1.1 The introduction of temporary measures

As countries navigated the initial weeks of large-scale lockdowns in the Spring of 2020, it became apparent that a programme of road-space reallocation would need to be initiated. This would create more space for social distancing for pedestrians but also help encourage and support those no longer travelling by public transport to travel by active modes. The necessity of such a programme was a result of the entrenched imbalances in space provided for walking, cycling and other active modes (collectively referred to as “active travel” throughout this paper) compared to that provided for motor vehicles (Nurse and Dunning, 2020). As a result, policymakers were forced to act to address this, at least in the short-term.

In the UK, efforts to introduce these temporary measures were supported and enabled through statutory guidance published by the UK government along with several temporary legislative changes giving increased powers to local authorities (Department for Transport, 2020c). Financial support was also provided through the Emergency Active Travel Fund (EATF), which was a repurposing of existing promised spending on cycling and buses (Department for Transport, 2020a). This provided – through two “tranches” or phases of funding – a total of £217.5 million to local authorities in England (Northern Ireland, Scotland, and Wales were funded separately by their devolved governments) to deliver these temporary interventions. The EATF was renamed the Active Travel Fund (ATF) for the second, more substantial, tranche of funding in England.

Local authorities responded rapidly at the onset of the pandemic to reallocate road space. Indeed, as Dunning and Nurse (2020) observed, the ability to quickly repurpose existing infrastructure was notable and demonstrable of the scale of change that can be achieved in the short-term, if required. Although not all local authorities moved quickly to implement these measures, with a considerable lag in some cases. Often, it was those with existing infrastructure that moved first, perhaps having a local policy disposition to promoting active travel. Considerable debate about the ambition, or over-ambition, of the measures followed between the “pro” and “anti” camps, with some measures being removed after a negative reception.

It is beyond the scope of this project to examine in significant detail broader UK responses to the pandemic with regards to active travel. However, a detailed account has been produced separately by the authors (Gore et al., 2021).

Further afield many major cities introduced measures, not just those traditionally associated with cycling such as Paris, which created almost 50 kilometres of temporary cycle lanes, but also places like Barcelona (over 20 kilometres) and Milan (around 35 kilometres).¹ Similar responses were also seen across the globe in places such as New York (NYC Department of Transport, 2021) and Bogotá (WHO, 2020). In many places these measures are seen as a first step in reducing private car use and improving air quality. Although not all of the temporary measures saw the same levels of success. A study in Lausanne and Geneva, Switzerland, found that around a third of cyclists were likely to cycle more as a result of the cycle lanes (Rérat et al., 2022). An Australian study found that the cycle lanes were “somewhat” or “extremely” important in increasing the level of cycling (Fuller et al., 2021). However, this was one of the least important factors, compared with fitness or social interaction, for example. This is in contrast to the experience of Thessaloniki, Greece, where the failure of the cycle lane led to its removal after 18 months (Katsavounidou et al., 2023). Using secondary data, Buehler and Pucher (2021) found widespread increases in cycling across the US, Europe and Australia, attributing much of this to leisure cycling (i.e., mostly occurring outside the working week), they conclude that this increase was supported by the “expansion and improvement of protected cycling infrastructure” (ibid: p.398).

In the UK, efforts to support more active travel because of the pandemic, coincided with a longer-term ambition to establish a step-change in the role of active travel in the wider transport system. In May 2020 the Government published a walking and cycling plan (Gear Change) which included a commitment for half of all trips in towns and cities to be made by active modes by 2030 (Department for Transport, 2020b).

1.2 Travel behaviour change and the pandemic

The sudden restriction to movement caused by the pandemic created a substantial shift in travel behaviour, and the impact of the Covid-19 pandemic was felt across all forms of mobility. The need to ensure social distancing and the imperative to stay at home to help prevent the spread of the disease meant that there was a significant decline in mobilities generally (Hadjidemetriou et al., 2020), with large town and city centres seeing much reduced footfall (Enoch et al., 2022). Public transport was particularly hard hit by Covid-19, with shared mobilities not conducive to efforts to social distance (Kłos-Adamkiewicz and Gutowski, 2022). Indeed, governments largely encouraged a shift away from public transport at the height of the pandemic (e.g., Mason, 2020). Whilst levels of cycling were shown to have increased as a result of the pandemic (Buehler and Pucher, 2021), such levels have been difficult to maintain. In the medium-term, the anticipated return to normal has not played out quite as expected, with levels of car use not back to pre-pandemic levels (Anable et al., 2022; Department for Transport, 2023). Increased working from home is likely to be a factor with this, although there does remain some uncertainty as to what the longer-term picture will look like.

Whilst the Covid-19 pandemic presented an unprecedented degree of disruption, the approach used to study its impact builds from previous experience. Marsden et al. (2020) conceptualise transport disruptions as a breakdown of the activities that the transport system itself enables. Rather than examining the physical breakdown of a certain piece of infrastructure (i.e., a bridge closure), they argue it is important to examine the broader system of complex interdependencies that make it possible for people to undertake the activities of their daily life (Ibid). This might be travelling to work or school; completing shopping or leisure trips; or undertaking caring responsibilities for others. The disruption of the pandemic altered how travel was undertaken, with variable restrictions over the course of the two years creating a system that was in flux and therefore highly uncertain.

Large-scale planned and unplanned events have been increasingly studied, owing to the opportunities for change they present. These opportunities offer a chance to shift the population away from less desirable travel behaviours (specifically private passenger vehicle use) towards modes that are regarded as more sustainable, such as walking, cycling and public transport. For example, in examining a specific and large-scale disruption to travel (the London 2012 Games), Parkes et al. (2016) found that 54% of commuters in London made at least one change to their usual travel in response to the disruption. Notably, this work also highlighted how people exist in different stages of preparedness to change and this can influence their ultimate response to a disruption to their usual travel.

The impact of the pandemic on travel behaviour has played out across a sustained temporal dimension. For instance, Gutiérrez et al. (2020) examined public transport use in the shorter-term – when the messaging was to avoid public transport – and then in the medium and longer-term. This longer-term perspective considers the shift in focus from immediate concerns to thinking about the future impacts on public transport provision and funding, and how it might be integrated into a healthier and more sustainable future transport system.

Furthermore, responses to the pandemic have also been multifaceted, with a range of policies enacted. Indeed, transport policy responses to the pandemic have been varied and combinations of measures have been applied to help deliver the desired changes in behaviour. Shortall et al. (2021) categorised measures across three objectives: (1) avoidance of travel (e.g., work-from-home, stay-at-home orders); (2) modal shift (from public/shared); and (3) improvement of quality (more capacity, encourage masks in shared space).

In this paper, we examine specifically road-space reallocation measures, which falls into the “modal shift” category. Understandably, the introduction of such measures has garnered interest in the literature. Such road-space reallocation is of great interest to transport planners as it offers a quick and relatively low-cost approach to implementing measures to increase active travel, particularly when contrasted with the substantial costs of large-scale infrastructure projects (i.e., light-rail upgrades) (Dunning and Nurse, 2020). Indeed, Covid-19 has prompted a discussion of how road-space is allocated across different modes of travel and whether the pandemic offers a turning point in this (Nurse and Dunning, 2020; Wright and Reardon, 2021), and more broadly, in terms of how we design our cities (Honey-Rosésa et al., 2020).

Attention in the literature has ranged from questions of how Covid-19 can prompt us to think differently about the way we travel around our cities, to what we allocate space for. Some research has focused on cycling specific issues (cf. Nikitas et al., 2021), whilst others have explored issues around equity of the provision of temporary interventions (Aldred et al., 2021; Firth et al., 2021; Fischer and Winters, 2021). This reflects the challenge of balancing the need to respond quickly in deploying interventions with ensuring existing inequalities in transport access are not exacerbated.

1.3 Case studies: Lancashire and Sheffield

To explore the impact of the temporary road-space reallocation measures, this study examined two contrasting local authorities in the North of England. Few studies have approached the introduction of temporary measures using a place-based approach such as this (cf. Anable et al., 2022; Firth et al., 2021; Fischer and Winters, 2021). A place-based approach explicitly considers the context when trying to understand the impact of interventions, for example, the existing physical infrastructure, geography, or local administrative structures. This compares with a one-size-fits-all approach, where local context is largely ignored.

Lancashire, in the North-West of England, is a large non-metropolitan county with a population of 1.2 million distributed across a number of towns and cities of varying size surrounded by an extensive rural hinterland. In contrast, Sheffield is a deindustrialising city in South Yorkshire with a population of almost 600,000. As a city, Sheffield is considerably more compact than Lancashire, with nearly four times as many inhabitants per square kilometre. Data on the access to large (5000+ people) employment centres also demonstrates the diverging geographical structures of the two case-study locations (Department for Transport, 2021a). For inhabitants in Lancashire, average travel times to such employment centres before the pandemic were 31 minutes by bicycle and 17 minutes by car, whilst in Sheffield the equivalent figures were only 20 minutes by bicycle and 14 minutes by car (Ibid).

The motivation for examining these two case-study areas stems from a desire to explore the extent to which place-based perspectives can shed light on how best to approach efforts to help increase levels of active travel. Whilst these places have contrasting spatial structures, the type of measures introduced were broadly similar and included footpath widening; road pedestrianisation; road closures or restrictions to through traffic; rerouting of bus routes; temporary cycle/active travel lanes; and a low-traffic neighbourhood (in Sheffield).

Initially, Lancashire introduced 14 road-space reallocation measures, whilst Sheffield introduced seven. The Lancashire measures were weighted towards increasing cycling, with roads closed to vehicles, or restricted through traffic and temporary cycle lanes. Sheffield’s measures were more balanced between walking and cycling, with greater use of footpath widening and pedestrianisation. the low-traffic-neighbourhood and closure to through traffic benefits both pedestrians and cyclists.

This difference in the measures introduced reflects to some extent the different geographical contexts, although half of the Lancashire interventions were in, or linked to, Preston, the county town. This of course means that many of the Lancashire measures, certainly outside Preston, operated in isolation, rather than as a network, or as improvements to a network of existing infrastructure, in contrast to Sheffield. Sheffield also has a considerably higher population density than Preston, and a tram system that has been operating for almost three decades. These will undoubtedly have an impact on previous and evolving travel patterns.

This paper provides insights from this study of the road-space reallocation measures in these two local authorities, with data collected through a three-wave panel study. The remainder of the paper will first explain the methods in more detail. It will then present the findings and accompanying discussion, before presenting conclusions from this work.

3.1 Characteristics of the sample

The sample achieved through the Wave One survey was equally split across the two local authorities; 50.6% of valid responses were from Lancashire and 49.4% from Sheffield. In terms of socio-demographic characteristics there were some differences between the two cohorts. More people identified as female in Lancashire (70.6%) than in Sheffield (56.1%). The Sheffield cohort was older (a quarter of respondents were over 65) and this also meant that a much greater proportion of the Sheffield cohort were retired (27.0%) compared to those in Lancashire (5.3%). The majority of participants in both cohorts were white with comparable proportions in each location (96.9% in Lancashire and 96.5% in Sheffield). These figures are outlined in Table 2.

The dependence on convenience sampling and using online only surveying – owing to resource limitations and the pandemic – created an enhanced risk that certain population groups would be underrepresented. Significant effort was made by the project team to promote the survey widely given concerns around ensuring representativeness. Community groups and charities operating in wards of each local authority that are typically underrepresented in research (e.g., those with higher levels of deprivation or higher levels of ethnic minority residents) were targeted to promote the survey, along with efforts by the local authorities’ partners to promote the survey widely using their own mailing lists. In addition, the survey was presented as a broader travel survey rather than a specific “active travel” survey to avoid self-selection of those opposed to measures to restrict or reduce motor traffic (such as Low-Traffic Neighbourhoods) and risking biasing the sample as such.

Ultimately, there are some limitations to this sample that we acknowledge and therefore cannot claim it to be representative of the wider population. In Table 2, we have provided comparators for both case-study locations and at the national level. This helps to provide context around where the samples either align or diverge from the population. Despite these limitations, this research provides valuable and timely insights into the impacts of the pandemic and associated road-space reallocation measures, which will be explored in the remainder of this section.

3.2 The behavioural shift caused by the pandemic

The scale of the reduction in mobility caused by the pandemic was substantial. Nearly half of the overall sample (47%) reported working from home during the first lockdown. A further third (31%) reported travelling to their usual workplace or elsewhere.

Across the overall sample, there were substantial levels of change in behaviour in response to the first lockdown. Of those still making journeys to a place of work, 23.2% of the Lancashire respondents who commuted had made a change to how they travelled to work prior to the pandemic. In Sheffield, this shift was even greater with 52.8% of commuters changing modes.

For shopping trips, there was also a shift in usual behaviours but more equitable across the two case-study areas (31.6% in Lancashire and 42.1% in Sheffield). Leisure trips saw the highest degree of change in both areas. Nearly half (49.1%) of leisure travellers in Lancashire had changed their behaviour, compared to almost two-thirds (60.5%) in Sheffield. Figures 2 and 3 visualise these differences for each location.

Figure 2

Figure 2

Changes made to usual mode of travel in Lancashire during the first lockdown.

Figure 3

Figure 3

Changes made to usual mode of travel in Sheffield during the first lockdown.

3.3 Use and attitudes towards measures

The types of temporary road reallocation measures introduced across both Lancashire and Sheffield were quite similar. At the time the Wave One survey was deployed, Lancashire had 14 individual temporary measures in place and Sheffield had seven. Given Lancashire’s more fragmented spatial structure as a non-metropolitan county council, these measures were more dispersed across different towns. Most of Sheffield’s measures in contrast were situated in or close to the city centre.

Awareness of these measures was high amongst both the Lancashire and Sheffield cohorts (see Table 3). Overall, 77.7% of respondents were aware of at least one temporary measure (86.9% in Sheffield, 68.6% in Lancashire). This difference is likely to be due to a combination of different factors. Firstly, that the national programme of road-space reallocation was high-profile at the beginning of the pandemic (both in national and local news). Secondly, Sheffield covers a smaller geographical area, so respondents were more likely to live closer to an intervention and although there were fewer measures they were concentrated in and around the city centre, rather than being dispersed around several locations (e.g., Lancaster, Ormskirk, Preston) as in Lancashire.

Actual use of the measures was in contrast quite low. Overall, 33.0% of respondents who were aware of the measures went on to use at least one of them as part of a journey. For respondents who were aware of the measures but did not use them, the most common response for not using them was that the measures were not on or near their usual route (36.6%), so they may have not regarded them as useful or relevant. A further 19.1% were not walking or cycling regularly and therefore did not use them.

Two-thirds of the overall sample did not use the measures, with over half of this group not doing so because of a lack of proximity to the measures themselves. This is perhaps contrary to expectations that people will travel out of their way to utilise new and improved infrastructure (Pritchard et al., 2019; Rissel et al., 2015). There are likely to be multiple facets of a journey that are traded off when comparing alternative routes (see for example, Dill and Gliebe, 2008; Prato et al., 2018). This may be explained by the relatively low number of measures deployed, particularly across Lancashire. Furthermore, given the presence of the pandemic and the fact that discretionary travel was discouraged or people were primarily working from home, it could be that they had little opportunity to actually make journeys to allow them to make use of the measures.

The follow-up surveys examined whether respondents had started using or continued to use any of the measures that were introduced by the council (of those that were still in use). Of those who were aware of the temporary measures at Wave One but did not use them, a fifth (20.7%) subsequently went on to use them at a later point. Of those who had used them at the time the first survey wave was issued, 81.6% reported continuing to utilise them at the time the second survey wave was circulated.

The use of specific measures was also variable within each case-study area. In Lancashire, there was overall a much lower use of the measures by the respondents compared to Sheffield (see Table 4). For most measures, only 1% or 2% of the sample reported using them. The exceptions to this were the Liverpool Road temporary cycle lane in South Ribble (3%), which provided a safe route to Preston city centre from a residential area, and the Fishergate closure in Preston (6%), one of the main shopping streets within the city centre. The spatial structure of Lancashire, being a county council spread over 2,894 sq. km (Lancashire County Council, 2023) means that measures were more dispersed, and it was likely that individuals may only experience one such measure.

The use of measures by the Sheffield cohort was greater in comparison to Lancashire and this may be a reflection of the increased likelihood of encountering multiple measures, particularly if navigating the city centre where the majority were located. Nearly a quarter of the Sheffield sample (23%) had made use of the Division Street pedestrianisation (which involved a closure to motor traffic of a section of a popular shopping street. The same proportion using the Pinstone Street closure, which widened the footpath along a popular route connecting two shopping areas. The A61 Shalesmoor temporary cycle lane ran along a section of the ring-road and involved the removal of a traffic lane up to a busy junction. This was removed only a few months after being installed but was well utilised during this time, with 18% of the Sheffield sample reporting having used it.

A challenge with the rapid deployment of such temporary measures is how they integrate with existing infrastructure. For example, Dunning et al. (2021) undertook video analysis of cycling journeys along temporary infrastructure in the Liverpool City Region and identified issues including the lack of connectedness with the existing infrastructure and the quality of the existing road surface. Whilst it is beyond the scope of this paper to explore this for our case-studies, it is likely – given the disjointed nature of the measures introduced – that similar challenges were experienced in both Lancashire and Sheffield.

To establish the quality and value of the temporary measures, the respondents who had used them were asked to rate their quality. A scale of 1 to 5 was provided (1 = Excellent and 5 = Poor) with respondents providing a score for each one they had used. Median scores were calculated based on this to give the average rating. All measures were rated either Good or Excellent. A trend was observed in the data that those measures rated Excellent were also the most used measures, which suggests a possible relationship between the perceived quality of temporary measure and frequency of use. It is worth noting that there is potential for self-selection bias – discussed in the methodology section – to be a factor here. Specifically, whether those using the measures are predisposed to have favourable views of them regardless of their actual experiences, because they are supportive of such interventions. However, if this analysis was opened beyond those who had experienced the measures themselves then there is equally a risk that those with unfavourable views of such interventions biasing the results, without actually having experienced them.

3.4 Impacts of temporary measures on travel behaviour

3.4.1 Cycling

The evidence above demonstrates that respondents had broadly positive attitudes towards the temporary measures and were supportive of more deliberate urban design to prioritise pedestrians. However, it is important to understand whether the introduction of measures actually led to an increase in cycling (and walking).

Analysis showed that the uptake of cycling during the first lockdown (for those who had not previously cycled), was higher in those who had used at least one of the temporary measures (8.0%), than those who had not used any (1.4%). To test whether this difference might be due to the existence of the temporary measures a chi-square test was performed. This indicated that there was a significant association between use of the measures and starting cycling for commuting (x² (1) = 19.521, p < .001) with respondents 6.29 times more likely to have taken up cycling if they used the temporary measures. Tests for leisure trips (x² (1) = 40.863, p < .001; 7.50 times more likely) and shopping trips (x² (1) = 49.412, p < .001; 10.93 times more likely), showed comparable results.

This relationship is further highlighted in Figure 4, which shows the increase in respondent’s who cycled during the first lockdown period compared to their pre-pandemic travel. This clearly indicates a trend that those who were aware of and had used the measures saw a larger increase in cycling for all three purposes shown than the other three groups.

Figure 4

Figure 4

Percentage increase in cycling between pre-pandemic levels and during the first lockdown.

This suggests that the temporary measures did have some impact during lockdown for the respondents. The extent to which these trends persist is also of interest. Figure 5 shows the increase in cycling between pre-pandemic and post the first lockdown and demonstrates how those respondents both aware and using the temporary measures saw greater increases in cycling levels.

Figure 5

Figure 5

Percentage increase in cycling between pre-pandemic levels and post-lockdown.

It is worth noting that these trends are also likely to be part of a wider story. Interestingly there was also an increase in commuter cycling for those who were aware of but had not used them and increases in both commuter and leisure cycling for those who were not aware of the road-space reallocation (see Figure 5). This would suggest that even without the measures there would have been some increase in cycling. There are several factors that may have contributed to the increase in the “not aware” group. Firstly, there were significantly lower traffic levels during the first lockdown, as a result many people would have considered it safer to cycle, particularly those with some cycling experience. Secondly, even if they were unaware of any local measures (in some Lancashire locations there were no “local” measures), there was significant coverage by the national news media to raise awareness of the increase in popularity of cycling nationally during this period, which may have encouraged others to consider cycling or start cycling again.

3.4.2 Walking

Although there were fewer temporary measures introduced to facilitate social distancing, and these were largely in town and city centres, there was still a change in the level of walking during and post-lockdown, although the picture was different from that for cycling.

Figure 6 shows that the largest increase in walking within the sample was for leisure trips during the first lockdown period, which is perhaps unsurprising as the “stay at home” directive applied to leisure as well as working, and leisure trips were only permitted near to home. Walking around local areas and greenspaces provided easy access to outdoor exercise that was strictly limited to once a day under the lockdown regulations.

Figure 6

Figure 6

Percentage increase in walking between pre-pandemic levels and during the first lockdown.

There was also an increase in the number of shopping trips. This may indicate that large weekly shopping trips by car were replaced by more frequent trips to local shops: certainly responses to the survey indicated that many changed their normal shopping (over 80%) and leisure (over 60%) destinations during the lockdown period. However, it is unlikely, given the data presented, that the increases in walking during the lockdown period were associated with the introduction of the temporary measures.

Unlike cycling, the levels of walking fell after the lockdown period, although for leisure and shopping they remained above pre-pandemic levels. This is unsurprising given around half of respondents indicated that they were still maintaining their lockdown habits in terms of leisure and shopping destinations.

It is likely that a combination of factors encouraged the shift to cycling and walking and that these vary depending on the journey purpose. However, in the case of cycling, the temporary measures were certainly an important factor in this.