The aim of the project is to assess how life transitions influence travel behaviour and to identify opportunities from this for policy interventions to achieve desirable transport outcomes. Life transitions (major changes in personal circumstances) are associated with life events such as joining the labour force, moving home, having children or retiring. Emerging research has established that significant changes in travel behaviour are often associated with life events. However, there remains much to learn about the extent to which different life events trigger behavioural change and the conditions under which life events are more likely to trigger change. It is of interest to policy makers to better understand this so that policies can be formulated to influence travel behaviour and achieve objectives such as tackling congestion and carbon emissions from transport. The Understanding Society (US) survey offers a previously unavailable opportunity to investigate interactions between life events and travel behaviour for a large, representative sample of the UK population (40,000 households and 100,000 individuals). We know that about 10% of the sample move home each year with about 700 births per year and a similar number of retirements. Data from the first three waves of US will be available during the project. Information collected on travel behaviour includes household car ownership (number of cars owned) and commuting behaviour (mode, distance and time of travel) as well as information on housing mobility, changes in employment and other key life events. British Household Panel Survey (BHPS) respondents have been incorporated into US since wave 2 and it will also be possible to create a longer data history for these respondents. In addition, the project will be innovative in linking the survey data to local spatial data drawn from Office for National Statistics (ONS) sources such as the census and Department for Transport (DfT) accessibility indicators. This data will allow new insights into the effect of the spatial context on changes to travel behaviour associated with life events. The project will start by creating the required bespoke data sets drawn from US/BHPS and ONS data. The first stage of data analysis will identify the prevalence of life events amongst different population groups and the relative importance of different life events for changes being made to car ownership level and commuting behaviour. It will then be analysed under what circumstances life events are more likely to result in changes in these types of travel behaviour. Consideration will be given to individual, household and geographical circumstances and statistical models will be used to quantify the relative importance of different factors. The final part of the analysis will assess the stability of car ownership and commuting behaviour and whether people whose behaviour has been stable are less likely to change their behaviour when life events occur. The project team comprises travel behaviour experts from Centre for Transport & Society, University of the West of England (UWE), longitudinal data experts from Institute for Social and Economic Research (ISER), University of Essex, and policy researchers from DfT. The project will extend the skills base in the analysis of large and complex data sets to the field of transport. DfT will be involved in all stages of the research to ensure there is a close link with policy and practice needs. The research co-investigator from DfT will chair an Advisory Group that also includes representatives from transport practice and academia. UWE will manage a website for the project, including a researcher forum, and will draw on their existing networks in academic and policy circles to promote engagement with the research. ISER will contribute media communications support and disseminate findings from the research through the ISER website and US communications team at ESRC.

Promoting cycling, including promoting cycling among children, would be expected to deliver substantial benefits in terms of population health and environmental sustainability. Many children do not meet government recommendations in terms of the amount of exercise they do, and increasing levels of cycling would be one way in which they could incorporate additional physical activity into their everyday lives. In addition, many children are currently driven relatively short distances by their parents to be dropped off at school or other destinations. If more children instead made these trips by bicycle then this would also be expected to reduce the congestion, air pollution and greenhouse gas emissions associated with motorised transport. Since 2007, one of the Department for Transport's flagship policies to promote cycling is the delivery of 'Bikeability' cycle proficiency training in schools. Currently, around half of children in England are offered the training for free before they leave primary school, and the annual cost of the programme to the Department for Transport is £11 million. However there exists very little robust evidence regarding which particular children get the cycle training, or regarding the effect the scheme has on subsequent cycling behaviour. In collaboration with the Department for Transport, this proposal will seek to fill these gaps in the evidence using data from the Millennium Cohort Study (MCS). MCS is a nationally-representative birth cohort which has now been surveyed five times, most recently in 2012 at age 10/11. In this most recent sweep, 8,700 parents in England were asked if their children had "ever done any formal cycling proficiency training such as 'Bikeability'". Parents were also asked how often their children used their bicycles. Using these data, we will seek to answer two broad research questions. First we will examine which individual, family and area characteristics help explain why some children do cycle training and some do not; for example, are boys more likely than girls to get the training, or are children from richer areas more likely to get the training than those from poorer areas? Second, we will examine whether children do in fact cycle more often if they have been offered cycle training in school. Answering this second question will involve comparing children whose schools had already offered cycle training at the time of the MCS survey with children whose schools offered cycle training later in the same year. If cycle training is effective, our prediction is that the first set of children would report cycling more often than the second set of children. As a part of this second research question, we will examine whether there is any evidence that cycle training works better for some sorts of children than for others - for example, whether it has a bigger effect on boys than on girls. Together, answering these two questions will provide the most robust evidence to date regarding the effectiveness of cycle training in schools, and regarding whether all children benefit equally. Our non-academic partners, the Department for Transport, will then be able to use this evidence as part of deciding how best to pursue their goal of increasing cycling in childhood in an effective, cost-effective and equitable manner. Our findings will also have broader relevance for the international evidence base, helping to address the current lack of robust studies examining the effectiveness and equity of different types of policies to promote cycling.

Efforts to reduce the emissions from car travel have been hampered by a lack of specific information on car ownership and use. In 2010, the Department for Transport released a dataset containing annual MOT test records for cars from 2005 onwards, with regular updates promised. By providing relatively comprehensive information about British car ownership and use, this dataset provides a key opportunity to address a number of issues in transport and energy debates. For the first time precise links can be made between car use and car type, and changes in use over space and time can be examined on a relatively complete basis. When combined with a wealth of other existing data sets (not least the new information from the 2011 Census), a range of new and important insights should emerge. Having already worked together as a project team to scope the use of this data in a small EPSRC-funded study in 2011, we now propose to use it as a platform upon which to develop a set of interlinked modelling and analysis tasks using multiple sources of vehicle-specific and area-based data. A set of interdependent workpackages will span three years to investigate spatial and temporal differences in car ownership and use, the determinants of those differences, and how levels may change over time and in response to various policy measures. The relationships between car ownership, car use, fuel use and vehicle emissions, and the demographic, economic, infrastructural and socio-cultural factors influencing these will be tested mathematically using spatial statistics, regression modelling and scenario analyses. Linkages will also be made with spatial patterns of domestic gas and electricity usage in order to understand relationships within and between these end-user energy demands. The new analysis capability will be tested through case study evaluation of local transport policies. By enabling car ownership and use to be examined at relatively fine spatial and temporal scales, and via techniques to identify areas sharing important 'background' characteristics, it should be possible to answer key questions for sustainable transport policies such as, what difference to car ownership and use have particular policies achieved (compared with areas where these policies were not in place)? It will also be able to calculate figures for fuel use and emissions to contribute to the development of policies specifically targeted at the most energy intensive or polluting drivers or localities. We will also be able to link energy use from cars, with domestic energy usage through household electricity and gas. This will allow us to build up a much better picture of energy and carbon footprints across the country. When linked to patterns of income, multiple deprivation and other socio-economic factors, there will be insights for the design of much more effective climate and energy policies, and to ensure that the burden of these is borne equitably. The project will be supported by an Applied Statistics Expert Panel, and includes provision for workshops with key stakeholders to help shape the work. The project will also help develop a specification for a possible web-based tool to enable a wide community of users to undertake their own analysis on these sorts of issues, using the data and tools that we develop. In order to achieve our goals, we will develop methods to overcome the challenges of merging a range of important but disparate datasets, based on varying spatial, temporal and other characteristics, and subject to varying issues of data protection and sensitivity. Our scoping study demonstrated that there are very significant technical challenges to be overcome in working with datasets of this size and nature, and a wide range of disciplines may be able to learn from this work. The analysis frameworks and the new scientific understanding delivered will be the important legacies of this project.

Efforts to reduce the emissions from car travel have been hampered by a lack of specific information on car ownership and use. In 2010, the Department for Transport released a dataset containing annual MOT test records for cars from 2005 onwards, with regular updates promised. By providing relatively comprehensive information about British car ownership and use, this dataset provides a key opportunity to address a number of issues in transport and energy debates. For the first time precise links can be made between car use and car type, and changes in use over space and time can be examined on a relatively complete basis. When combined with a wealth of other existing data sets (not least the new information from the 2011 Census), a range of new and important insights should emerge. Having already worked together as a project team to scope the use of this data in a small EPSRC-funded study in 2011, we now propose to use it as a platform upon which to develop a set of interlinked modelling and analysis tasks using multiple sources of vehicle-specific and area-based data. A set of interdependent workpackages will span three years to investigate spatial and temporal differences in car ownership and use, the determinants of those differences, and how levels may change over time and in response to various policy measures. The relationships between car ownership, car use, fuel use and vehicle emissions, and the demographic, economic, infrastructural and socio-cultural factors influencing these will be tested mathematically using spatial statistics, regression modelling and scenario analyses. Linkages will also be made with spatial patterns of domestic gas and electricity usage in order to understand relationships within and between these end-user energy demands. The new analysis capability will be tested through case study evaluation of local transport policies. By enabling car ownership and use to be examined at relatively fine spatial and temporal scales, and via techniques to identify areas sharing important 'background' characteristics, it should be possible to answer key questions for sustainable transport policies such as, what difference to car ownership and use have particular policies achieved (compared with areas where these policies were not in place)? It will also be able to calculate figures for fuel use and emissions to contribute to the development of policies specifically targeted at the most energy intensive or polluting drivers or localities. We will also be able to link energy use from cars, with domestic energy usage through household electricity and gas. This will allow us to build up a much better picture of energy and carbon footprints across the country. When linked to patterns of income, multiple deprivation and other socio-economic factors, there will be insights for the design of much more effective climate and energy policies, and to ensure that the burden of these is borne equitably. The project will be supported by an Applied Statistics Expert Panel, and includes provision for workshops with key stakeholders to help shape the work. The project will also help develop a specification for a possible web-based tool to enable a wide community of users to undertake their own analysis on these sorts of issues, using the data and tools that we develop. In order to achieve our goals, we will develop methods to overcome the challenges of merging a range of important but disparate datasets, based on varying spatial, temporal and other characteristics, and subject to varying issues of data protection and sensitivity. Our scoping study demonstrated that there are very significant technical challenges to be overcome in working with datasets of this size and nature, and a wide range of disciplines may be able to learn from this work. The analysis frameworks and the new scientific understanding delivered will be the important legacies of this project.