A cross-sectional study was performed on the basis of the results of the 2006 Daily Mobility Survey (EMQ2006) carried out by the Catalan regional government and the Metropolitan Transport Authority; the CATI method (computer-assisted telephone interview) was used to survey a representative sample of the population of Catalonia, a region in the northeast of Spain with approximately 7 million inhabitants. Individuals who reported having made any journey on the workday referred to in the interview (N = 23,424) were selected from the population of working individuals between 30 and 44 years. Those who declared that they did not make any journey on this day (N = 769), were excluded. For each individual, journeys made for work- and home-related reasons were selected.

Using multistage stratified sampling, the EMQ2006 collected data from throughout the territory of Catalonia. Data were collected, firstly, selecting the municipalities from transport zones, and then the individuals using a simple random sample.13 A correction system based on the population distribution was applied in each territory, in order to avoid over-representation of the least mobile individuals. In this way data regarding each of the 406,366 journeys made on the day before the interview by the 106,091 individuals surveyed, thereby obtaining representativity at the level of the 41 counties into which Catalonia is divided.

The dependent variables analysed were time, measured in minutes, spent per person per working day travelling for work-related and home-related reasons (daily shopping, non-daily shopping, personal affairs and accompanying others). Time only included the minutes spent going to the destination. The explanatory variables were educational level (none or primary, secondary, university), and household size not-including the interviewee (none, one, and two or more). Vehicle availability was used as an adjusting variable.

Gender and residential area were used as stratification variables, with residential area defined as Barcelona, the second belt and rest of Catalonia. Barcelona included the city and the first belt, formed by the 16 municipalities closest to the city, whose mobility is conditioned by the city. The second belt included the municipalities of the six counties neighbouring Barcelona, which, despite the proximity of many municipalities to the city, showed higher levels of self-containment (journeys with origin and destination within the same county or municipality). The rest of Catalonia showed patterns of mobility that were independent of Barcelona.14

A descriptive analysis of the time spent travelling according to type of journey (work- or home-related) was performed, using the mean and the corresponding 95% confidence intervals (95%CI), and median and the interquartile range. The mean time for each type of journey was calculated with respect to the total population of the study. Where individuals travelled for only one reason, their time spent travelling for other reasons was considered to be zero.

Percentage of men and women who spent more time travelling than the median was described for each reason of travel and according to educational level and household size, and p-value of the chi-square test was used in order to compare differences by gender.

A bivariate and multivariate logistic regression analysis was performed to determine the factors associated with greater time spent travelling according to reason for travel, with calculation of the odds ratio (OR) and 95%CI, adjusted for vehicle availability and stratifying by gender and residential area. Travel time was categorised as being above or below the median. In the case of time spent travelling for home-related reasons, the median time was zero, and thus the categorization corresponded to having travelled for these reasons or not.

All analyses were carried out using the statistical package STATA 10.0.15

A higher proportion of men reported have travelled for work-related reasons than women, 90.2% compared to 82.6% (p <0.001), and spend more time, 29.6 (95%CI: 28.9-30.2) and 23.2 (95%CI: 22.7-23.7) minutes respectively (p <0.001). Conversely, a higher proportion of women reported have travelled for home-related reasons and spend more time than men (42.5% compared to 19.3%, and 7.7 compared to 3.2 minutes; p <0.001) (Table 2).

The percentage of men who spent more time travelling for work-related reasons than the median was significantly greater than that among women in all residential areas and for all levels of education (Table 3). These percentages increased with educational level in both men and women and in all residential areas (Table 3).

Conversely, the percentage of women who spent any time travelling for home-related reasons was significantly greater than that among men in all residential areas and for all levels of education (p <0.001). The percentage in women decreased with increase in educational level in all three residential areas, being university-level women who statistically present lower proportion (ORUniversity= 0.75 [0.59-0.96]; ORUniversity= 0.84 [0.71-0.99] in Barcelona and the rest of Catalonia, respectively). Among men the reverse relationship was observed among residents of the rest of Catalonia, being university-level men who statistically present higher proportion (ORUniversity= 1.31 [1.07-1.61]).

For all three residential areas, only among individuals living with two or more people, the proportion of woman who spent more time on work-related journeys than the median travelling for work-related reasons was significantly minor than that among men (p <0.01). The percentages decreased as the household size increased, in both men and women and in Barcelona and in the second belt, approaching statistical significance only in women living with two or more people (OR = 0.72 [0.54-0.97]; OR = 0.38 [0.24-0.60, respectively] (Table 3).

Among both men and women from all three residential areas, those who lived with two or more people represented the greatest proportion of individuals who travelled for home-related reasons. However, this proportion was significantly greater among women (p <0.01).

In both men and women from all three residential areas, the only factor associated with greater time spent travelling for work-related reasons was higher educational level. But, only in women from the second belt, living with two or more people was associated with less time spent on work-related journeys (OR = 0.45; [0.27-0.73]) (Table 4).

In men, the only factor associated with greater time spent on home-related journeys was higher educational level and only among men living in the rest of Catalonia (ORSecondary= 1.21 [1.01-1.46], ORUniversity= 1.31 [1.06-1.61]). Conversely, in women, living with one more person was associated with longer time travelling for home-related reasons in the rest of Catalonia (OR = 1.68 [1.18-2.39]) and living with two or more individuals was associated in all residential areas (OR = 4.16 [2.71-6.38] in Barcelona, OR = 3.53 [2.09-5.96] in the second belt, OR = 3.20 [2.30-4.47] in the rest of Catalonia) (Table 4).

It is important to highlight the usefulness of the EMQ2006 for analysing the mobility of different social groups, since it allows us to simultaneously study journeys made throughout the day and for different reasons. It offers a very large sample, representative of both the urban and non-urban areas, and allows a high level of disaggregation.

One limitation of this study was that, when analysing journey time as a function of household size, it is not possible to know the characteristics of the co-habitants, nor the interviewee's position within the family unit. Thus, it was not possible to limit this study to married co-habiting individuals or to analyse how family responsibilities are distributed among working individuals. However, limiting the study to working individuals between 30 and 44 years of age allowed us to approximate the analysis to a population of productive and reproductive age, which makes the results obtained consistent with those of studies carried out only with married or co-habiting individuals.

Another limitation is the lack of data regarding the reason for the return journey. As a result, the times described may underestimate the real time spent on each type of journey throughout the day. Nevertheless, as the objective was not to estimate times but to evaluate their association with reason for travel, we believe that this will have a minimal impact on the conclusions of this study.

The results obtained are consistent with other studies of daily mobility.16 A study carried out in the metropolitan area of the city of Pamplona (Spain)17 analysed the daily mobility of individuals over 15 years of age using a questionnaire that collected data on all journeys made on the day before the interview. Among individuals between 30 and 50 years of age, men made more work-related journeys per day, and women made more journeys in the accompaniment of others and for shopping.

We are not aware of any study that has directly analysed time spent on work-related journeys according to educational level, although the study by Salom and Delios18 performed in region of Valencia (Spain) used the rate of inter-municipal mobility (percentage of individuals who travelled to work outside of their home town) as an indirect indicator of time spent travelling to work. This study showed that, in both men and women, the tendency to work outside the town of residence increased with educational level, which is therefore associated with longer work-related journeys. Similarly, Weinberger19 in a study carried out in the region of Philadelphia showed that time spent travelling to work increased with increase in salary in both genders. Both results are consistent with our analysis, in terms of greater time spent travelling to work among individuals with a higher educational level.

Some studies of time use, such as that carried out by The Women's Institute of Spain in 2006,20 indicated how inequalities in time spent on household tasks between men and women decreased in university-educated individuals explained by a reduction in time spent by women. Our study, although it was concerned only with travel time, finds significant differences between men and women for all educational levels. There was no significant reduction in time spent by university-educated women in any of the residential areas studied, while, on the contrary, the percentage of university-educated men who travelled for this reason was significantly greater in the rest of Catalonia.

To date, several studies have analysed the greater role of working women in care of the home as a function of household size and/or of dependent persons, self-perception of participation in household tasks and care of children, or through the study of time use,21,22 but to date few studies have explored the effect of this greater responsibility on mobility. Camarero et al.17 showed that the presence of young children in the home had no effect on work-related mobility for men, but reduced work-related mobility among women. They also found differences as a function of residence area. Our study obtained similar results for men and women resident in the second belt of the city.

Working men and women of 30-44 years in productive and reproductive age present very different mobility patterns. The fact that women continue to be mainly responsible for care of the home and family23 is well recognised, as well as the fact that this has consequences for their health (in terms of poorer perceived health,24 greater work leave,25 less physical activity,11 chronic disorders or poorer mental health26). However, the fact that caring for the home implies travelling27 is not always considered. In all residential areas, greater time for home-related reasons is spent by women who live with two or more individuals. On the other hand, time spent in domestic work decreases with increase in educational level, supporting the hypothesis that educational level also contributes to determining the amount of domestic and family work assumed by women, fundamentally because those with a higher socio-economic level have more resources for domestic or family work.11

However, as mentioned by Díaz-Muñoz et al.,2 it is also important to understand how daily travel is integrated into the sequence of activities that are carried out during the course of a day, since taking charge of the care of the home could affect other areas of daily life. Our results, together with previous work18 support this hypothesis since work-related travel among women is seen to diminish when they live with two or more individuals, which could be also a reflection of the fact that greater family responsibility could affect women's access to the labour market.28,29

Therefore, we may conclude that gender inequalities in mobility within the working population aged from 30 to 44 years are determined to a large extent by the greater responsibility of women in the domestic and family sphere, as well as by educational level. In contrast, daily mobility among men is primarily conditioned by educational level. This study reinforces the concept of mobility as a reflection of men and women's distinct work and families responsibilities. For future mobility surveys we recommend the introduction of new concepts that analyse social and economic aspects, such as the characteristics of the individuals in the household, since these could provide indispensible information to improve the study of mobility. Similarly, it is clear from this study that gender, work, family and social class must be taken into account when planning effective transport policies that facilitate reconciliation of work with family responsibilities, and which do not limit women's access to work or personal achievements, or that of more disadvantaged classes.