Paper on Cycling and Diversity

Recently we published a paper incorporating analysis conducted as part of the project, on cycling and diversity. The project has enabled us to do some new analysis of Census data, with some interesting new findings about what’s happened to age and gender balance, where cycling has increased. The paper is available open access here.

Does More Cycling Mean More Diversity in Cycling?

Rachel Aldred, James Woodcock & Anna Goodman
In low-cycling countries, cycling is not evenly distributed across genders and age groups. In the UK, men are twice as likely as women to cycle to work and cycling tends to be dominated by younger adults. By contrast, in higher cycling countries and cities, gender differences are low, absent, or in the opposite direction. Such places also lack the UK’s steady decline in cycling among those aged over 35 years. Over the past fifteen years some UK local areas have seen increases in cycling. This paper analyses data from the English and Welsh Census 2001 and 2011 to examine whether such increases are associated with greater diversity among cyclists. We find that in areas where cycling has increased, there has been no increase in the representation of females, and a decrease in the representation of older adults. We discuss potential causes and policy implications. Importantly, simply increasing cycling modal share has not proved sufficient to create an inclusive cycling culture. The UK’s culturally specific factors limiting female take-up of cycling seem to remain in place, even where cycling has gone up. Creating a mass cycling culture may require deliberately targeting infrastructure and policies towards currently under-represented groups.

Progress Update

A brief update on progress, as of the start of August 2014.

Model One
We have built our first model, Model One. We believe the model is quite innovative, not only being different from traditional transport models but also other newer approaches.

It doesn’t look at trade-offs between cost and time, but rather seeks to model:
(a) the relationship between use of ‘safety stuff’ (e.g. helmets, high-visibility clothing), perceptions of danger and the quality of the cycling environment, and
(b) ways in which people influence their others in local peer groups (colleagues, friends, neighbours).

It’s influenced by sociological perspectives known as ‘practice theories’, which explore the cultural and material resources people need to participate in different practices. For example, where practices are seen as requiring lots of skills or lots of stuff, this can exclude people from participating.

Example of cluster plots used in analysis of Model One

Example of cluster plots used in analysis of Model One

It’s been challenging bringing agent-based modelling together with practice theories, and we feel like we’ve learned more about the strengths and limitations of both. One issue has been model scope: initially, we wanted to include seven meanings that people might associate with cycling, but, we realised this would make the model too complex to be interpretable.

Focusing on ‘safety stuff’ and danger has allowed us to concentrate on what we see as a crucial aspect of cycling in countries such as the UK: cycling feels dangerous (and is, compared to high-cycling countries), it seems like something only for the risk tolerant and highly skilled, and something that required a lot of protective clothing. Although that’s perhaps the overall picture, the model lets us look at ‘clusters’ developing: in other words, there can be diversity within a local area in how much ‘stuff’ people use to cycle, and how dangerous they think it is. This is something that we’ve observed anecdotally but which hasn’t had much research attention.

The model allows us to explore in an abstract way two different types of intervention.
- The first involves workplaces giving employees ‘safety stuff’ that people may feel they need to cycle – you could imagine this as your employer giving everyone a high-visibility tabard.
- The second involves improving local cycling environments in part of the area – you could imagine this as the local authority building high quality cycle tracks in part of a city. In our model, this reduces the negative experiences people have cycling (not necessarily injuries but also near misses), and helps to reduce their perception of danger.

We can explore both the short and longer-term dynamics that may result from the two different types of interventions, given the social networks that exist and the way people learn from others (for example, seeing many people with high-visibility jackets riding to work).

In autumn Model One will be presented at the European Social Simulation Association conference and as a poster to the Public Health England Conference. Our work on it will also be discussed at the conference on Uncertainty in Computer Models 2014. We’ve already spoken about it to audiences at the University of the West of England, Manchester Metropolitan University, the London School for Hygiene and Tropical Medicine, the University of Surrey and the Royal Geographical Society Conference.

We are going to put Model One online for people to try out themselves. In the meantime we can offer demonstrations or presentations of the model and our work on it.

The audience at Rachel's Research Methods Festival session.

The audience at Rachel’s Research Methods Festival session.

We have been working on six papers that either have recently been submitted or which will be submitted later this year. These papers will later be summarised in more readable format for the website.

Paper 1: We have recently submitted a journal article entitled ‘Does more cycling mean more diversity in cycling? An analysis of English and Welsh Census data on cycle commuting, exploring shifts in gender and age composition’ (with Anna Goodman, LSHTM). This topic has great policy interest and Rachel has presented findings at the Bristol Cycle Festival and the Hackney Cycling Conference, and been invited to speak at the upcoming Cycling and Society Symposium and the London Cycling Show.

Paper 2: We’re working on a paper describing and analysing Model One, explaining what we did and what we can learn from it.

Paper 3: We are working on a paper analysing data from three qualitative research projects, looking at interview material about ‘safety stuff’ in relation to cycling. This paper explores people’s attitudes to, and experiences of, ‘safety stuff’ in different cycling contexts. Preliminary findings have already been presented by James at the UKCRC Centres of Excellence Conference and by Rachel at the 2014 ESRC Research Methods Festival (and also as part of the UWE talk).

Paper 4: Earlier in the project, we drafted a paper about using practice theory in relation to transport modelling, which is being redrafted after input from advisory board members.

Paper 5: A fifth paper will be based on the analysis of Model One.

Paper 6: Another paper explores how cycling (and walking) commuting vary over time within individuals, and what factors predict returning to cycling or walking after a period of travel by other modes (with Adam Martin, CEDAR).

Model Two

Having created Model One, focusing on perceptions of danger and the ‘stuff’ needed to cycle, we’ve decided to do something a bit different and hopefully complementary for our second model.

Model Two will include more realistic geography. Currently we are working with Robin Lovelace (University of Leeds) to generate synthetic populations at the small area level based on census data. The model will also include a closer link to quantitative data. It will explore movements into and out of cycling, using longitudinal data from the British Household Panel Survey. We will be able to explore how cycling rates have changed over time, and look at how that might be understood in relation to social influence (as cycling tends to be clustered in localities and in workplaces).

We will be developing Model Two over the summer, and are keen to discuss its progress with a variety of experts.

Modelling human behaviour and social interactions can help evaluations to explain the unexpected

Cross-posted courtesy of LSHTM’s School for Public Health Research.

Zaid Chalabi and Theo Lorenc discuss why simulations based on known human behaviour could help us to understand complex interventions like Glasgow’s urban regeneration programme.

When big sports stadiums are built, designers must ask an important question: ‘How are people going to get out if there’s a fire?’ It’s not enough just to include plenty of exits. Designers need to understand how individuals are likely to act during a large-scale fire, to ensure that exits would not become blocked and the evacuation plan would actually work.

This question can be understood using a method called ‘agent-based modelling’. It involves simulating real-life situations in a ‘bottom-up’ way, with agents who can respond to their environment and to each other, based on what is known about individual behaviours. It’s a more intricate form of modelling than, for example, chronic disease modelling in which people are simply passive victims of environmental exposures and risk factors. Crucially, agent-based modelling sees individuals as decision-makers and tries to create social simulations of their behaviours.

It’s an approach that can’t predict every eventuality, but it gets you closer to likely outcomes than simply relying on hunches or intuition. Like other forms of simulation that harness available evidence, it is cheaper than realising, too late, that you’ve made a big mistake. It helps to avoid the unintended consequences of change in complex systems involving multiple actors.

However, relatively few researchers have tried to apply this approach to large-scale social programmes, such as urban regeneration. For example, Glasgow is currently demolishing vast neighbourhoods and rehousing people in regenerated environments. The process is likely to take up to 20 years, the entire childhood of some residents and a third of the lifetime of many from these deprived communities. LSHTM’s Matt Egan is part of the 10-year GoWell longitudinal study of this programme. Matt has already detailed initial evaluations in a previous blog and SPHR@L seminar (read Matt’s blog here).

We know that improving the physical fabric of people’s lives can be good for health. But the demolitions and regeneration of the 1960s demonstrated that maintaining and even strengthening the social fabric can be more problematic. As Matt argues, if you destroy a neighbourhood, you need a strategy to support those who rely on it most.

That’s where we believe there is a place for agent-based modelling. An agent-based model simulating the population of the regenerated areas could help us to anticipate the behaviour of different types of residents over time. We could also investigate how the model matches up with and explains what has been observed so far in the Glasgow regeneration project. Testing the model against the reality would help to validate it and also refine it for use elsewhere.

How might agent-based modelling help researchers? It could provide an important tool to support evaluation. For example, the GoWell evaluation of the demolition programme expected to find a negative health impact on residents in the initial phase, resulting from remaining in a deteriorating environment and being left behind by friends, family and neighbours. It didn’t. As Matt explains in his blog, the evaluation team found virtually no change in the mental and physical health of residents living for two years in an area that was being destroyed around them.

There are lots of possible reasons for this surprising finding; perhaps things had already got so bad on the estates that the demolition programme did not make matters much worse. However, agent-based modelling, which builds in known evidence about human actions and interactions in certain circumstances, might help to disentangle and inform these findings. Such modelling could also potentially help policy strategists to plan such big projects far better in the future, by simulating how large-scale changes to the environment might impact on residents’ behaviour and wellbeing.

The need for such insights is clear for the later stages of the Glasgow regeneration project. Research is beginning to suggest that the mental health of white Scottish residents seems largely unthreatened by the disruption of relocation. This may be because these residents may already have had weak community links. In contrast, immigrant communities, whose cohesiveness probably protected their health from the adversity of living on slum estates, may now be heading for trouble as their networks risk being broken up in new housing configurations. A good agent-based model might be able to simulate such potential downsides for different populations and help policy-makers to build in alleviating measures.

Modelling cannot answer all questions. The world is complex, particularly when you are modelling a society and trying to build in all the different decisions that people make in various circumstances. Modelling is also only as good as the evidence base. But if we start to build models, they will gradually get better and more informative as researchers gather the missing data.

Dr Zaid Chalabi is Senior Lecturer in Health Impact Analysis and Mathematical Modelling at LSHTM. Dr Theo Lorenc is Honorary Lecturer at LSHTM and Provost Fellow at the Department of Science, Technology, Engineering, and Public Policy (STEaPP), University College London. On May 20, they will present a seminar at the NIHR School for Public Health Research at LSHTM, entitled ‘Can Agent-Based Models Inform the Evaluation of Complex Natural Experiments?’

Workshop in Lund – cycling meanings (2)

This is the second part of the writeup from the Changing Commutes project workshop in Lund. For more on the workshop goals and participants see my previous post. Again it’s my reflections on, and interpretations of, the discussions at the workshop.

For me one of the most interesting things about the meaning-related discussions were the comments about stereotyping and stigma. I have elsewhere written about cycling and stigma in a UK context – so it was fascinating to find out about images of cyclists in very different country contexts.

My ‘Incompetent, or Too Competent’ paper suggested that in the UK, cycling is doubly stigmatised. As DfT research also found, there is a stereotype of cyclists as being incompetent road users. This is in turn is linked to cycling’s historic association with poverty (in the postwar decades, policy prioritised the car and increasingly it was assumed that only those unable to afford a car would cycle to work) and with children (as adult transport cycling disappeared, bikes became seen as toys for children). Unsurprisingly, in this context many cyclists react to their cultural and infrastructural marginalisation (including, for example, regular experiences of fast moving motor traffic passing close enough to touch) by putting on what one Cycling Cultures interviewee called ‘the uniform of a cyclist’. In other words, what we in the UK know as ‘cycling gear’ – high vis, helmet, maybe even a face mask. But, with cycling seen for so many decades as not a ‘proper’ form of transport, this unfortunately doesn’t help – you may ‘look like a cyclist’, but, you’re still seen as a problem (e.g. a ‘Lycra Lout’) because cyclists are by definition seen as low status and low priority travellers. This may be one reason why bike hire schemes are relatively safe, in London and many other cities, compared to own-bike cycling. Hire bike users are often dressed in everyday clothing and may not even own a bike; in the words of a couple of HGV drivers interviewed by my colleagues:

W: You’ve got the Barclays bikes as well…and they haven’t been trained on the road, you can see them, you see them wobbling along the road like that…And you just hold back until, you know, they’ve disappeared or done whatever they want…

S: [...] On the…Barclays bikes, you just don’t know what you’re going to get. You can get the suited gentleman, you know, with his shoes on and his bowler hat and gets on there to do his 300 metres. Or you may get a family of four, who’ve just come over from Japan on holiday and they decide they want to go on, you know, …look at the touristy things.

I found this exchange very interesting, because while the drivers were complaining about hire bike users’ being unskilled and unpredictable, at the same time they said that they would behave in a more considerate manner towards these riders!

Some of the contexts people drew upon were similar to the UK in this respect. In Belarus, for example, we were told that many people saw cycling as incompatible with professionalism; conversely, cycling then became a kind of resistance identity.

In Belarus one company hired a book-keeper who cycled and another company member wrote in the [company] blog ‘She cycles to work! Should I fire her?’ People were asking ‘is she sweaty, does she smell?’ – it was considered unprofessional. And of course she changes into her suit when she gets in but it’s still seen as unprofessional. In Belarus when I was cycling I felt special like I was trying to deliver an identity message, here [in Lund] it’s just normal.

Comments made about Stockholm suggested some similarities to the London context: a tradition of marginalising cycling, alongside a recent revival in cycling still so far dominated by a relatively limited social demographic; the ‘MAMIL’ stereotype (‘Middle Aged Man in Lycra’) was recognised as comprehensible in a Stockholm contexts, albeit with fewer negative connotations than in the UK. In relation to Skåne/Copenhagen, by contrast, people expressed disbelief at the pejorative nature of the ‘MAMIL’ stereotype. One participant asked why these people would be negatively stereotyped when they are doing the culturally ‘correct’ thing of getting exercise and protecting their health.

In Skåne and Copenhagen though, there did seem to be boundaries in terms of what was considered normal and what was exceptional cycling. We heard that 15km (10 miles) was not considered a ‘normal’ distance to commute by bike in Copenhagen, with ‘normal cyclists’ (‘people like us’) driving such distances. In Lund, those distances might be much shorter, with people perhaps seeing 5km as a cut-off point at which driving would become normal. Some of these kinds of distinctions were reminiscent of Cambridge Cycling Cultures interviews, with many of our Cambridge interviewees owning cars and using them to journeys outside the city centre. It did seem that in Skåne and Copenhagen there were some weak stereotypes relating to people who commuted ‘abnormal’ distances; but without the pejorative and negative connotations associated with this in the UK.

Here there is a clear contrast with Cambridge. Even in Cambridge, with over 30% of all journeys to work cycled, the Cycling Cultures research found cycling is still stigmatised; but in Skåne and Copenhagen, this wasn’t the case. People did describe negative views about cyclists; in Copenhagen, we heard about the ‘shouty cyclist’ who believes in their own priority; we also heard about Swedish perceptions of cyclists as rule-breaking. However, these perceptions weren’t seen to undermine the credibility of cyclists as a group or the credibility of cycling provision and funding. We heard that to some extent, there is instead a discourse linking cyclist misbehaviour to problematic environments – for example, having to stop and push a button at cycle crossings, which, it was argued, leads to cyclists becoming impatient. The irony is that of course in Skåne and Copenhagen the cycling infrastructure is generally superior to that found in the UK, yet in the UK, precisely because of the greater marginalisation of cycling, cyclist rule-breaking is less often seen as an indicator of the need to improve cycling environments. Instead, in the UK, it’s often seen as a reason for not improving cycling environments. (As advocates have put it, motorists are not told that they won’t get any investment in car infrastructure until there’s no more speeding – even large amounts of individua)

In contexts like Copenhagen and Skåne, while there are negative views about cyclists in some quarters (e.g. conflict with pedestrians), this doesn’t seem to lead to the kind of local media hostility and broader local anti-cycling public discourse that we found even in a high-cycling UK city like Cambridge. Even where there are signs of stereotyping of cyclists in Skåne and Copenhagen, it doesn’t lead to stigma and marginalisation.

For another project, I’m currently looking into how media representations of cycling in London has changed. It’ll be interesting to see to what extent the stories have changed, and to what extent advocates have succeeded in challenging the reinforcing loops of stigma and stereotype. I think there’s likely to be some indications of change, but, we’re still some way off the situation in Copenhagen, where ‘shouty cyclists’ might well be criticised, but are not seen as undermining the importance of cycling as a transport mode.

Workshop resources from Lund (1)

This is the first part of a write-up of our project workshop in Lund.

James and I held an expert workshop in Lund, Sweden, which fitted in with my Erasmus visit there. The experts were students and academics based in Sweden and Denmark; we anticipated they’d be able to talk about those contexts, but people could also draw on experiences of experiences of cycling in contexts as diverse as Ukraine, Canada, and China.

The main purpose of the workshop was to find out how participants thought processes around cycling ‘meanings, skills and stuff’ worked in their contexts. It also aimed to get feedback on the idea of using practice theory for agent-based modelling and to see how well using workshops to develop ABM rules might work. For more on practice theory and cycling, and what is meant by ‘meanings, skills, and stuff’ see an article I’ve written with Kat Jungnickel discussing practice theory concepts in relation to cycling.

This is my write-up of the workshop, so while it relies on James’ and participants’ comments, notes, etc. they might not agree on all points – it is also me thinking through ideas as I go along.

We created three sheets for small group discussions focusing on cycling ‘meanings’, ‘skills’, and ‘stuff’ . In the end we only actually used two of these, with one small group discussing ‘meanings’ and one ‘skills’. The sheets sought to give people from different country contexts a sense of how (we think) things work in the UK for cycling, using data from the Cycling Cultures project. We would use slightly different resources for a UK audience which might be more familiar with some of these issues and experiences. The sheets also contained questions for discussion. James sat in on one group and I sat in on another.

The discussion about skills/competences was interesting in that neither word quite seemed to connect with participants. The problem with ‘skill’ is firstly that in everyday use it tends to foreground something that individuals can acquire (particularly in a context where people are all meant to continuously upskill, be employable, and so on!), and secondly that ‘lower skilled’ is seen as derogatory. Therefore, people were not always comfortable talking about skill in relation to cycling practices. Competence was if anything more difficult; ‘less competent’ carries echoes of ‘incompetent’; given cyclists are often cast as incompetent road users, this was also problematic. Perhaps ‘abilities’ might be a better word to use in this context as implying something that might be more or less flexible or fixed for different people. Although it may be possible for individuals to increase their ability to cycle, being unable or less able to do something doesn’t carry the same pejorative weight as being ‘unskilled’ or ‘less skilled’, ‘less competent’ or ‘incompetent’.

That said, there were very interesting discussions in the ‘skills’ group, for example, comparing the two relatively high cycling contexts of Malmö and Copenhagen. Copenhagen’s cycling environment was seen as more consistent than Malmö’s, with one participant talking about being ‘surprised’ to suddenly find herself in the middle of motor traffic while riding in Malmö. People thought that more skills were needed to ride in Malmö because of the inconsistent environment – sometimes a cyclist needed to act ‘like a cyclist’, but sometimes like a pedestrian or like a driver, depending on the cycling environment. However, it was also noted that this led to a greater acceptance of rule-breaking, precisely because of the lack of clarity! (While in Copenhagen, it was argued that the greater consistency of environment has led to a stronger consensus about how cyclists should behave, and hence greater disapproval by cyclists of rule-breaking).

One aspect of practice theory that’s always attracted me is the idea of relocation of competences or skills (or abilities, in fact) from individuals to objects, or vice versa. So, in-car GPS means that passengers and drivers no longer have to acquire/possess map-reading abilities. In a sense it can be seen as deskilling. However, one nagging thought has always been, but, as one type of skill becomes obsolete isn’t it replaced by others? Aren’t there abilities related to using GPS? In what sense is the need to use GPS less problematic than the need to read a map? (You could imagine if there were only speaking GPS and no written maps, how difficult this would make navigation for Deaf people, for example).

This issue came up in relation to cycling skills, in the discussion around levels of skill needed to ride in Copenhagen. While in some respects cycling in Copenhagen was seen as relatively easy, the skills needed to signal and communicate with other cyclists were raised as potentially problematic for those unused to this. To what extent can we identify a skill gradient outside of the prioritisation of certain abilities over others?

So, for example, in the UK the tradition of vehicular cycling has prioritised the ability to cope with riding in motor traffic; not, for example, the ability to negotiate and coexist with pedestrians or with other cyclists, which may be seen as more important in cycling environments where cyclists are less often expected to act ‘as a vehicle’. Similarly with ‘confidence’, a popular term in UK cycling discourse: when cycling in The Netherlands, I’ve been impressed to see people cycling sociably and slowly two abreast on narrow streets, blithely blocking overtaking by a car behind. In a sense, this represents confidence by the cyclists that they have a right to occupy the street; and a lack of confidence by the car driver in asserting a ‘right’ to overtake. Confidence isn’t purely an individual attribute; it expresses the legitimacy of different modes which in turn is linked to policy and provision.

I’m getting slightly off topic now, so bringing it back, during this project we will continue working on how we think about and model skill/competence/ability. For creating our model, we will need to narrow down and focus on what we think is essential to the workings of the system, and to the potential for change. We will no doubt have to simplify many of the complexities around the social construction of skill, and the relationships between inclusive design and different types of skills and competences.

The next post in this series will discuss some ideas that came out of the discussion on meanings.