Previous studies of commuting in urban areas mainly in North America and Europe during the past two decades, listed in Table 1, have confirmed that women commute times and distances shorter than men. In Japan, however, there has been few investigation about this topic so far. This paper, therefore, focuses on the differences of commuting patterns between men and women. The study area (Figure 1) is the Osaka Metropolitan Area (OMA). The data used here are obtained from the 3rd Person-trip Survey of the Keihanshin Metropolitan Area conducted in 1990. This data source includes information on age, residence, workplace and trip purpose of every subject trip-maker.
In the third section, the overall distributions of workers' residence and employment in the OMA for both sexes are outlined. With regard to residence of the workers, the rate of employed women shows an uneven distribution pattern and is highest in the CBD and declines toward the suburbs (Figure 2): it presents a great contrast to the one of employed men, whose distribution is relatively high and shows an even pattern across the study area. As for workplace of the workers, it can be seen that a large proportion of women are employed in the outer suburbs as well as in the CBD (Figure 3): on the other hand, a significant portion of the men's workplaces are located not only in the CBD but also in the inner city and the inner suburbs which are basically industrial areas.
The fourth section is devoted to a gender comparison of commuting patterns in terms of time, distance and trip mode. The results obtained (Figures 4 through 8) are consistent with the previous finding: women show shorter commuting times and distances than men. It also turned out that women tend to use walking/bicycle and not to use automobile.
A remarkable feature of female employment in Japan is the M-shaped curve (see Figure 9 for the OMA) for the relationship between labour participation rate and age, and the curve has two peaks for women in their twenties and fourties. These two peaks are thought to imply two different types of labour: whereas the former represents full-time jobs, the latter are part-time jobs, mainly due to the heavy responsibility of housework. When the female commuting trips of the groups in their 20's and 30-40's are selected, it is clear that the commuting distances and times of the former are much longer than the latter and similar to men's (Tables 2 and 3). Hence, a general statement that women commute shorter times and distances than men does not hold when women are subdivided into different age groups.
