International relations	Masculine	0.002	0.0013	0.0017	−0.0004
Aggregate unclassified issues		21.990	22.080	21.981	0.099
Administration	Unclassified	16.414	16.560	16.400	0.160
Legislative functions	Unclassified	3.253	3.291	3.249	0.042
Debt obligations	Unclassified	2.023	1.964	2.029	−0.064
Judiciary functions	Unclassified	0.172	0.161	0.173	−0.012
Communications	Unclassified	0.066	0.045	0.068	−0.023**
Essential to justice	Unclassified	0.063	0.060	0.063	−0.003

Difference = women mayors’ average expenditures – men mayors’ average expenditures.

p < .05.

Women’s Issues in Brazil

The second way we measure women’s issues is by accounting for how the Brazilian context affects the gendered nature of certain policy issues. In Brazil, many issues that have traditionally concerned women continue to disproportionately affect women today. We emphasize three areas—education, health care, and social assistance—and compare these to the other largest expenditure areas (urban development, administration, transportation, plus all other areas combined) as a second test of whether women mayors prioritize spending on women’s issues.

Education, health care, and social assistance are considered women’s issues in Brazil for multiple reasons. In addition to corresponding to women’s traditional roles, these issues continue to affect Brazilian women to a greater extent than they affect men. For example, women constitute the majority of employees in the education, health care, and social service sectors in Brazil. While just 1.21 percent of men employed in 2010 worked in education, around 3.5 times as many women held education-related jobs. Likewise, about six times as many women as men are employed in health care or social service occupations: 4.38 percent of employed women work in health care compared with 0.74 percent of employed men; 4.46 percent of employed women hold jobs related to health and social services compared with 0.75 percent of employed men (see the Supplementary Materials for tables supporting this section.)

Survey data also indicate that Brazilian men and women are affected differently by and have different attitudes toward these issues.⁸ With respect to health care, women report their health is worse than men on average and they frequently lacked medications. Women are also more likely to identify lack of health services as the most important problem in Brazil and believe health care should be a government spending priority. In contrast, men are more likely to agree with the statement “government should offer less public services, like health and education, to reduce taxes.” Regarding education, Brazilian women continue to be primarily responsible for child care and ensuring their children are educated. Women report they have more children than men and they worry about their children’s education. Women support increasing taxes to facilitate spending on secondary and higher education and more often identify lack of quality education as Brazil’s most important problem. Women also attend more parent–teacher meetings and disagree that a “university education is more important for a boy than a girl.”

Women are also disproportionately affected by social-assistance spending. In Brazil, “[f]emale-headed households are more likely to be in poverty at any point in time than male-headed households” (Barros, Fox, and Mendonça 1997, 231). Women report lower household and personal incomes, express dissatisfaction with their financial situation, and self-identify into a lower social class than men. Women are more likely to believe government does not do enough to combat poverty, report receiving Bolsa Família benefits, and favor expanding Bolsa Família. Women also frequently identify poverty, slums, or hunger as the most important problems in Brazil, and believe government should prioritize spending on “helping the poor.”

Thus, in Brazil, education, health care, and social assistance not only fall under women’s traditional domain but also continue to be important to women today. Women are more likely than men to be employed in these sectors, place greater priority on these issues, and are more affected by policy changes in these areas. Given these realities, increases in municipal spending on education, health care, or social assistance should benefit Brazilian women as a group, and consequently, these areas can be classified as women’s policy issues in the context of Brazil.

Local Government Budgeting in Brazil

Brazil is highly decentralized with 5,570 municipal governments (nested within twenty-six states plus a federal district). The 1988 Constitution deems municipalities’ autonomous political units with legislative and executive powers. While some municipalities have large populations, the vast majority (95% in 2010) have less than 100,000 residents, and around 50 percent have less than 11,000. Each municipality has a mayor and city council that are elected concurrently every four years. Voting is mandatory in Brazil, so voter turnout tends to be very high. Mayors are limited to one immediate reelection but can compete again after sitting out one term. Municipalities have a “strong mayor, weak council” form of government, known as executivismo, which is common throughout Latin America. This means mayors have significant policy-making powers and discretion. Federal legislation dating before the military coup and subsequent redemocratization gives mayors nearly complete control over local budgetary decisions (Federal legislation 4.320, March 1964).

The city council acts as an oversight body and must approve the mayor’s proposed budget by two-thirds vote (1988 Constitution, Art. 31, Paragraph 2). However, the council is limited in its ability to alter the mayor’s proposed budget (Wampler 2007, 50–51). Councilors are prohibited from introducing legislation that would require spending additional funds. They can pass budgetary amendments, but amendments cannot increase the budget’s size (they must reallocate from one area to another) and can be vetoed by the mayor. Moreover, amendments that are approved often alter a very small portion of the budget and may not actually be implemented. Furthermore, the mayor can increase spending between 5 and 40 percent on any budget item without prior council approval (Wampler 2007, 50).⁹ If the council fails to approve the mayor’s proposed budget, the budget from the previous year is reinstated.

Crafting the municipal budget is a very technical process. Transfers received from higher levels of government must be spent appropriately and the Constitution also sets spending floors for education and health care, and a spending ceiling for city councilors’ salaries. Municipalities must spend at least 25 percent of tax revenues in the maintenance and development of education (Art. 212), and at least 15 percent of all revenues on health care (Art. 198; Temporary Constitutional Provisions Act: Art. 77, Item III). The distribution of responsibility for different policy areas is also specified in the Constitution. All levels of government are jointly responsible for disability services; access to culture, education, and science; the environment; agriculture and food supply; housing and sanitation; and combating poverty (Art. 23, Item II). Municipalities are primarily responsible for infant and elementary education, health services, urban development, and preserving local historic and cultural heritage (Art. 30). Beyond these functions, municipalities also allocate expenditures to other areas, such as social assistance, energy, or public safety (see Table 1).

While the Constitution places some limits on municipal expenditures and specifies which policy areas municipalities should prioritize, mayors still have much control over discretionary expenditures. Mayors decide, with few limitations, which policy areas should receive more or less discretionary funds. “Institutional rules encourage mayors to initiate reform efforts and policies that reflect their interests; the rules concentrate legislative, budgetary, and administrative authority in the mayor’s office, which gives the mayor great flexibility to govern” (Wampler 2007, 48). Even in municipalities with participatory budgeting, mayors can manipulate the budget to advance their agenda (Wampler 2007, 65). Thus, if a mayor prioritizes a particular issue, she or he should be able to allocate expenditures to reflect this prioritization, so long as the constitutional and legal requirements are met.

Data, Variables, and Method

We use data covering two electoral terms (2005–2012) for more than 5,400 municipalities.¹⁰ The dependent variable we use is also known as a compositional variable, as each category in the composition (e.g., the proportion spent on masculine, feminine, and unclassified areas) sums to one. This means that increasing the share of expenditures in one category requires decreasing expenditures in at least one other category (the sum of the proportions must always equal one), which can make models of compositional dependent variables relatively complicated to estimate. The most common strategy is to model one category as a proportion of the total. However, using this method would prevent us from obtaining valuable information about trade-offs that occur simultaneously between policy areas; for example, the (1) feminine and masculine, (2) feminine and unclassified, and (3) masculine and unclassified categories.

To capture these trade-offs, we use a modeling strategy similar to the one proposed by Philips, Rutherford, and Whitten (2016a), which has been previously used to analyze budgets (e.g., Lipsmeyer, Philips, and Whitten 2017). Such estimation strategies are gaining popularity as they allow all components of a composition to be modeled simultaneously, meet standard regression assumptions by “unbounding” the compositions (which are naturally bounded between 0% and 100%), and lend themselves to graphical presentations. We use the additive log-ratio transformation to express each category relative to the others. This involves estimating the log-ratio of feminine and unclassified expenditures relative to the reference category (masculine) via the following models:

l n {(\frac{F e m i n i n e}{M a s c u l i n e})}_{i t} = f (\begin{array}{l} woman mayor, controls, \\ year fixed-effects \end{array})

l n {(\frac{U n c l a s s i f i e d}{M a s c u l i n e})}_{i t} = f (\begin{array}{l} woman mayor, controls, \\ year fixed-effects \end{array})

Above, the log-ratio of the feminine (equation (1)) or unclassified (equation (2)) proportions of total expenditures (relative to the masculine proportion) for municipality i in year t is a function of a dichotomous indicator for woman mayor, a set of controls, and year dummies. By taking the log-ratio transformation, we can estimate all three categories using the standard multivariate normal distribution (Philips, Rutherford, and Whitten 2016a). We also estimate seemingly unrelated regressions to account for contemporaneously correlated errors between models. Year fixed effects account for unobservable shocks common to all municipalities.

In terms of controls, we include the mayor’s age (logged), level of schooling (none = 1, graduate = 7), and whether the mayor’s party is left-leaning (from Power and Zucco 2009, 2012) as younger, better educated, and leftist mayors are more likely to care about gender equality (Inglehart and Norris 2003).¹¹ We also control for the mayor’s term in office (1 = second, 0 = first) and the mayor’s margin of victory as popular mayors may more easily implement changes. We control for the percent city councilors from the mayor’s party and percent women councilors, which also likely affect the mayor’s decisions. In addition, mayors with greater revenues per capita and more own-source revenue (vs. transfers) may have greater flexibility in modifying expenditures. Finally, we account for demands for spending in certain areas by controlling for municipal population size (logged), percent of the population in rural areas, percent women in the population, municipal human development index, percent living in poverty, average life expectancy, average years of schooling, average income per capita, percent with access to clean water and toilets, percent of women age 15 to 17 years with children, and the illiteracy rate.¹²

Analyses and Results

We first present results for the aggregated feminine, masculine, and unclassified areas, and then present results for the individual policy areas and women councilors.

Results for Feminine, Masculine, and Unclassified Areas

Results for the aggregate categories are presented in Table 2. Model 1 presents spending on feminine issues relative to masculine issues. Model 2 presents spending on unclassified issues relative to masculine issues. A positive (negative) coefficient indicates that the variable increases (decreases) the proportion of expenditures allocated to the numerator category at the cost of the denominator category. Thus, the positive and statistically significant coefficient on the woman mayor variable in model 1 suggests that women allocate more to feminine policy issues compared with masculine issues. Yet model 2 suggests that a woman mayor does not significantly affect the proportion of expenditures allocated to unclassified areas, relative to masculine ones. These findings lend support to our first hypothesis and suggest that women mayors allocate more to women’s issues at the expense of traditionally masculine issues.¹³

Table 2.

Women Mayors Prioritize Spending in Traditionally Feminine Areas.

	Model 1		Model 2
	$l n (\frac{F e m i n i n e}{M a s c u l i n e})$		$l n (\frac{U n c l a s s i f i e d}{M a s c u l i n e})$
	Coefficient	SE	Coefficient	SE
Woman mayor	0.0493***	(0.0131)	0.0322	(0.0208)
% Women councilors	0.0007**	(0.0003)	0.00001	(0.0005)
% Councilors mayor’s Party	0.0001	(0.0002)	−0.0003	(0.0004)
Left party	0.0455***	(0.0086)	0.0754***	(0.0137)
Second term	−0.0273***	(0.0080)	−0.0359***	(0.0126)
ln(age)	0.0394**	(0.0188)	0.0443	(0.0298)
Schooling	0.0028	(0.0023)	0.0028	(0.0037)
Win margin	−0.0008***	(0.0003)	−0.0016***	(0.0004)
ln(population)	0.1160***	(0.0051)	−0.0138*	(0.0082)
Revenues per capita	−0.00002***	(0.000003)	−0.000003	(0.000005)
% transfers	0.0076***	(0.0005)	0.0039***	(0.0007)
% rural population	−0.0020***	(0.0003)	−0.0026***	(0.0004)
% Women population	−0.0028	(0.0030)	0.0004	(0.0047)
Human development Index	0.0757	(0.240)	−0.421	(0.379)
% Poverty	0.0025***	(0.0008)	0.0025*	(0.0013)
Life expectancy	−0.0177***	(0.0029)	−0.0215***	(0.0046)
Average schooling	−0.0019	(0.0055)	−0.0482***	(0.0088)
Income per capita	−0.0004***	(0.00004)	−0.0002***	(0.0001)
% Clean water	−0.0050***	(0.0004)	−0.0017***	(0.0006)
% Young mothers	0.0063***	(0.0009)	0.0066***	(0.0015)
Illiteracy rate	0.0081***	(0.0009)	0.00005	(0.0014)
Constant	0.944***	(0.266)	2.309***	(0.422)
Observations	41,858		41,858
Municipalities	5,407		5,407
R ²	.23		.05
χ²	12,363.8***		2,337.8***

Seemingly unrelated regression with standard errors in parentheses. Year intercepts included but not shown.

p < .10. **p < .05. ***p < .01.

Since interpreting the substantive significance of the log-ratio results presented in Table 2 is not straightforward, we turn to calculating the expected proportion of expenditures allocated to feminine, masculine, and unclassified areas by women and men mayors, setting all control variables to their means. To do so, we estimate stochastic simulations using CLARIFY (Tomz, Wittenberg, and King 2003), then “un-transform” the resulting compositional predictions back into proportions as done by Philips, Rutherford, and Whitten (2016b).¹⁴ We then plot the predicted average percent spent in each area by men and women mayors and include 95 percent confidence intervals.

Figure 1 suggests women mayors allocate about one percentage point more to feminine issues than do men mayors, all else equal. This increase comes almost exclusively at the cost of masculine issues, since, while women mayors spend statistically significantly less on masculine areas relative to men mayors, there is no statistically significant difference with respect to spending on unclassified areas. This can also be seen in the middle plot presented in Figure 1, which shows that women mayors allocate less to masculine areas, while there is no statistically significant difference in men and women mayors’ allocations to unclassified areas. Overall, Figure 1 supports the hypothesis that women mayors allocate a larger portion of expenditures to areas that have traditionally concerned women compared with men mayors.

Figure 1.

Predicted allocations to traditional feminine, masculine, and unclassified issues.

To further grasp the substantive effects of mayoral gender on expenditures, consider a typical municipality in our sample: Ibiassucê, a municipality in the state of Bahia with around 10,000 inhabitants. It is typical in that all control covariates for this municipality fall within 0.75 standard deviations of the sample mean. Our estimates suggest that if Ibiassucê had a woman mayor instead of a man, there would be an increase in spending of around R$86,700 (Brazilian Reals) in feminine policy areas and a decrease of around R$75,400 and R$11,300 in masculine and unclassified areas, respectively, per year.¹⁵ Although relatively small in per capita terms, these differences can have notable impacts, especially in smaller municipalities where changes in investments or public works projects are very noticeable. Furthermore, these changes can add up to large differences over time given that budgets usually change incrementally from year to year.

Results for Individual Policy Areas

Next, we disaggregate expenditures to examine whether gender differences persist in specific policy areas. We examine seven areas: education, health, social assistance, transportation, urban development, administration, and an “other” category comprised of the remaining areas. We hypothesize that women mayors allocate a larger proportion of expenditures to the women’s policy issues identified above—education, health care, and social assistance—compared with men mayors, which may result in significant decreases in the percent allocated to traditionally masculine (urban development, transportation) and/or unclassified (administration, other) areas. For brevity, we plot the predicted average percent spent in each area, although the full table of log-ratio coefficients is available in Table S2 in the Supplemental Materials. The proportions allocated to education, health, and social assistance are shown in Figure 2, while the proportions allocated to transportation, urban development, administration, and other issues are shown in Figure 3.

Figure 2.

Predicted allocations to women’s policy issues.

Figure 3.

Predicted allocations to other policy issues.

In terms of spending on women’s issues, Figure 2 shows that women mayors allocate a larger portion of expenditures to education. Specifically, women mayors allocate about one percentage point more to education than do men mayors. Table S7 in the Supplemental Materials shows that this increase in education expenditures is the result of decreases in spending on transportation, urban development, health care, and “other” areas. Figure 2 also suggests there is not a statistically significant difference for spending on social assistance or health care as a percent of total expenditures, as the 95 percent confidence intervals for men and women mayors overlap. However, there are important gender differences when considering the trade-offs between different areas. Results suggest women mayors increase health care expenditures by decreasing expenditures to transportation, but they also decrease health care spending to increase education and social assistance spending, compared with men mayors (see Table S8 in the Supplemental Materials). Furthermore, women mayors increase spending on social assistance by decreasing spending on transportation, urban development, health care, and “other” areas (see Table S9 in the Supplemental Materials).

Figure 3 shows women mayors spend less on transportation compared with men mayors, but there are not statistically significant differences for urban development, administration, or other issues when analyzed as a proportion of total expenditures. Estimates suggest women mayors allocate around 0.1 percentage points less to transportation compared with men mayors. These findings all support our hypothesis that women mayors will allocate more expenditures to women’s policy issues and also suggest these increases often come at the expense of spending on traditionally masculine issues, such as urban development and especially transportation.

To better understand the size of these effects, consider these results in the context of Mirante da Serra, another typical municipality in the state of Rondônia with 14,000 residents. If Mirante da Serra had a woman mayor rather than a man, our estimates suggest education expenditures would increase nearly R$130,000 and transportation expenditures would decrease about R$55,050. Additionally, the estimates predict increases of around R$17,900 in social assistance and R$40,700 in administration, and decreases of around R$35,400 in health care, R$36,600 in urban development, and R$61,550 in other expenditures. Again, while these amounts may seem small in per capita terms, they could result in very noticeable differences in outcomes.

Results for Women Councilors

Although mayors have significant control over expenditures, the city council is responsible for approving the mayor’s proposed budget and may influence the mayor’s decisions directly or indirectly. Thus, if there are many women on the city council, we might expect mayors (both men and women) to be more likely to prioritize women’s issues. Estimates presented in Table 2 suggest that increasing the percent women councilors increases expenditures in feminine issues at the expense of masculine issues; however, women councilors do not significantly impact expenditures to unclassified issues. This finding suggests the presence of women councilors, like women mayors, leads to spending increases on issues that have traditionally concerned women.

Disaggregating expenditures further, we see similar results. More women councilors increase spending in all three women’s policy issues, education, health care, and social assistance, at the expense of spending on administration and “other” issues (see the Supplemental Materials). These findings provide further evidence that increasing the presence of women councilors significantly increases the proportion of expenditures devoted to women’s issues. These are especially notable findings given that the council does not have substantial control over expenditures and lacks power relative to the mayor.

Results for Women Councilors and Women Mayors

We also theorized that the effect of women mayors might be conditional on the percent women city councilors. We test this by interacting the women mayor variable with the percentage of women on the city council. For brevity, the results are summarized here but reported more fully in the Supplemental Materials. Across all our measures of women’s policy issues, we see that the effect of women mayors is not conditional on the percent women councilors. In other words, women mayors do not appear to need a “critical mass” of women on the city council before they are able to change the distribution of expenditures. However, we do find a significant conditional effect for men mayors. As the percent women councilors increases, men mayors become more likely to spend on feminine issues in general and on education in particular. In fact, when the percent women councilors is sufficiently large (around 28.57%), men mayors’ spending becomes statistically indistinguishable from women mayors’ spending. Overall, these findings suggest that women councilors do not have a significant effect on women mayors’ allocation decisions but may push men mayors to spend more on women’s policy issues, like education.

Discussion and Conclusion

This study adds to the growing body of research about the consequences of women’s representation by examining whether women’s presence in local elected offices influences the allocation of government expenditures in ways that favor women’s interests. Overall, our findings echo those of numerous studies illustrating the link between women’s descriptive and substantive representation. Our analysis of spending on traditional feminine and masculine policy issues reveals women mayors allocate a larger proportion of expenditures to traditionally feminine issues and a smaller proportion to traditionally masculine issues, compared with men. We also find that allocations to feminine issues increase as the percent women councilors increases. Second, disaggregating expenditures into specific policy areas, we find women mayors allocate more to women’s issues including education, health care, and social assistance by decreasing expenditures in masculine areas like urban development and transportation, and that increasing the percent women councilors also increases the proportions allocated to education, health care, and social assistance.

Furthermore, we show that women represent women, even in a context where representation is difficult to achieve. The municipal budgeting process is highly technical and constrained by legal and practical considerations. Women city councilors, in particular, have little formal say over budgeting decisions. In addition, the context of our analysis should be one in which men mayors are also inclined to spend on women’s issues as Brazilian municipalities are responsible for education and health care provision. Yet, even in this constrained context, we find that women elected officials better represent the interests of women.

Regarding our particular findings, we see that women mayors allocate even more to education than they do to either health care or social assistance. One potential explanation for these findings is the distribution of policy responsibilities among levels of government. While municipalities are primarily responsible for elementary education, the state and federal levels take on a larger share of responsibilities for health care and social assistance. Thus, women mayors may view education as a higher spending priority than other women’s issues that might be funded by higher levels of government. A second potential explanation is that women mayors, especially those with more “feminine” backgrounds (e.g., careers as homemakers or caregivers), may hesitate to prioritize social assistance and health care because it feeds into traditional stereotypes of women as softhearted, whereas spending on education is seen as an investment in human capital. There is not much evidence for this explanation, however, given that a woman mayor’s employment background does not appear to affect her spending decisions (as shown in the Supplemental Materials). A third explanation is that mayors are motivated by election concerns, and so women mayors allocate more to education because they expect women voters to reward them electorally for doing so. Finally, one of the most plausible explanations for our results is that women mayors spend more on education (and also on health care and social assistance relative to masculine issues) because they desire to represent women’s interests.

These findings also suggest that women elected officials have effects beyond the “rising tide” of gender equality that comes with slow-moving societal changes (Inglehart and Norris 2003). In a large way, women’s elected representation itself is a result of the rising tide, as changes in cultural attitudes and socioeconomics contribute to women’s elections. Yet it seems that women elected officials make “the tide rise faster,” as it were, as they improve gender equality in the representation of interests. Women mayors and city councilors promote more spending in areas that matter and can have direct consequences for the lives of women. In the context of education, for example, increasing spending on early childhood education helps women shoulder the responsibility of child care, which increases their employment options (Bratton and Ray 2002; Wängnerud and Sundell 2012). These results also underscore the importance of improving women’s numeric representation in Brazil and elsewhere as this study shows that electing women will result in different outcomes than electing men, and that women’s substantive interests will be better represented when women are present in governing bodies.

Supplemental Material

Funk_Philips_SI_Apr2018_online_supp – Supplemental material for Representative Budgeting: Women Mayors and the Composition of Spending in Local Governments

Supplemental material, Funk_Philips_SI_Apr2018_online_supp for Representative Budgeting: Women Mayors and the Composition of Spending in Local Governments by Kendall D. Funk and Andrew Q. Philips in Political Research Quarterly

Footnotes

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

Supplemental Materials

Supplemental materials for this article are available with the manuscript on the Political Research Quarterly (PRQ) website.

References

Acker

Joan

. 1992. “From Sex Roles to Gendered Institutions.” Contemporary Sociology 21 (5): 565–69.

Alexander

Amy C.

2012. “Change in Women’s Descriptive Representation and the Belief in Women’s Ability to Govern: A Virtuous Cycle.” Politics & Gender 8 (4): 437–64.

Baekgaard

Martin

Kjaer

Ulrik

. 2012. “The Gendered Division of Labor in Assignments to Political Committees: Discrimination or Self-Selection in Danish Local Politics?” Politics & Gender 8 (4): 465–82.

Baldez

Lisa

. 2011. “The UN Convention to Eliminate All Forms of Discrimination against Women (CEDAW): A New Way to Measure Women’s Interests.” Politics & Gender 7 (3):419–23.

Barnes

Tiffany D.

2016. Gendering Legislative Behavior: Institutional Constraints and Collaboration in Argentina. New York: Cambridge University Press.

Barros

Ricardo

Fox

Louise

Mendonça

Rosane

. 1997. “Female-Headed Households, Poverty, and the Welfare of Children in Urban Brazil.” Economic Development and Cultural Change 45 (2): 231–57.

Bolzendahl

Catherine

. 2009. “Making the Implicit Explicit: Gen-der Influences on Social Spending in Twelve Industrialized Democracies, 1980–99.” Social Politics 16 (1): 40–81.

Bolzendahl

Catherine

. 2011. “Beyond the Big Picture: Gender Influences on Disaggregated and Domain-Specific Measures of Social Spending, 1980–1999.” Politics & Gender 7 (1): 35–70.

Braendle

Thomas

Colombier

Carsten

. 2016. “What Drives Public Health Care Expenditure Growth? Evidence from Swiss Cantons, 1970–2012.” Health Policy 120 (9): 1051–60.

10.

Bratton

Kathleen A.

Ray

Leonard P

. 2002. “Descriptive Representation, Policy Outcomes, and Municipal Day-Care Coverage in Norway.” American Journal of Political Science 46 (2): 428–37.

11.

Campa

Pamela

. 2011. “Gender Quotas, Female Politicians and Public Expenditures: Quasi-Experimental Evidence.” Unpublished Manuscript.

12.

Chattopadhyay

Raghabendra

Duflo

Esther

. 2004. “Women as Policy Makers: Evidence from a Randomized Policy Experiment in India.” Econometrica 72 (5): 1409–43.

13.

Chen

Li-Ju

. 2010. “Do Gender Quotas Influence Women’s Representation and Policies?” The European Journal of Comparative Economics 7 (1): 13–60.

14.

Chen

Li-Ju

. 2013. “Do Female Politicians Influence Public Spending? Evidence from Taiwan.” International Journal of Applied Economics 10 (2): 32–51.

15.

Childs

Sarah

. 2004. New Labour’s Women MPs: Women Representing Women. New York: Routledge.

16.

Childs

Sarah

Krook

Mona Lena

. 2008. “Critical Mass Theory and Women’s Political Representation.” Political Studies 56 (3): 725–36.

17.

Clots-Figueras

Irma

. 2011. “Women in Politics: Evidence from the Indian States.” Journal of Public Economics 95 (7): 664–90.

18.

Courtemanche

Marie

Green

Joanne Connor

. 2017. “The Influence of Women Legislators on State Health Care Spending for the Poor.” Social Sciences 6 (2): 1–24.

19.

Eagly

Alice H.

Karau

Steven J.

2002. “Role Congruity Theory of Prejudice toward Female Leaders.” Psychological Review 109 (3): 573–98.

20.

Ennser-Jedenastik

Laurenz

. 2017. “How Women’s Political Representation Affects Spending on Family Benefits.” Journal of Social Policy 46 (3): 563–81.

21.

Escobar-Lemmon

Maria C.

Funk

Kendall D.

2018. “Women’s Representation in Subnational Governments: Replicating National Institutions or Forging a Unique Path?” In Gender and Representation in Latin America, edited by Schwindt-Bayer

Leslie A.

, 99–118. New York: Oxford University Press.

22.

Escobar-Lemmon

Maria C.

Schwindt-Bayer

Leslie A.

Taylor-Robinson

Michelle M.

2014. “Representing Women: Empirical Insights from Legislatures and Cabinets in Latin America.” In Representation: The Case of Women, edited by Escobar-Lemmon

Maria C.

Taylor-Robinson

Michelle M.

, 205–24. New York: Oxford University Press.

23.

Escobar-Lemmon

Maria C.

Taylor-Robinson

Michelle M.

2009. “Getting to the Top: Career Paths of Women in Latin American Cabinets.” Political Research Quarterly 62 (4): 685–99.

24.

Ferreira

Fernando

Gyourko

Joseph

. 2014. “Does Gender Matter for Political Leadership? The Case of U.S. Mayors.” Journal of Public Economics 1 (12): 24–39.

25.

Funk

Kendall D.

2015. “Gendered Governing? Women’s Leadership Styles and Participatory Institutions in Brazil.” Political Research Quarterly 68 (3): 564–78.

26.

Funk

Kendall D.

Morales

Laura

Taylor-Robinson

Michelle M.

2017. “The Impact of Committee Composition and Agendas on Women’s Participation: Evidence from a Legislature with Near Numerical Equality.” Politics & Gender 13 (2): 253–75.

27.

Funk

Kendall D.

Silva

Thiago

Escobar-Lemmon

Maria C.

Forthcoming. “Leading toward Equality: The Effect of Women Mayors on Gender Equality in Local Bureaucracies.” Politics, Groups, and Identities.

28.

Funk

Patricia

Gathmann

Christina

. 2015. “Gender Gaps in Policy Making: Evidence from Direct Democracy in Switzerland.” Economic Policy 30 (81): 141–81.

29.

Halim

Nafisa

Yount

Kathryn M.

Cunningham

Solveig A.

Pande

Rohini P.

2016. “Women’s Political Empowerment and Investments in Primary Schooling in India.” Social Indicators Research 125 (3): 813–51.

30.

Heath

Roseanna Michelle

Schwindt-Bayer

Leslie A.

Taylor-Robinson

Michelle M.

2005. “Women on the Sidelines: Women’s Representation on Committees in Latin American Legislatures.” American Journal of Political Science 49 (2): 420–36.

31.

Hinojosa

Magda

Fridkin

Kim L.

Kittilson

Miki Caul

. 2017. “The Impact of Descriptive Representation on ‘Persistent’ Gender Gaps: Political Engagement and Political Trust in Uruguay.” Politics, Groups, and Identities 5 (3): 435–53.

32.

Holman

Mirya R.

2014. Women in Politics in the American City. Philadelphia: Temple University Press.

33.

Inglehart

Ronald

Norris

Pippa

. 2003. Rising Tide: Gender Equality and Cultural Change around the World. New York: Cambridge University Press.

34.

Inter-Parliamentary Union. 2017. Women in Parliaments: World and Regional Averages. http://www.ipu.org/wmn-e/world.htm.

35.

Iversen

Torben

Rosenbluth

Frances

. 2006. “The Political Economy of Gender: Explaining Cross-National Variation in the Gender Division of Labor and the Gender Voting Gap.” American Journal of Political Science 50 (1): 1–19.

36.

Jalalzai

Farida

. 2013. Shattered, Cracked, or Firmly Intact? Women and the Executive Glass Ceiling Worldwide. New York: Oxford University Press.

37.

Jones

Mark P.

1997. “Legislator Gender and Legislator Policy Priorities in the Argentine Chamber of Deputies and the United States House of Representatives.” Policy Studies Journal 25 (4): 613–29.

38.

Kanter

Rosabeth M.

1977. “Some Effects of Proportions on Group Life: Skewed Sex Ratios and Responses to Token Women.” American Journal of Sociology 82 (5): 965–90.

39.

Keiser

Lael R.

Wilkins

Vicky M.

Meier

Kenneth J.

Holland

Catherine A.

2002. “Lipstick and Logarithms: Gender, Institutional Context, and Representative Bureaucracy.” American Political Science Review 96 (3): 553–64.

40.

Kerevel

Yann P.

Atkeson

Lonna Rae

. 2015. “Reducing Stereotypes of Female Political Leaders in Mexico.” Political Research Quarterly 68 (4): 732–44.

41.

Kittilson

Miki Caul

. 2008. “Representing Women: The Adoption of Family Leave in Comparative Perspective.” Journal of Politics 70 (2): 323–34.

42.

Koch

Michael T.

Fulton

Sarah A.

2011. “In the Defense of Women: Gender, Office Holding, and National Security Policy in Established Democracies.” Journal of Politics 73 (1): 1–16.

43.

Krook

Mona Lena

O’Brien

Diana Z

. 2012. “All the President’s Men? The Appointment of Female Cabinet Ministers Worldwide.” Journal of Politics 74 (3): 840–55.

44.

Latin American Public Opinion Project. (n.d.). AmericasBaro-meter. www.lapopsurveys.org.

45.

Lipsmeyer

Christine S.

Philips

Andrew Q.

Whitten

Guy D

. 2017. “The Effects of Immigration and Integration on European Budgetary Trade-Offs.” Journal of European Public Policy 24 (6): 912–30.

46.

Little

Thomas H.

Dunn

Dana

Deen

Rebecca E

. 2001. “A View from the Top: Gender Differences in Legislative Priorities among State Legislative Leaders.” Women & Politics 22 (4): 29–50.

47.

Lovenduski

Joni

. 2005. Feminizing Politics. Oxford: Polity Press.

48.

Meier

Kenneth J.

Funk

Kendall D.

2017. “Women and Public Administration in Comparative Perspective: The Case of Representation in Brazilian Local Governments.” Administration & Society 49 (1): 121–42.

49.

Osborn

Tracy

Mendez

Jeanette Morehouse

. 2010. “Speaking as Women: Women and Floor Speeches in the Senate.” Journal of Women, Politics & Policy 31 (1): 1–21.

50.

Philips

Andrew Q.

Rutherford

Amanda

Whitten

Guy D.

2016a. “Dynamic Pie: A Strategy for Modeling Trade-Offs in Compositional Variables over Time.” American Journal of Political Science 60 (1): 268–83.

51.

Philips

Andrew Q.

Rutherford

Amanda

Whitten

Guy D.

2016b. “Dynsimpie: A Program to Examine Dynamic Compositional Dependent Variables.” Stata Journal 16 (3): 662–77.

52.

Poggione

Sarah

. 2004. “Exploring Gender Differences in State Legislators’ Policy Preferences.” Political Research Quarterly 57 (2): 305–14.

53.

Power

Timothy J.

Zucco

Cesar

. 2009. “Estimating Ideology of Brazilian Legislative Parties, 1990–2005: A Research Communication.” Latin American Research Review 44 (1): 218–46.

54.

Power

Timothy J.

Zucco

Cesar

. 2012. “Elite Preferences in a Consolidating Democracy: The Brazilian Legislative Surveys, 1990–2009.” Latin American Politics and Society 54 (4): 1–27.

55.

Rehavi

M. Marit

. 2007. “Sex and Politics: Do Female Legislators Affect State Spending?” Unpublished Manuscript.

56.

Reingold

Beth

. 1992. “Concepts of Representation among Female and Male State Legislators.” Legislative Studies Quarterly 17 (4): 509–37.

57.

Reingold

Beth

. 2000. Representing Women: Sex, Gender and Legislative Behavior in Arizona and California. Chapel Hill: The University of North Carolina Press.

58.

Reyes-Housholder

Catherine

. 2016. “Presidentas Rise: Consequences for Women in Cabinets?” Latin American Politics and Society 58 (3): 3–25.

59.

Rigon

Massimiliano

Tanzi

Giulia

. 2012. “Does Gender Matter for Public Spending? Empirical Evidence from Italian Municipalities.” Working Paper No. 862, Temi Di Discussione, Bank of Italy. Rome, Italy. Available at SSRN: https://ssrn.com/abstract=2057858

60.

Schild

Christopher-Johannes

. 2013. “Do Female Mayors Make a Difference? Evidence from Bavaria.” IWQW Discussion Paper Series No. 07/2013. Nuremberg, Germany

61.

Schwindt-Bayer

Leslie A.

2010. Political Power and Women’s Representation in Latin America. New York: Oxford University Press.

62.

Smith

Adrienne R.

2014. “Cities Where Women Rule: Female Political Incorporation and the Allocation of Community Development Block Grant Funding.” Politics & Gender 10 (3): 313–40.

63.

Stivers

Camilla

. 2002. Gender Images in Public Administration: Legitimacy and the Administrative State. 2nd ed. Thousand Oaks: SAGE.

64.

Sundström

Aksel

Stockemer

Daniel

. 2015. “What Determines Women’s Political Representation at the Local Level? A Fine-Grained Analysis of the European Regions.” International Journal of Comparative Sociology 56 (3–4): 254–74.

65.

Svaleryd

Helena

. 2009. “Women’s Representation and Public Spending.” European Journal of Political Economy 25 (2): 186–98.

66.

Swers

Michele L.

2002. The Difference Women Make: The Policy Impact of Women in Congress. Chicago: The University of Chicago Press.

67.

Swers

Michele L.

2014. “Unpacking Women’s Issues: Gender and Policymaking on Health Care, Education, and Women’s Health in the US Senate.” In Representation: The Case of Women, edited by Escobar-Lemmon

Maria C.

Taylor-Robinson

Michelle M.

, 158–82. New York: Oxford University Press.

68.

Taylor-Robinson

Michelle M.

Heath

Roseanna M

. 2003. “Do Women Legislators Have Different Policy Priorities than Their Male Colleagues? A Critical Case Test.” Women & Politics 24 (4): 77–101.

69.

Tomz

Michael

Wittenberg

Jason

King

Gary

. 2003. “Clarify: Software for Interpreting and Presenting Statistical Results.” Journal of Statistical Software 8 (1): 1–30.

70.

Vengroff

Richard

Nyiri

Zsolt

Fugiero

Melissa

. 2003. “Electoral System and Gender Representation in Sub-National Legislatures: Is there a National—Sub-National Gender Gap?” Political Research Quarterly 56 (2): 163–73.

71.

Wampler

Brian

. 2007. Participatory Budgeting in Brazil: Contestation, Cooperation, and Accountability. University Park: Penn State Press.

72.

Wängnerud

Lena

Sundell

Anders

. 2012. “Do Politics Matter? Women in Swedish Local Elected Assemblies 1970–2010 and Gender Equality in Outcomes.” European Political Science Review 4 (1): 97–120.

73.

Wildavsky

Aaron B.

1986. Budgeting: A Comparative Theory of the Budgeting Process. New Brunswick, NJ: Transaction Publishers.

74.

World Values Survey. 2015. World Value Survey 1981-2014 Longitudinal Aggregate v.20150418, 2015. Aggregate File Producer: JDSystems Data Archive, Madrid, Spain. World Values Survey Association. www.worldvaluessurvey.org.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.27 MB