Evolution of mobility patterns by gender and its transversalities in Fortaleza, BrazilAbstract
Discussions about gender have become relevant in the investigation of commuting patterns in Brazilian metropolises. Differences in mobility patterns between women and men are essentially influenced by the sexual division of labor, considering its transversalities with different social markers. This study investigates the evolution of gender differences in travel patterns in the Metropolitan Region of Fortaleza and their relationship with income and family structure between 1996 and 2019. Origin-destination household surveys and statistical tests validated the differences in each segment. In Fortaleza, with trends similar to those already observed in other state capitals, women have travelled more and for shorter distances, used fewer individual motorized modes of transport, and made more trips outside of peak hours.
Keywords
gender and transversalities; mobility patterns; sexual division of labor; income; family structure
Resumo
Discussões sobre gênero têm se tornado relevantes na investigação dos padrões de deslocamento nas metrópoles brasileiras. Diferenças nos padrões de mobilidade entre mulheres e homens são influenciadas essencialmente pela divisão sexual do trabalho, considerando suas transversalidades com diferentes marcadores sociais. Este trabalho investiga a evolução das diferenças relativas aos padrões de viagens por gênero para a Região Metropolitana de Fortaleza e sua relação com renda e arranjo familiar entre os anos de 1996 e 2019. Pesquisas domiciliares origem-destino e testes estatísticos foram considerados para validar as diferenças em cada recorte. Em Fortaleza, com tendências semelhantes às de outras capitais, mulheres têm se deslocado mais e por menores distâncias, usado menos modos individuais motorizados e realizado mais viagens fora do pico.
Palavras-chave
gênero e transversalidades; padrões de mobilidade; divisão sexual do trabalho; renda; arranjo familiar
Introduction
The production of knowledge about differences in mobility patterns between women and men has become important in the context of transport planning and urban mobility in Brazilian metropolises (Nunes and Pereira, 2020). Since the end of the last century, there has been an effort to understand changes in transport demand arising from women's growing insertion into the labor market, the distribution of household duties, and the complexity of women's travel patterns, since they still take on an important or full share of reproductive work (Hanson and Johnston, 1985; Rosenbloom, 1978). Understanding gender differences relative to the interactions shaped by urban structure and their implications for travel patterns therefore becomes fundamental within a new paradigm of integrated urban transport and land-use planning (Lopes et al., 2020).
Most studies on mobility differences between men and women explore travel patterns that are influenced by the sexual division of labor, characterized by the assignment of the productive sphere to men and the reproductive sphere (domestic chores and childcare) to women (Kergoat, 2000). Moreover, it should be stressed that these differences can be amplified or reduced when analyzed transversally alongside other social markers (Moutinho, 2014), such as income, family arrangement, among others (Kearl, 2014; Mauch and Taylor, 1997). However, research on this problem and its intersectional dimensions refers mostly to cities in the Global North (Silveira Neto, Duarte and Páez, 2015).
In Brazil, Svab (2016) and Macêdo et al. (2020) characterized the travel patterns of women and men in the cities of São Paulo, Salvador and Florianópolis. For São Paulo, the temporal evolution of these patterns was tracked through analyses of a sequence of household origin-destination (OD) surveys. However, a gap in knowledge still needs to be explored, especially regarding the characterization of the influence of intersectional dimensions on gender-differentiated mobility patterns, as well as the evolutionary analysis of this phenomenon in other Brazilian metropolises.
In this context, the general objective of this study is to investigate the evolution of gender differences in mobility patterns in Fortaleza, considering income and family- -arrangement segments, as well as to discuss the evolution of these differences by comparing the results obtained with those from other Brazilian cities. This objective is achieved initially through a literature review of the factors that influence mobility differences between women and men and how these factors relate to the social markers of income and family arrangement in the Brazilian urban context. From the literature review, the hypotheses under investigation in this study are drawn. Next, the analysis method is presented, including a description of the data used, as well as the definition and justification for choosing the indicators. The results obtained for each indicator are presented and discussed, along with a comparative analysis with other Brazilian cities. Finally, the conclusions and recommendations of this study are presented.
Travel patterns by gender and intersectionalities
Origins of gender differences in mobility
The concept of gender emerges from feminist women's movements and research, having been articulated and progressively contested and theorized, with the aim of consolidating a term that would be seen by the academic community as a category of analysis more scientific, neutral and objective than the category "women" (Scott, 1986). Despite the different approaches to conceptualizing "gender," all its meanings are grounded in Beauvoir's (1970, p. 11) observation that "one is not born, but rather becomes, a woman" (Haraway, 2004). Thus, being a man or being a woman goes well beyond a reflection of biological sexuality, but rather of social relations based on structures of power (de Moraes, 1998).
An important social issue at the root of gender inequalities is the sexual division of labor, characterized by the assignment of the productive sphere to men and the reproductive sphere to women, grounded in the principles of separation – there are "men's jobs" and "women's jobs" – and of hierarchization – productive work is worth more than reproductive work (Kergoat, 2000). This historical designation still has consequences in women's lower insertion into the productive market and their workload, as well as in their greater participation in household tasks.
Despite advances in women's insertion into the labor market, it is still marked by the existence of gender inequalities. Such differences, related to domestic roles, employability, occupation and income are reflected in distinct mobility patterns between men and women (Bowman and Ben-Akiva, 1996; Macêdo et al., 2020; Rosenbloom, 2004; Sermons and Koppelman, 2001; Svab, 2016).
Gender inequalities in the economy, as well as in the assignment of household tasks, therefore, reverberate into differences in gender mobility patterns. These differences can be further intensified by other "social markers of difference" (Moutinho, 2014), such as income class, race, caste, age, sexual orientation, ethnicity, disability, language or religion (Hera, 1998), making it necessary to investigate how these intersectional dimensions relate to gender inequalities.
Gender inequality and intersectional dimensions
Income class
Saffioti (2016, p. 111) points to the need to integrate the question of social classes into the discussion on gender from a Marxist inspiration, stressing the need for a "[...] theoretical treatment of the questions posed by the intersection between the mode of operation and the relations between the common determination of sex and the essential determination of the capitalist system of production – the division of society into social classes." For her, associating the concept of class with that of gender makes hierarchies explicit and shapes the unequal social relations between men and women (Saffioti, 2004). Income class and gender relations intertwine and animate all fields of the social, intervening in the structures of society, thinking of relations between the sexes as "classist" and class relations as "sexed" (Matos, 2009).
Two socioeconomic indicators, broken down by gender and income class, can be considered: the proportion of the employed population and the weekly workload of domestic labor. In Brazil at the start of the twenty-first century, more than half of women (56.2%) were in the poor class (up to 1 minimum wage), while for men this class corresponded to 37.9% of them. For classes above 1 minimum wage, the percentage of men was higher than that of women, indicating greater participation of men in the wealthier classes. There was also a tendency toward a decrease in gender inequalities by class as income increased, except for the wealthiest class (above 9 minimum wages), whose gender gap became larger than for the previous class (Pinheiro et al., 2009).
As for the weekly workload of domestic labor, it was found that the lower the income, the greater the burden of unpaid domestic chores for both genders. However, the number of work hours was greater for women in all income classes, and gender inequality in the distribution of household tasks was greater the lower the income (the 14.4-hour difference for the class up to 1 minimum wage fell to an average difference of 6 hours for the class above 8 minimum wages) (Pinheiro et al., 2009). This inversely proportional relationship between gender inequalities and income suggests the possibility that gender inequalities in other domains, such as mobility, may also widen at lower income levels.
Family arrangements
Family arrangements are directly linked to income dynamics in Brazil. Leone, Maia and Baltar (2010) state that Brazilian family composition has been undergoing changes related to demographic shifts – declining fertility and an aging population – and social shifts – fewer marriages, more separations, more family arrangements made up of mother and children, more people living alone, and greater prominence of women. Between 1995 and 2015, there is a trend toward a percentage decline in families formed by couples with children and an increase in the share of childless couples and single-person households. In addition, between 2005 and 2015, families shrank in size from an average of 4.2 to 3.7 residents per household (Ribeiro, Curi and Cassuce, 2019).
Pinheiro et al. (2009) found that, among male-headed families, the couple-with-children arrangement was the most representative in Brazil, while among female-headed families, the most relevant arrangement was that of woman with children. Regarding activity status and occupation, there was a higher percentage of families with an actively employed head when the head was a man than when the head was a woman, indicating lower labor-market participation among female heads of household relative to male heads, which could contribute to greater precariousness and social vulnerability for these families.
Furthermore, at the turn of the century in Brazil, women and men had different possibilities for contributing to family income, since this depended on how family life was structured, on how responsibility for household chores was divided among family members, as well as on the existence of public services that freed (generally the woman) from domestic chores and caregiving tasks for family members. Income by gender also depended on the labor market, which defined opportunities for individuals according to sex, age, schooling and position in the family hierarchy (Leone, 2000). Therefore, family arrangements that depend on the labor of women and young people were already more exposed to situations of instability and precariousness due to their lower labor- -market participation for reasons of gender and age (Goldani, 2002; Guimarães, 1994; Montali, 1991; Telles, 1992).
Age
The age/generation category, like other social categories, is also expressed as a marker of social power relations. Being young or being old is a situation experienced differently according to the gender and income class of individuals within each age group (da Motta, 1999). Brazil has been undergoing a population aging process in the face of rising life expectancy, associated with the phenomenon of the feminization of old age, with a growing number of women relative to men in the elderly population (Mesquita, 2017). According to Pinheiro et al. (2009), for all age categories, women's labor-force participation rate was, at the start of the twenty-first century, already lower than that of men and increased gradually for both genders from ages 10 to 44, before beginning to decline from age 45 onward. Income inequality for this indicator tended to increase as age increased, except for the last category (60 years or more), in which the difference in participation rate was smaller than in the previous category (45 to 59 years).
In terms of weekly time spent on domestic chores, it was already apparent that the workload fell on women across all age groups and tended to increase with age for both genders. Gender inequality in the assignment of household tasks tended to increase with age, considering the age range between 10 and 59 years. For the last category (60 years or more), however, the difference in weekly domestic workload was smaller than in the previous category (45 to 59 years) (Pinheiro et al., 2009).
It is therefore possible to show that age is a factor that influences gender inequality in Brazilian society, and it can increase or decrease consistently with age within the economically active age range, depending on the aspect analyzed. The elderly category (from age 60 onward), however, differs from the working- -age categories in the socioeconomic indicators, especially due to the greater share of retired people in that category. Thus, the differing levels of gender inequality in the economy and in income security by age can result in differentiated mobility patterns by gender and age, which can be widened or narrowed depending on the age category.
Race
For Crenshaw (1989), a leading author of Black Feminism and a pioneer of intersectional studies, it is increasingly necessary to analyze social processes by exploring gender and race, both by crossing these variables and by analyzing them separately. Pinheiro et al. (2009) also analyzed indicators on the Brazilian labor market, at the turn of the twentieth to the twenty-first century, broken down by race and gender, finding that white women had the lowest labor-force participation rate, while Black men had the highest value for this indicator. Moreover, inequalities in participation rate between Black men and women were greater than for white men and women.
Regarding the unemployment rate, in the 2000s the highest level of unemployment corresponded to Black women, followed by white women, Black men and white men, with gender inequality on this indicator being greater for Black people than for white people. Regarding average monthly income, the same hierarchy of categories was observed, with Black women at the lowest level and white men at the highest income level. In absolute terms, gender inequality in average monthly income was smaller for Black people than for white people (Pinheiro et al., 2009).
The weekly paid work hours were also lower for Black women, followed by white women, Black men and white men, indicating greater participation of Black women in part--time jobs. As for total work hours (including unpaid work), it was Black women who worked the most hours, followed by white women, Black men and white men. These gender inequalities in paid and unpaid working hours remained greater for Black people than for white people (Pinheiro et al., 2009).
It is thus apparent that most of the indicators showed, in the 2000s, a social hierarchy in which Black women occupied the lowest position, followed by white women, Black men and then white men, in terms of occupation and employability. Such gender inequalities generally widened for Black people relative to white people. In terms of income, the hierarchy remained with Black women at the bottom and white men at the top. However, it was Black men who ranked just above Black women, followed by white women; in this case, gender inequality was greater among white people (Pinheiro et al., 2009).
Evidence of distinct mobility by gender
Most studies that use gender as a category of analysis for mobility patterns have shown the same differences between the travel behavior of women and men across many countries, in both the Global North and South (Law, 1999; Peters, 2001; Rosenbloom and Plessis-Fraissard, 2009). The sexual division of labor is pointed to as the main determinant of these differences (Kergoat, 2000; Peters, 2001; Siqueira, Loureiro and Lopes, 2022), which can manifest in an amplified or attenuated form when crossed with the social markers of difference (Moutinho, 2014) – income, family arrangement, age and race.
According to Siqueira et al. (2022), there is evidence in Brazilian metropolises that gender mobility differences caused by the sexual division of labor are attenuated for high income relative to low income. According to Mauch and Taylor (1997), in the US context, the gender variation in trips serving children tends to be negatively correlated with income, such that women in low-income families make a much larger proportion of this type of trip.
Regarding family arrangement, Rosenbloom and Burns (1994) and Silveira Neto et al. (2015) point out, in both the Global North and South, that gender mobility differences are greater in families with children than in other family arrangements. Fan (2015) and Boarnet and Hsu (2015) show that the presence of children is the main contributing factor explaining gender differences in household- -serving travel. Moreover, the travel patterns of women and men in single-adult households without children appear to be more similar than in other types of arrangements (Mauch and Taylor, 1997). When two single adults marry, gender differences in travel increase: men lengthen their trips after marriage, while women reduce their trips to accommodate household responsibilities, even without the presence of children (McLafferty, 1997; Silveira Neto et al., 2015).
As for the social marker of age, according to Crane (2007), the smallest mobility differences between genders are observed among young people, while the largest differences are observed among adults. Moreover, mobility is more restricted for older women than for men of comparable age (Basarić et al., 2016). In terms of race, gender inequalities in employability and income and their respective mobility differences are greater among white people than among Black people (Huato and Woodbury Zeno, 2009). However, the similar employability between Black women and men is associated with their worse economic conditions. Specifically, white women earn much lower wages than white men, but the main reason is the high wages of white men (Kaufman, 2010). Thus, empirical research shows that gender differences in travel tend to be greater among white people and smaller among Black people (Crane, 2007; Hu, 2020; Mauch and Taylor, 1997; Johnston--Anumonwo, 2014; McLafferty, 1997; Preston and McLafferty, 2016).
The sexual division of labor therefore implies greater assignment of household tasks to women (Fenstermaker and West, 2002), resulting in more complex mobility behaviors, since, besides making trips for their own needs, women also take on the trips needed for food provisioning and household upkeep, as well as accompanying children or elderly people on trips to their respective activities (school and medical appointments, for example) (Crane, 2007; McGuckin and Nakamoto, 2005; Rosenbloom, 2004; Shirgaokar and Lanyi- -Bennett, 2020). Taylor, Ralph and Smart (2015) point out that, even in families where women earn more, have better education, or work more hours than their partners, they still make more trips to serve their children and go to the supermarket than men do.
Thus, gendered domestic roles may also explain women's lower labor-market insertion (Siqueira et al., 2022). Because of family responsibilities, women have lower labor-force participation and lower employment/occupation rates (Blau and Kahn, 2013) and, even when employed, are more likely to work part-time (Rosenbloom, 2004). The need for flexible work schedules leads to a greater tendency for women to be employed in lower-paying occupations and industries (Blau and Kahn, 2017). This, combined with persistent wage discrimination against women (Cobb-Clark, 2001; Correll, Bernard and Paik, 2007; Hedija, 2015), results in gender wage disparities, with women having lower average income than men.
Thus, in urban contexts in both the Global North and South, women's greater assignment of and responsibility for household duties, and the resulting lower female employability, lead women to make fewer work trips and more trips for other purposes (shopping, health, education, escorting) than men (Macêdo et al., 2020; Root and Schintler, 1999; Rosenbloom, 2004; Srinivasan and Rogers, 2005; Svab, 2016). Consequently, women make more trips than men (Crane, 2007; Hanson and Johnston, 1985; Rosenbloom, 1998; Rosenbloom and Burns, 1994; Svab, 2016). Furthermore, the greater variety of trip purposes translates into a greater diversity of origins and destinations, and a higher share of trips made during off-peak periods in women's daily travel profile than in men's (Rosenbloom and Plessis-Fraissard, 2009; Ng and Acker, 2018).
Moreover, women's lower average income means they have less access to owning and using vehicles (cars, motorcycles and bicycles) and use cheaper modes of transport, such as public transit or active mobility (Anand and Tiwari, 2006; Crane, 2007; Root and Schintler, 1999; Rosenbloom, 1998; Srinivasan and Rogers, 2005; Svab, 2016; Harkot, 2018; Macêdo et al., 2020; Siqueira et al., 2022). The combination of slower modes of transport associated with women's lower income, together with the burden of household responsibilities and their corresponding trips, therefore generates greater pressure on their time budgets, resulting in women traveling shorter distances than men (MacDonald, 1999; Rosenbloom, 2004) and traveling in a more chained rather than commuter-like fashion, so as to combine several trips into single journeys to cope with time constraints (McGuckin and Murakami, 1999; Root and Schintler, 1999).
It is worth noting that previous studies have considered different indicators, disaggregated by gender, to address the characteristics of trips: distance, modal split, duration, trip chaining, time of day, immobility, trip purpose and trip rate. There is significant variability in the indicators used to analyze the different mobility patterns by gender. However, results are similar, showing, in general, the same mobility differences between men and women across the various aspects analyzed.
Crane (2007) takes the analysis further and offers a characterization of gender mobility differences that also covers aspects of race, age and family arrangement, exploring the gender gaps (the percentage difference between female and male results for the indicator) in relation to intersectional aspects. The study found that the gender gap (a) is greater for white people than for Black people; (b) increases with age but falls for the elderly group (over 64 years); (c) is smaller for single people without children than for single people with children; and (d) is smaller for married people without children than for married people with children.
The analysis methods for the different indicators relating to gender mobility patterns vary according to the objectives of each study. When the objective is to characterize gender mobility differences, comparisons between means for a given confidence interval are generally used, to establish whether the difference between genders for the indicator is statistically significant.
When the objective is to diagnose gender mobility differences, the aim is to investigate hypotheses about causal relationships between explanatory factors and observed mobility patterns. Among the methods used in the literature presented in this review are multiple linear regression, random-effects panel regression, multinomial logistic regression, binary logistic regression, ordered probit model, combined log-linear regression, and structural equation modeling (SEM).
As a synthesis of this literature review, the diagram proposed in Figure 1 seeks to conceptually complement the problem represented by Macêdo et al. (2020), starting from social premises and combining relationships between the sexual division of labor and the markers of income and family arrangement with differences in gender mobility patterns.
Methodological proposal
In order to empirically investigate the hypotheses raised, the method proposed by Macêdo (2022) was applied to characterize the evolution of gender mobility pattern differences based on samples from the household OD surveys of the Metropolitan Region of Fortaleza (RMF), collected in 1996 (4,948 households, 20,270 individuals and 32,427 trips) and in 2019 (23,801 households, 66,336 individuals and 105,182 trips). First, a descriptive analysis of the samples was carried out for the categories, time periods and analysis variables (indicators). Then, statistical inference techniques were used to compare sample means (t-test) or proportions (Chi-square test) for each indicator across its stratifications, always with a significance level of 5%.
Figure 2 presents the indicators proposed to characterize the hypotheses on differences in gender mobility patterns. To analyze the social marker of income, hypotheses were raised that differences between men and women increase the lower the household income, categorized according to the 2019 Brazil Criterion. Regarding the social marker of family arrangement, evidence was sought that mobility differences between genders for single (or married) people without children are smaller than for single (or married) people with children, as well as that differences between married people are greater than between single people.
Besides the methodological contribution of including hypotheses regarding intensification of the aspects investigated from the social markers of income and family arrangement, it is worth noting that the indicators relating trip rate to employment status seek to advance methodologically relative to those used by Macêdo et al. (2020) and Svab (2016), since employability has a strong influence on the number of trips, with a tendency for the trip rate to increase for employed people. When using an indicator for mobility rate without distinguishing between employed and unemployed people, lower employability tends to decrease the trip rate for women, while household responsibilities tend to increase the number of trips, biasing the characterization.
Results and discussion
This section discusses the evidence found on the impacts of intersectional dimensions on gender mobility differences, as well as significant changes in the respective mobility patterns observed in the RMF over the last two decades, highlighting behaviors similar to or different from those observed in other Brazilian cities. Table 1 presents the results obtained for the indicators in Fortaleza, for the years 1996 and 2019, discussed in the following subsections. Cells filled in pink indicate that the hypothesis was supported the respective hypothesis under analysis for those years (Table 1).
Trip rate (trips per day)
Regarding the mobility of employed individuals, it is noted that there are no significant differences in trip rate between working women and men in 2019, whereas in 1996, men showed a trip rate 2% higher than women's, which contradicts hypothesis H1. Thus, for both employed groups in the years analyzed, the hypothesis was not supported the hypothesis of a higher trip rate for women. Still looking at Table 1, regarding unemployed people, women's trip rate is 4% higher than men's in 2019, while there was no significant gender difference in 1996. Hypothesis H2 is therefore confirmed for 2019, which can be explained by women's greater responsibility for household tasks and the corresponding trips for provisioning and caregiving.
Indicators H1 and H2 advance beyond the indicators used by Macêdo et al. (2020) and Svab (2016) of the average trip rate of individuals who made trips. Splitting the trip-rate indicator between employed and unemployed individuals seeks to isolate the effects of lower female employability (which tends to decrease their trip rate) from household responsibilities (which tends to increase the number of trips). Using a single indicator of average trip rate for people aged 15 or older who travel, without distinguishing between employed and unemployed people, Macêdo et al. (2020) obtained results that did not show significant differences in this indicator between genders for São Paulo/SP (2017), Salvador/BA (2012) and Florianópolis/SC (2014). Svab (2016) analyzed the average trip rate of those who made some trip, considering all ages, for São Paulo, between 1977 and 2007. Her results show that women's average daily number of trips became higher than men's starting in 1997.
Considering income (Figure 3), for employed women and men, in none of the income categories in the years analyzed was it possible to verify the hypothesis of a higher trip rate for women, at a 95% confidence level (CL). For unemployed people, however, women's trip rate is higher than men's for all income brackets in 2019, especially in the lower classes, corroborating the hypothesis regarding the income transversality.
As for family arrangement, for employed people, the t-test results (95% CL) indicate no significant differences in the indicator in either year. For unemployed people, women's trip rate is higher than men's for all types of family arrangement in 2019 (Figure 4). An increase in the differences is noted with the presence of children only for single people, so the hypothesis was not supported the hypothesis established regarding the effects of family arrangement.
Modal split
Results show that women walk more, use public transit more, and are car passengers more often than men (Table 1). Men, on the other hand, drive cars more and travel more by motorcycle and bicycle than women. Thus, the hypothesis was supported hypotheses H3 through H8 for both years of analysis. Similar results were obtained by Macêdo et al. (2020) in São Paulo (2007 and 2017), Salvador (2012) and Florianópolis (2014).
These results can be attributed to women's lower average income, which gives them less access to owning and using vehicles (cars, motorcycles and bicycles), so that they need to use cheaper modes of transport. These discrepancies point to travel patterns with shorter reach and, consequently, less access to opportunities for carrying out activities on the part of women, since the cheaper modes are also the slower ones.
As for temporal evolution, there was an increase in the percentage gender differences between 1996 and 2019 for most modes, except driving a car and cycling. This raises the hypothesis of increasing income inequality over the period of analysis, corroborating the observed results of increasing gender differences in modal split. Moreover, one may ask whether other factors could have contributed to the decrease in gender differences specifically for driving a car and cycling, such as a possible increase in the number of women licensed to drive or who have learned to ride a bicycle.
Comparing this temporal evolution with the result found by Macêdo et al. (2020) in São Paulo, between 2007 and 2017, a similar result is noted for public transit, with an increase in the percentage gender differences. For walking and driving, however, the trend results in São Paulo are opposite to those observed in Fortaleza, since there was a decrease in gender differences for walking and an increase for driving. One may therefore ask whether the trend in the evolution of gender differences in transport mode would also be affected by local factors, such as levels of congestion, trip distances, among others. Moreover, for a deeper understanding of the differences in trends between the two cities, it is relevant to look in detail at the methodological differences between the surveys in the two years of analysis in each city.
Considering income classes and the results of the Chi-square test (95% CL), the hypotheses are confirmed for all income classes in the years analyzed (Figure 5). A decrease is noted in the percentage gender differences between 1996 and 2019 for most modes in class AB, an increase in gender differences for walking and public transit in class C, and, in classes D and E, an increase in gender differences in almost all modes, except driving a car and cycling.
Regarding family arrangement (Figure 6), there are no gender differences in relation to mode, with greater use of public transit, walking and being a car passenger for women's trips in all categories, while driving cars, motorcycles and bicycles predominates among men. An increase is also noted in the percentage gender differences between 1996 and 2019 for walking, public transit and being a car passenger for families with children, considering both married and single people. Thus, there is evidence confirming the hypothesis of an increase in the gender gap in the presence of children.
Trip purpose
The distribution of trip purposes by gender corroborates the hypotheses established earlier (H9 through H11). It is noted that the proportion of trips for work purposes is higher for men, while the proportion for shopping purposes is higher for women in both years (1996 and 2019). The proportion of trips for the purpose of "escorting someone”, in 2019, is higher for women. This indicator could not be analyzed for 1996 because the OD database did not include this trip purpose.
Similar results were obtained by Macêdo et al. (2020) and Svab (2016) in São Paulo (1977, 1987, 1997, 2007 and 2017), Salvador (2012) and Florianópolis (2014), which can be attributed to women's greater assignment to household activities and their resulting lower labor-market insertion. It is worth stressing the importance of household OD surveys including this breakdown of trip purposes so that these gender differences intrinsic to travel purposes and needs can be analyzed. In OD surveys that do not include the purpose "escorting someone" or "serving a passenger”, it is possible that escort trips are allocated to other trip purposes that are not the purpose of the person doing the escorting, but rather the purpose of the person being escorted (for example, a mother's trip taking her child to school may be assigned the purpose "education" for the mother's trip, when this purpose belongs to the child's trip, not the mother's, whose purpose is only to escort the child).
In terms of temporal evolution, the percentage gender differences for the work purpose remained stable, while the differences for the shopping purpose decreased between 1996 and 2019. In São Paulo, Macêdo et al. (2020) and Svab (2016) show trends of decreasing gender differences for the work purpose due to increased women's participation in the labor market. The difference between the trends for the work purpose between São Paulo (decrease) and Fortaleza (stable) raises the hypothesis that women's insertion into the labor market in Fortaleza may have been smaller.
Moreover, Macêdo et al. (2020) and Svab (2016) also observed a trend of decreasing gender differences for shopping purposes, which raises the hypothesis of a greater sharing of responsibility for household upkeep between men and women over the period of analysis in both cities. For trips to "serve passengers" or "escort someone”, increases in gender differences were observed in São Paulo, which may have resulted from women's growing responsibility for escorting family members to their travel destinations, such as children to school or elderly people to the doctor, for example.
Considering the income segment for Fortaleza, the hypotheses are confirmed that women show a lower percentage of trips for work purposes than men, while they show a higher percentage of trips for shopping and "escorting someone." Moreover, comparing the gender differences in each income class, it is possible to validate the hypotheses that the differences increase the lower the income, except for shopping. The results thus point to a worsening, as income decreases, of household task responsibilities (Figure 7).
Regarding family arrangement (Figure 8), it is also noted that the proportion of trips for work purposes is higher for men, while the proportion for shopping purposes is higher for women in both years and for all types of family arrangement. Comparing the gender differences in each type of family arrangement in general, no trend of increasing gender differences in trip purpose is observed with the presence of children or marriage in the years and categories analyzed.
Total travel distance
Regarding average trip distance by gender (Figure 9) on the respective transport network, hypothesis H12 can be confirmed, namely that women's trip distances are shorter than men's in both years. Shorter distances for women were also found by Macêdo et al. (2020) and Svab (2016) in other Brazilian cities and other time periods. This result reinforces the hypothesis that women have a shorter reach in their trips, which occurs due to the use of slower modes and greater time constraints from household responsibilities, which increase the pressure on their time budgets.
The temporal evolution in Fortaleza shows an increase in distance differences over time. This result is the opposite of that obtained by Macêdo et al. (2020) and Svab (2016) for Euclidean travel distances in São Paulo. This difference may be related to the fact that in São Paulo there was a decrease in gender differences in trips for work purposes, while in Fortaleza there was a trend toward stability. In São Paulo, therefore, women's greater insertion into the labor market may have led to a larger increase in women's trip distances than in men's over the period of analysis, since work trips are the longest ones. Moreover, in São Paulo a decrease was observed in gender differences for walking (generally shorter than other modes), while in Fortaleza an increase was observed, contributing to greater differentiation in trip distances.
Considering the income segment, women's trip distances are shorter than men's across all income categories in the years analyzed, with the exception of women in classes A and B who, in 1996, had slightly longer trip distances than men of the same income class (Figure 9). There is also a general decrease, between the years analyzed, in trip distances, and an increase in the differences between men and women for all income categories, with a notable increase in this difference for class C, which showed the greatest evolution.
Regarding family arrangement, the results of the tests on trip distance by gender confirm the hypotheses for all types of family arrangement in 2019, though only for married people with children in 1996, with the gender differences having increased over the period of analysis for all family arrangements (Figure 10). No worsening of differences was observed overall with the presence of children, so the hypothesis was not supported the effect of this transversality on the phenomenon analyzed.
Time of day (off-peak trips)
Due to the lower proportion of trips for work purposes among women and their greater participation in part-time jobs, women are expected to have more schedule flexibility for making trips related to caregiving and household provisioning. The results presented in Table 1 show a higher proportion of women's trips in the off-peak period than men's in both years. Macêdo et al. (2020) obtained a similar result for the cities of Salvador (2012) and Florianópolis (2014), while for São Paulo (2017) the differences were not significant.
Hypothesis H13 is confirmed, and an increase is noted in the percentage gender differences between 1996 and 2019 regarding the proportion of off-peak trips. Since in Fortaleza the gender differences for the work purpose remained stable, and the other established cause for making off-peak trips is greater participation of women in part-time jobs, this raises the hypothesis that, over this period, there was an increase in the number of women in this type of job.
Considering the income segment, the hypotheses established are confirmed, and there is also an increase in the percentage gender differences between 1996 and 2019 for income brackets A and B, while they remained stable for class C and decreased for classes D and E. There are also, to a modest degree, larger differences in the lower income classes (Figure 11).
Regarding the proportion of off-peak trips by gender, it was possible to validate the hypotheses that women's proportions are higher than men's for all types of family arrangement in 2019, though only for people with children in 1996 (Figure 12). Moreover, the hypotheses were validated that the gender gap increases with the presence of children in 1996, but the opposite occurs in 2019. In neither year was an increase found in gender differences for off-peak trips in relation to married people.
Conclusions and recommendations
The general objective of this study, to investigate the evolution of differences in the mobility patterns of women and men in Fortaleza, based on income and family-arrangement segments, was satisfactorily achieved. The objective of discussing the evolution of such differences, comparing the results obtained with other Brazilian cities, was also achieved, which allowed for an understanding of the evolution of the phenomenon in the city investigated, evaluating the hypotheses based on the literature.
The analysis of OD survey data in Fortaleza, in 1996 and 2019, confirms most of the hypotheses established, in line with what has been observed in other Brazilian and foreign studies. It should be noted that comparing the evolution of this problem in Fortaleza with that of São Paulo showed differences in the evolutionary patterns of gender mobility. On one hand, in Fortaleza, there was a widening or stability of gender mobility pattern differences considering the aspects investigated. On the other hand, in São Paulo, there was a decrease in this problem, which may indicate that the effect of women's insertion into the labor market in Fortaleza was smaller than in São Paulo. This hypothesis cannot be verified in this study and remains a recommendation for future work.
Regarding trip-generation rate, no significant differences were observed between employed women and men. However, it was noted that unemployed women tend to make more trips than unemployed men, and that this aspect intensifies as income class decreases, corroborating the understanding that income constraints contribute to the worsening of gender mobility differences. This intensification of the aspects investigated in relation to declining income was also observed for modal split, with greater use of public transit and walking among women. Differences between men and women were also identified regarding work trips and off--peak trip rate. The same pattern was not found regarding family arrangement, which, in many of the aspects examined, did not show significant differences that would support the hypothesis that the presence of children may intensify mobility differences. In this case, the overlap of income with family arrangement could reveal the effect of this transversality on the aspects investigated.
Despite the relevant phenomenological contributions of this study to understanding the problem of gender inequalities in urban mobility in Fortaleza, its methodological limitations should be acknowledged. First, the analytical effort to characterize mobility pattern differences undertaken here was directed at verifying the hypotheses established for the Metropolitan Region of Fortaleza, without offering a diagnosis of their cause-and-effect relationships. To deepen the understanding of this problem, it is recommended to use the formulation and operationalization of the causal inference model of Siqueira et al. (2022), proposed as representative of Brazilian metropolises.
Furthermore, the literature recognizes other segments as intensifiers of differences in mobility patterns between men and women, especially race. However, the data analyzed here could not offer the necessary inputs to verify such effects, since the two household OD surveys for Fortaleza, like the great majority of surveys of this kind carried out in Brazil, do not include self-declared race information for individuals, making it impossible to analyze the impacts of this social marker on gender mobility differences using this important database.
It is therefore strongly recommended that all surveys on mobility demand data (not only household OD surveys) begin collecting information on race, since, as clearly shown by the literature, this social marker has significant influence on distinct mobility behaviors. The availability of mobility pattern data disaggregated by race becomes essential for analyses of gender inequalities and their intersectionalities in accessibility and urban mobility, allowing for properly grounded and effective public policies to address them, such as the National Urban Mobility Policy and its respective urban mobility plans for municipalities and metropolitan regions, which, in general, still consider the population from the idea of a universal subject (male, white and middle-class) and render invisible people who travel in different ways (Nunes and Pereira, 2020).
Acknowledgments
This work was carried out with the support of the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brasil (Capes) – Finance Code 001, as well as the Conselho Nacional de Desenvolvimento Científico e Tecnológico – CNPq (Process n. 404.754/2021-0).
References
- ANAND, A.; TIWARI, G. (2006). A gendered perspective of the shelter-transport-livelihood link: the case of poor women in Delhi. Transport Reviews, v. 26, n. 1, pp. 63-80.
- BASARIC´, V.; VUJICIC´, A.; SIMIC´, J.M.; BOGDANOVIC´, V.; SAULIC´, N. (2016). Gender and age differences in the travel behavior: a Novi Sad case study. Transportation Research Procedia, v. 14, pp. 4324-4333.
- BEAUVOIR, S. (1970). Segundo sexo São Paulo, Difusão Europeia do Livro.
- BLAU, F. D.; KAHN, L. M. (2013). Female labor supply: Why is the United States falling behind? American Economic Review, v. 103, n. 3, pp. 251-56.
- BLAU, F. D.; KAHN, L. M. (2017). The gender wage gap: Extent, trends, and explanations. Journal of economic literature, v. 55, n. 3, pp. 789-865.
- BOARNET, M. G.; HSU, H. (2015). The gender gap in non-work travel: the relative roles of income earning potential and land use. Journal of Urban Economics, v. 86, pp. 111-127.
- BOWMAN, J. L.; BEN-AKIVA, M. (1996). Activity based travel forecasting. In: Conference of activity based travel forecasting, Texas. Travel Model Improvement Program. Houston, US Department of Transportation and Environmental Protection, pp. 1-32.
- COBB-CLARK, D. A. (2001). Getting ahead: the determinants of and payoffs to internal promotion for young US men and women West Yorkshire, Emerald Group Publishing Limited.
- CORRELL, S. J.; BENARD, S.; PAIK, I. (2007). Getting a job: Is there a motherhood penalty? American Journal of Sociology, v. 112, n. 5, pp. 1297-1338.
- CRANE, R. (2007). Is there a quiet revolution in women's travel? revisiting the gender gap in commuting. Journal of the American Planning Association, v. 73, n. 3, pp. 298-316.
- CRENSHAW, K. (1989). Demarginalizing the intersection of race and sex: a black feminist critique of antidiscrimination doctrine feminist theory and antiracist politics. University of Chicago Legal Forum. Illinois, v. 1989, article 8, pp. 139-167.
- DA MOTTA, A. B. (1999). As dimensões de gênero e classe social na análise do envelhecimento. Cadernos Pagu. Campinas, n. 13, pp. 191-221.
- DE MORAES, M. L. Q. (1998). Usos e limites da categoria gênero. Cadernos Pagu. Campinas, n.11, pp. 99-105.
- FAN, Y. (2015). Household structure and gender differences in travel time: spouse/partner presence, parenthood, and breadwinner status. Transportation, v. 44, n. 2, pp. 271-291.
- FENSTERMAKER, S.; WEST, C. (2002). Doing gender, doing difference: inequality, power, and institutional change Nova York, Psychology Press.
- GOLDANI, A. M. (2002). Família, gênero e políticas: famílias brasileiras nos anos 90 e seus desafios como fator de proteção. Revista Brasileira de Estudos de População. Rio de Janeiro, v. 9, n. 1.
- GUIMARÃES, I. B. (1994). As articulações possíveis: família e reprodução em um segmento operário. Revista Brasileira de Estudos de População. Rio de Janeiro, v. 11, n. 2.
- HANSON, S.; JOHNSTON, I. (1985). Gender differences in work-trip length: explanations and implications. Urban Geography, v. 6, n. 3, pp. 193-219.
- HARAWAY, D. (2004). "Gênero" para um dicionário marxista: a política sexual de uma palavra. Cadernos Pagu. Campinas, n. 22, pp. 201-246.
- HARKOT, M. K. (2018). A bicicleta e as mulheres: mobilidade ativa, ge^nero e desigualdades socioterritoriais em Sa~o Paulo. Tese de doutorado. São Paulo, Universidade de São Paulo.
- HEDIJA, V. (2015). Do women really face wage discrimination on the labour market? An analysis using intra- -household specialization. Acta Universitatis Agriculturae et Silviculturae Mendelianae Brunensis, v. 62, n. 6, pp. 1279-1286.
- HERA - Health, Empowerment, Rights & Accoutability (1998). Direitos sexuais e reprodutivos e saúde das mulheres: ideias para ação. Nova York, IWHC.
- HU, L. (2020). Gender differences in commuting travel in the US: interactive effects of race/ethnicity and household structure. Transportation, v. 48, pp. 909-929.
- HUATO, J.; WOODBURY ZENO, K. (2009). Class, race, and the spousal income gap: the effects of family income, educational attainment, and race-ethnicity on the husband-wife income ratio in the United States, 1980, 1990, 2000. American Behavioral Scientist, v. 53, n. 2, pp. 261-275.
- JOHNSTON-ANUMONWO, I. (2014). "Women's work trips and multifaceted oppression". In: ASUMAH, S. N.; NAGEL, N. Diversity, social justice and inclusive excellence: transdisciplinary and global perspectives. Albany, State University of New York Press.
- KAUFMAN, R. L. (2010). Race, gender, and the labor market: inequalities at work. Boulder, Lynne Rienner Publishers.
-
KEARL, H. (2014). Unsafe and harassed in public spaces: a national street harassment report. Disponível em: https://www.stopstreetharassment.org/wp-content/uploads/2012/08/2014-National-SSH-Street-Harassment-Report.pdf Acesso em: 15 ago 2023.
» https://www.stopstreetharassment.org/wp-content/uploads/2012/08/2014-National-SSH-Street-Harassment-Report.pdf - KERGOAT, D. (2000). Division sexuelle du travail et rapports sociaux de sexe. Dictionnaire critique du féminisme. Paris, v. 2, pp. 35-44.
- LAW, R. (1999). Beyond 'women and transport': towards new geographies of gender and daily mobility. Progress in human geography. Londres, v. 23, n. 4, pp. 567-588.
- LEONE, E. T. (2000). "Renda familiar e trabalho da mulher na Região Metropolitana de São Paulo nos anos 80 e 90". In: ROCHA, M. I. B. (coord.). Trabalho e gênero: mudanças, permanências e desafios São Paulo, Editora 34.
- LEONE, E. T.; MAIA, A. G.; BALTAR, P. E. (2010). Mudanças na composição das famílias e impactos sobre a redução da pobreza no Brasil. Economia e Sociedade Campinas, v. 19, n. 1, pp. 59-77.
- LOPES, A. S.; CAVALCANTE, C. B.; VALE, D. S.; LOUREIRO, C. F.G. (2020). Convergence of planning practices towards LUT integration: seeking evidence in a developing country. Land Use Policy, v. 99, 104842.
- MACÊDO, B. (2022). Evolução das diferenças nos padrões de mobilidade por gênero em Fortaleza. Dissertação de mestrado. Fortaleza, Universidade Federal do Ceará.
- MACÊDO, B.; PINTO, D. G. L.; SIQUEIRA, M. F.; LOPES, A. S.; LOUREIRO, C. F. G. (2020). Caracterização das diferenças no padrão de mobilidade de mulheres e homens em grandes cidades brasileiras. Transportes, v. 28, n. 4, pp. 89-102.
- MACDONALD, H. I. (1999). Women's employment and commuting: explaining the links. Journal of Planning Literature, v. 13, n. 3, pp. 267-283.
- MATOS, V. C. S. (2009). Um estudo histórico das relações de gênero e classe. Saber Acadêmico. Presidente Prudente, n. 7, pp. 57-73.
- MAUCH, M.; TAYLOR, B. D. (1997). Gender, race, and travel behavior: Analysis of household-serving travel and commuting in San Francisco Bay area. Transportation Research Record, v. 1607, n. 1, pp. 147-153.
- MCGUCKIN, N.; MURAKAMI, E. (1999). Examining trip-chaining behavior: Comparison of travel by men and women. Transportation Research Record, v. 1693, n. 1, pp. 79-85.
- MCGUCKIN, N.; NAKAMOTO, Y. (2005). Differences in trip chaining by men and women. In: Conference proceedings, v. 35, pp. 49-56.
- MCLAFFERTY, S. (1997). Gender, race, and the determinants of commuting: New York in 1990. Urban Geography, v. 18, n. 3, pp. 192-212.
- MESQUITA, A. A. (2017). Envelhecimento populacional e relações de gênero: velhos dilemas e novos desafios. In: SEMINÁRIO INTERNACIONAL FAZENDO GÊNERO 11 & 13 WOMEN’S WORLDS CONGRESS. Anais.. Florianópolis.
- MONTALI, L. (1991). Família e trabalho na conjuntura recessiva. São Paulo em Perspectiva. São Paulo, Fundação Seade, n. 5.
- MOUTINHO, L. (2014). Diferenças e desigualdades negociadas: raça, sexualidade e gênero em produções acadêmicas recentes. Cadernos Pagu. Campinas, n. 42, pp. 201-248.
- NG, W.; ACKER, A. (2018). Understanding urban travel behaviour by gender for efficient and equitable Transport Policies. In: INTERNATIONAL TRANSPORT FORUM. Discussion Paper 2018 - 01. Paris, 2018.
- NUNES, A.; PEREIRA, M. (2020). A ause^ncia das perspectivas de ge^nero e raça nas poli´ticas públicas de mobilidade urbana. RBDU - Revista Brasileira de Direito Urbanístico. Belo Horizonte, Fórum, v. 6, n. 10, pp. 189-206.
- PETERS, D. (2001). Gender and transport in less developed countries: a background paper in preparation for CSD-9 Londres, Paper Commissioned by UNED Forum.
- PINHEIRO, L.; FONTOURA N. O.; QUERINO A. C.; BONETTI A.; ROSA W. (2009). Retrato das desigualdades de gênero e raça São Paulo, Instituto de Pesquisa Econo^mica Aplicada.
- PRESTON, V.; MCLAFFERTY, S. (2016). Revisiting gender, race, and commuting in New York. In: AMERICAN ASSOCIATION OF GEOGRAPHERS, Washington, DC, Annals … v. 106, n. 2, pp. 300-310.
- RIBEIRO, M. P.; CURI, R. L. C.; CASSUCE, F. C. C. (2019). Desigualdade e estrutura familiar: uma análise comparativa. RDE - Revista de Desenvolvimento Econômico, v. 1, n. 42.
- ROOT, A.; SCHINTLER, L. (1999). Women, motorization and the environment. Transportation Research Part D: Transport and Environment, v. 4, n. 5, pp. 353-355.
- ROSENBLOOM, S. (1978). The need for study of women's travel issues. Transportation, v. 7, n. 4, pp. 347-350.
- ROSENBLOOM, S. (1998). Trends in women's travel patterns. In: Women's Travel Issues Second National Conference, 1998, Baltimore. Illinois, TRB, pp. 2-19.
- ROSENBLOOM, S. (2004). Understanding women's and men's travel patterns: the research challenge. Research on women's issues in transportation. Transportation Research Board of the National Academies, Washington, DC, Chicago, Illinois
- ROSENBLOOM, S.; BURNS, E. (1994). Why working women drive alone: Implications for travel reduction programs. Transportation Research Record, v. 1459, pp. 39-45
- ROSENBLOOM, S.; PLESSIS-FRAISSARD, M. (2009). Women's travel in developed and developing countries. Women's Issues in Transportation, v. 63.
- SAFFIOTI, H. (2004). Gênero, patriarcado, violência São Paulo, Fundação Perseu Abramo.
- SAFFIOTI, H. (2016). A mulher na sociedade de classes: mito e realidade São Paulo, Vozes.
- SCOTT, J. (1986). Gender: a useful category of historical analysis. The American Historical Review. Oxford, Oxford University, v. 91, n. 5, pp. 1053-1075.
- SERMONS, M. W.; KOPPELMAN, F. S. (2001). Representing the differences between female and male commute behavior in residential location choice models. Journal of Transport Geography, v. 9, n. 2, pp. 101-110.
- SHIRGAOKAR, M.; LANYI-BENNETT, K. (2020). I'll have to drive there: How daily time constraints impact women's car use differently than men's. Transportation, v. 47, n. 3, pp. 1365-1392.
- SILVEIRA NETO, R.; DUARTE, G.; PÁEZ, A. (2015). Gender and commuting time in São Paulo Metropolitan Region. Urban Studies, v. 52, n. 2, pp. 298-313.
- SIQUEIRA, M. F.; LOUREIRO, C. F. G.; LOPES, A. S. (2022). Diagnóstico das relações causa-efeito na mobilidade por gênero em grandes cidades brasileiras. Transportes, v. 30, n. 1.
- SRINIVASAN, S.; ROGERS, P. (2005). Travel behavior of low-income residents: studying two contrasting locations in the city of Chennai, India. J ournal of Transport Geography, v. 13, n. 3, pp. 265-274.
- SVAB, H. (2016). Evolução dos padrões de deslocamento na Região Metropolitana de São Paulo: a necessidade de uma análise de gênero Dissertaça~o de mestrado. São Paulo, Universidade de Sa~o Paulo.
- TAYLOR, B. D.; RALPH, K.; SMART, M. (2015). What explains the gender gap in schlepping? Testing various explanations for gender differences in household-serving travel. Social Science Quarterly, v. 96, n. 5, pp. 1493-1510.
- TELLES, V. S. (1992). Cidadania inexistente: incivilidade e pobreza. Tese de doutorado. São Paulo, Universidade de São Paulo.
Source: prepared by the authors, in 2026.
Source: prepared by the authors, in 2026.
Source: prepared by the authors, in 2026.
Source: prepared by the authors, in 2026.
Source: prepared by the authors, in 2026.
Source: prepared by the authors, in 2026.
Source: prepared by the authors, in 2026.
Source: prepared by the authors, in 2026.
Source: prepared by the authors, in 2026.
Source: prepared by the authors, in 2026
Source: prepared by the authors, in 2026.
Source: prepared by the authors, in 2026.