1. QUESTIONS

Countries such as the UK, where cycling levels are low, tend to see substantial demographic inequalities in cycling (Goel, Goodman, Aldred, et al. 2022). Analysing policy discourse and images from low-cycling Australia, Osborne and Grant-Smith (2017) find a “marked lack of diversity in depictions of cyclists [belying] claims about the importance of improving diversity in cycling”. Skewed visual images may reinforce inequalities in take-up (Andrews, Clement, and Aldred 2018), and/or wider stigmatisation of marginalised bodies, relationships, and identities. For instance, an award-winning 2006 campaign from Manchester, UK (still regularly cited positively online) contrasted a ‘Fast Lane’ for cyclists with a ‘Fat Lane’ for car users, representing fatness as undesirable.

Using a dataset gathered by the authors and sourced from Google UK’s Image search, this paper conducted exploratory research into the following questions:

  • What kind of family structures are represented in these 100 images of ‘family cycling’?

  • In terms of ethnicity, visible disability, and body size, how diverse are these 100 images of family members?

  • What kinds of cycling environments, clothing, and types of bike are depicted in the 100 images?

2. METHODS

On 13th November 2023, the lead author used a logged-out browser to search Google UK’s Image Search for “family cycling” (without the quotation marks) for results from the past year. (Google at that time provided search engines that were locally tailored, in this case with the co​.​uk domain). We saved the first 100 images, plus ten to allow for duplicates or images judged invalid. The image dataset reduced to 100 after removing duplicates (2 images), graphics and icons (7 images), and one image with no people. Both researchers coded the photos against a set of initially agreed categories relating to family structure, individual characteristics, and cycling environments. Discrepancies were discussed, with criteria refined to reduce ambiguity, and if agreement could not be reached, photos were coded as ‘unclear’ for a variable.

The study is exploratory because of the small number of images analysed and inevitable limitations, including being subject to ‘photographer bias’. Additionally, we are attributing characteristics such as gender identity and ethnicity (in broad terms) from images depicted. The aim is not to reflect self-identity but likely societal classifications. Further research could use techniques like focus groups with image prompts to explore the assumptions different audiences might make.

3. FINDINGS

Nearly 60% of images (58) showed two adults (Table 1). The next most frequent configuration (just over 25%) was one adult with one or more children (26 images). The typical number of children shown was two (63/100), while almost all the rest (31) showed one or three children. This suggests that most images were likely to be read as depicting a two-parent family, or if not, a single-parent family (or perhaps one member of a two-parent family); accompanied by one to three children. Images showing more than two adults tended to show cycling events.

Table 1.Numbers of adults and children in images
Number of children
Number of adults 0 1 2 3 4 Total
0 0 3 1 1 1 6
1 1 8 14 3 0 26
2 2 5 46 5 0 58
3 0 0 2 3 0 5
4 0 0 0 1 1 2
5 or more 1 1 0 1 0 3
Total 4 17 63 14 2 100

Applying a gender binary lens, of the 58 images with two adults, all showed (by our judgement) one man and one woman, while of the 26 images with a single adult, 11 were men and 15 women. This suggests that the cycling family is depicted as heterosexual, but the practice of cycling with children is shown as appropriate for men or women.

The images were coded for ethnicity, disability, and body size, again based on broad-brush, subjective judgements. No images showed anyone with a visible disability, while only 8 of 100 showed one or more people with what we called ‘larger bodies’. By comparison, a quarter of the English adult population is classified as obese, and two-thirds overweight or obese (Office for Health Improvement and Disparities 2024). 79 images showed only white people, 20 at least one person judged as non-white, and in one case this was unclear. Of the images with people of colour, around half (ten) were clearly from a non-UK context, for instance picture 9 (illustrating “How to Buy a Secondhand Bike”, from Cycling UK) apparently depicting two girls in a Latin American landscape containing cacti. By comparison, 20% of the English working-age population is non-white (Office for National Statistics 2025).

Turning to the cycling environment, 98 of 100 showed some greenery and only 18 buildings. 72 images showed people cycling somewhere separated from car traffic (like a forest trail). In 19 cases this was unclear, and in 9 cases they were on a street that permitted motor traffic. Categorising separation from pedestrians was much harder due to a lack of signage. Only four images, however, showed a route clearly separated from pedestrians. No images showed on-street infrastructure like cycle tracks separating cyclists both from pedestrians and motor traffic.

Only 3/100 showed any car traffic, with one more showing a car parked on a driveway. Defining pedestrians as someone likely to be travelling ‘on foot’ (including wheelchair or mobility scooter, although none were ever seen in these images), only 8-9 images showed any pedestrians (one was unclear). Most pedestrians shown were either parents accompanying very young children on bikes or were far away from cyclists, with only one image implying (not depicting) adult cyclists interacting with pedestrians. The lack of potential pedestrian interaction is notable given UK off-road cycle routes are usually shared with pedestrians.

Mostly, adults and children were dressed casually without high-visibility or cycle-specific clothing. Cycling was thus depicted more as a leisure pursuit than as a sporting activity, albeit requiring helmets. Specifically, 26 images (just over a quarter) depicted any unhelmeted cyclists (19 showed any unhelmeted adults, and 14 any unhelmeted children). Of the unhelmeted images, only 7/26 showed the UK, 11 were non-UK and the remaining 8 unclear. Hence the ‘typicality’ of helmet-wearing seemed strongest in a clearly UK context. The ‘leisure’ rather than ‘sport’ depiction was reflected in the bikes shown. Only 9 photos contained road bikes (drop handlebars and thin tyres, designed for speed), whereas all other photos featured hybrids (flat handlebars, a sturdier build than road bikes) or mountain bikes (flat handlebars with knobbly tyres and suspension).

The research suggests that families containing people with larger bodies, disabled people, people of colour, and LGBTQ+ parents and caregivers are rarely represented within the ‘cycling family’. Notably, no images out of the hundred analysed appeared to depict same-gender couples with children. Further research with a larger corpus of images from multiple sources would be useful given the limited and exploratory nature of this study.

Acknowledgements

This research was supported through an ongoing research collaboration with the climate action charity Possible, via the Car-Free Megacities project funded by KR Foundation. This study was devised and designed by the authors, and Possible and KR Foundation did not influence the study design, conduct, or reporting here.