What drives the gender-cycling-gap? Census analysis from Ireland

James Carroll,William Brazil,Bruno Morando,Eleanor Denny

doi:10.1016/j.tranpol.2020.07.007

Abstract

Cycling rates have been increasing in Ireland over the last ten years, but there is a large difference in male and female participation – only about a quarter of cyclists on Irish roads are female. This paper combines the latest census data with geospatial cycle lane data to explore the drivers of the gender-cycling-gap across 238 electoral districts in Dublin, Ireland. Our core hypothesis is motivated by previous literature which suggests that differences in female risk aversion could partly explain the gap. To test this hypothesis, we explore if areas with safer cycling routes to the city centre have relatively stronger effects for females, controlling for a range of area geographic and demographic factors. Both male and female bicycle participation is negatively correlated with an area's distance to the city, share of apartments and average income, and positively correlated with education. Comparing results across genders shows that the gender-cycling-gap is due to relatively larger negative effects for distance, income and apartments for females, which is partly offset by stronger positive education effects. Routes with very high shares of separated or off-road lanes (top quartile) have significantly higher cycling rates. This effect, although stronger for females, is not statistically different across genders. We highlight a number of gender-based policy recommendations related to cycling infrastructure, bicycle storage and bicycle supports (electric bicycles).

Highlights

Introduction and BackgroundIncreased cycling participation is a means to address a number of the social and environmental issues associated with a transportation system built upon the convenience of the internal combustion engine
We estimate two sets of models: in Model 1, we include the density of cycle lanes within a given electoral district (ED) acting as a proxy for cycle safety (“Lane Density”); in Model 2, we use the path density variable which accounts for lane density on a straight line between the centroid of the ED and the city centre (“Path Density”)
We divide the dependent variable by its mean so that male and female coefficients are directly comparable as percentage changes in the cycling share

Summary

Introduction

Households in Ireland, like most developed nations, are heavily dependent on the car for many household activities. As of 2016, 20% of Ireland’s greenhouse gas (GHG) emissions arose from the transport sector (CSO, 2018d), most of which is attributable to the private car. With this figure predicted to increase in the near future (EPA, 2018), there is an urgent need to transition towards more sustainable mobility practices, including walking, cycling and public transport. The latest Census figures (CSO, 2018b) show that 58.5% of Irish people commute to work by car, with a higher dependence for females (64.7% versus 53.1% - see Figure 1).

Results

Discussion

Conclusion