Statistical Release 12 July 2016

Local Area Walking and Cycling Statistics: England, 2014/15

The percentage of people walking frequently in England has increased significantly for a second consecutive year.

The proportion of adults % walk at least: 2014/15: % point change About this release since 2012/13

walking at least once, three This Statistical Release 1x a month 86.3 0.0 

presents information on walking and five times a week

and cycling by adults (aged 16

1x a week 80.6 3.3 

and over) in England for local increased by 3.3, 7.1 and authorities and regions for the 6.9 percentage points since 3x a week 61.8 7.1  year ending mid-October 2015. The statistics in this release 2012/13. 5x a week 50.6 6.9 are based on results from the Active People Survey (APS9), % Percentage point These changes were largely change 2012/13 an annual telephone survey to 2014/15 60 administered by Sport England. driven by the increase in Total 51% 6.9 Data on cycling rates are 50 people walking for utility 1.9 available on a consistent basis Utility and recreational purposes 9%  from the survey for 5 years and purposes. For example, adults 40 3 years for walking rates. walking at least five times 30 Utility purposes only 25%  4.4

a week for utility purposes 20

In this increased by 4.4 percentage 10 Recreational purposes only 17%  0.6 publication points since 2012/13. 0 Prevalence ...... p2 2012/13 2013/14 2014/15 Local authorities...... p4 Time spent...... p8 15% of adults in England Wandsworth 12.5 Purpose...... p9 cycled at least once a month Age/Gender...... p10 South Norfolk 10.6 Urban and rural ...... p11 in 2014/15, similar to previous Barrow-in-Furness 9.8 Ethnicity ...... p12 years. Some local authorities Elevation ...... p13 City of London 9.4 had large annual increases Bicycle ownership...... p15 Surrey Heath 7.4 Safety perceptions...... p15 including Wandsworth, Road safety...... p16 South Norfolk and Barrow-in- England -0.3 Background ...... p18 -2 0 2 4 6 8 10 12 14 Furness. Percentage point change between 2013/14 and 2014/15

What we can conclude What we can’t conclude There has been a statistically signficant increase in the % of We cannot be certain of the reasons behind any changes people reporting they are walking frequently. There has been no in reported walking or cycling levels at national or local statistically significant change in reported cycling prevalence in authority level, or if what respondents report translates into England overall but there is variation amongst local authorities. their actual walking and cycling behaviour.

RESPONSIBLE STATISTICIAN: Adam Evans 020 7944 6104 FURTHER INFORMATION: Media: 020 7944 3066 Public: 020 7944 6104 [email protected] Prevalence of walking and cycling

Walking Active People 86% of adults in England walk at least once per month for any purpose, Survey which is unchanged from 2013/14 and equivalent to 38.3 million adults1. Figures in this release are based on the Active The proportion who walk for recreation (54%) is slightly lower than the People Survey (APS) administered by Sport proportion who walk for utility purposes (60%). 17% of adults reported England unless otherwise having a disability which limits their ability to walk or cycle. stated. The survey consists of 165,000 respondents aged 16 and over (defined Cycling here as adults) in England For further details see the Cycling prevalence is considerably lower than walking prevalence, with ‘Background Information’ section. 15% of adults cycling at least once per month for any purpose. This is Recreation purposes 1 equivalent to about 6.5 million people . For the pleasure or value of the activity, or enjoyment of Chart 1: Percentage of residents walking and cycling at different the surroundings. frequencies, by purpose: England 2014/15 (table CW0103) Utility purposes Getting from A to B, which might be commuting, % Any purpose % Any purpose but would also include 100 16 90 purposes such as 14 80 shopping, going to the 70 12 library, college or hospital, 60 10 or visiting friends. 50 8 40 Utility walking and 6 cycling 30 4 20 The APS asks respondents 10 2 how many days they 86 81 62 51 14.7 9.5 4.4 2.6 0 0 have walked in the past 4 At least: 1 x per 1 x per 3 x per 5 x per 1 x per 1 x per 3 x per 5 x per weeks, and how many of month week week week month week week week these days they walked % Recreational % Recreational for recreational purposes. 100 16 Walking for utility purposes is calculated by subtracting 90 14 80 days walked recreationally 12 70 from total days walked. For 60 10 example if a respondent 50 8 said they have walked 40 6 for 20 days in the past 4 30 weeks but no recreational 4 20 walking days, then it is 10 2 54 43 25 17 10.0 5.4 1.6 0.6 assumed the respondent 0 0 walked for utility purposes At least: 1 x per 1 x per 3 x per 5 x per 1 x per 1 x per 3 x per 5 x per for 20 days (a frequency of month week week week month week week week “at least 5 times a week”). % Utility % Utility If the respondent walked 100 16 for 20 days, 15 of which 90 14 were recreational, then it is 80 12 assumed the other 5 days 70 involved utility walking. This 10 60 respondent walked 5 times 50 8 a week with a combination 40 6 of recreational and utility 30 4 walking. 20 10 2 1 Based on ONS 2015 Mid 60 53 36 25 6.5 4.5 2.6 1.5 0 0 Year Estimates for ages 16+ At least: 1 x per 1 x per 3 x per 5 x per 1 x per 1 x per 3 x per 5 x per month week week week month week week week

Statistical Release - Local Area Walking and Cycling Statistics - Page 2 of 19 Changes in walking prevalence over time Walking Detailed statistics The proportion of people walking at least once a month in England has Detailed statistics on remained at 86% for the past three years. However the proportion of walking, including at local authority level, can be people walking at least once a week, 3 times a week and 5 times a week found in table CW0105. have all experienced statistically significant increases for 2 consecutive years. This suggests that there is no change in the percentage of people Definitions that walk, but that people are walking more frequently. Significant The size of the value is sufficient that we can be % point change highly confident that the % walk at least: 2012/13: 2013/14: 2014/15: since 2012/13: measure reflects the whole

population, and is not just

a random outcome (see 1x a month 86.3 86.3 86.3 0.0 

Strengths and Weaknesses

* * of the Data). A 95% 1x a week 77.3 78.9 80.6 3.3  confidence threshold

* * was used, meaning iff 3x a week 54.7 57.9 61.8 7.1  we ran the same-sized * * sample 100 times, in 95 5x a week 43.8 47.2 50.6 6.9 cases we would correctly ascribe the change as large enough to count as These increases have largely been driven by increases in walking for utility statistically significant in the purposes. For example, the proportion of people walking at least 5 times population. In the publication, changes a week for utility purposes has increased significantly by 4.4 percentage that are statistically points since 2012/13 whereas the proportion walking for recreational pur- significant are marked with an asterisk (*). poses increased by 0.6 percentage points. Time Series We have three years Chart 2: Percentage of adults walking at least 5 times a week for of comparable data for walking and 5 years of each purpose: England 2014/15 (table CW0105) comparable data for cycling (albeit only 3 years of data % Percentage point for recreational and utility change 2012/13 to 2014/15 cycling). The APS does 60 have walking and cycling Total 51%  6.9 * questions for earlier years, however the questions 50 are different and therefore  1.9 the data are not directly Utility and recreational purposes 9% comparable with each 40 other.

4.4 30 Utility purposesonly 25%  *

20

10 Recreational purposes only 17%  0.6

0 2012/13 2013/14 2014/15

Statistical Release - Local Area Walking and Cycling Statistics - Page 3 of 19 Walking prevalence in local authorities

Highest walking prevalence Nine out of the top 10 authorities with the highest percentage of adults walking at least five times a week were London authorities. Overall prevalence in England was 51%.

Chart 3: Percentage of adults walking at least 5 times a week: top 10 local authorities, England, 2014/15 (table CW0105)

Southwark 68.1 Detailed statistics Detailed statistics on Tower Hamlets 68.1 walking, including at local Hammersmith and Fulham 68.0 authority level, can be found in table CW0105. Kensington and Chelsea 67.5 Camden 65.6 Islington 65.2 Isles of Scilly 64.6 Richmond upon Thames 64.3 Wandsworth 64.2 Westminster 62.5 England 50.6 0 20 40 60 %

Largest increases Test Valley, Gosport and Coventry had the highest annual increases in the percentage of adults walking at least 5 times a week with increases of more than 14%. The overall annual increase in England was 3.4%.

Chart 4: 10 local authorities with largest increases in percentage of adults walking at least 5 times a week: England, 2013/14 to 2014/15 (table CW0105)

2014/15 rate Test Valley 57.0 15.4 Gosport 54.6 14.6 Coventry 58.0 14.6 King's Lynn & W. Norfolk 50.3 14.1 Cheshire East 55.7 14.0 Nuneaton and Bedworth 54.6 13.7 Southwark 68.1 13.7 Hartlepool 53.7 13.2 Bedford 53.3 13.1 Rother 56.3 12.7 England 3.4

0 4 8 12 16 Percentage point change, 2013/14 to 2014/15

* All changes in the chart are statistically significant

Statistical Release - Local Area Walking and Cycling Statistics - Page 4 of 19 Walking prevalence in local authorities

Changes across local Map 1: Change between 2012/13 and 2014/15 in proportion of authorities adults walking at least 5 times a week: local authorities in England This increase in frequent walking has occurred across most local authorities in England. The proportion of residents walking at least 5 times a week has increased in 301 of 326 authorities since 2012/13 as can be seen in Map 1. It is also noticeable that among authorities where there have been decreases, a number are coastal authorities. It is unknown whether this is coincidental or whether there are genuine reasons behind this, or what those reasons would be if so.

* Not all changes are statistically significant. See Table CW0105 for more detail. The increase in frequent walking appears to be driven in part by authorities that in the previous year had relatively lower walking levels compared to other authorities. Chart 5 shows that in general the lower the percentage of adults walking at least 5 times a week in 2013/14, the higher percentage point increase they experienced in 2014/15. This pattern is just a general indication and does not apply to every individual local authority.

Chart 5: Percentage of adults walking at least 5 times a week in 2013/14 and Scatter chart Each dot is a local the percentage point change to 2014/15: local authorities in England authority. The axes compare the % walking at least 5x a week in 2013/14 Percentage point change, 2013/14 to 2014/15 and the percentage point 20 change to 2014/15. The trend line shows 15 the average relationship between these. Authorities with relatively lower 10 percentages of residents walking 5x a week in 2013/14 tended to have 5 higher percentage point increases into 2014/15. 0 Because they are a 30 40 50 60 70 significant outlier, The Isles of Scilly were excluded -5 from the analysis.

-10 Walking prevalence, 2013/14

-15 * Not all changes are statistically significant. See Table CW0105 for more detail.

Statistical Release - Local Area Walking and Cycling Statistics - Page 5 of 19 Possible reasons for the increases in frequent walking

Possible reasons for increases in frequent walking Data sources There are many reasons that could have caused the increases in frequent Sunshine walking in England, although we cannot be sure what factors are having Total hours sunshine for the months October- an effect and if so how much of an effect they are having. Some possible September for each year reasons could include better weather, changes in the employment rate and was calculated from data from the Met Office. initiatives taken by public and private organisations to encourage walking. The total number of hours of sunshine has increased during a similar time Employment rate Employment rates are period in which frequent walking has increased and it could be that people available from ONS. are more likely to walk on sunnier days. The employment rate has also increased, which may result in more people walking to work. However we National Travel Survey Data for long walks (more do not know whether these patterns are related or purely coincidental. than 20 minutes) is taken from Table NTS0312. Long term trend One of the most notable With only 3 years of comparable APS data available, it is not possible to differences between the NTS and the APS is that tell if the increases in frequent walking are part of a long term trend. The the NTS asks respondents National Travel Survey does have a longer time series and indicates that to record in a diary how many walking trips they whilst the overall number of walking trips are decreasing, the percentage make, whereas the APS of people walking frequently for longer periods (defined as at least 20 asks in a telephone survey retrospectively how many minutes) has generally been increasing since 2008. days the respondent has walked in the last 4 weeks. Chart 6: Percentage of people (aged 2+) who walked for at least 20 minutes This means someone could undertake more than at least 3 times a week England 2006-2014 one walking trip in a day and this would be recorded Percentage % differently in the NTS and APS. The methodological 45 difference (keeping a diary 40 vs memory recall) may also impact how accurately 35 respondents report their 30 walking behaviour. 25 20 15 10 5 0 2006 2007 2008 2009 2010 2011 2012 2013 2014

The NTS is reporting an overall decrease in walking trips (and some increase in longer, frequent walking trips as demonstrated above), whilst the APS reports an increase in frequent walking. The reasons for this are not clear, but it may be that the methodological differences result in the same walking behaviours being recorded differently, or that instead they are capturing different kinds of walking behaviours. In general, the NTS and APS are not directly comparable due to their differences in methodologies and how walking is defined and measured.

Statistical Release - Local Area Walking and Cycling Statistics - Page 6 of 19 Cycling prevalence in local authorities Cycling In England the proportion of the population who cycled at least once per Detailed statistics Detailed statistics on month has remained at around 15%. Cycling rates at higher frequency cycling, including at local authority level, can be levels have also had no significant change. found in table CW0104.

However, cycling rates (at least once a month) vary greatly between Cycling authorities, ranging from 5% in Burnley to 43% in Oxford and 58% in prevalence Since cycling prevalence Cambridge. Cambridge and Oxford have high student populations and is relatively low, it is best students cycle more frequently than other groups. This may partly explain to use figures for those who cycle at least once why these local authorities have higher cycling rates. per month, to increase the reliability of the results. Chart 7: Percentage of adults cycling at least once a month: top 10 In England as a whole, the rate has remained local authorities, England, 2014/15 (table CW0104) at around 15% and has done so since 2010/11. Cambridge 58.0 Oxford 43.2 Students York 34.2 As shown in table Richmond upon Thames 33.1 CW0203, 22% of students Wandsworth 30.7 report cycling at least South Cambridgeshire 29.3 once per month, more Isles of Scilly 28.2 than any other NS-SEC Gosport 27.6 group. Census data Exeter 26.2 from NOMIS reveals Norwich 26.1 that students make up England 14.7 27% and 28%% of the Cambridge and Oxford 0 10 20 30 40 50 60 populations respectively. % Demographics Wandsworth had the highest annual cycling increase in England in Detailed statistics for different demographics 2014/15, increasing from 18% to 31%, followed by South Norfolk and can be found in tables CW0203. Barrow-in-Furness.

2014/15 rate Wandsworth 30.7 12.5 South Norfolk 25.2 10.6 Barrow-in-Furness 19.1 9.8 City of London 16.9 9.4 Chart 8: 10 local authorities Surrey Heath 18.7 7.4 with largest increases in Rushmoor 19.1 7.2 percentage of adults cycling Mid Devon 21.2 6.9 at least once a month: Rotherham 15.3 6.8 Worcester 18.7 6.0 England, 2013/14 to 2014/15 Croydon 12.8 5.9 (table CW0104) England 14.7 -0.3 -2 0 2 4 6 8 10 12 14 Percentage point change, 2013/14 to 2014/15

* All changes in the chart are statistically significant except England and City of London

Statistical Release - Local Area Walking and Cycling Statistics - Page 7 of 19 Time spent walking and cycling

About 14.7% of the population cycled at least once a month. About 3.4% Detailed statistics of the population usually cycled for less than half an hour, 4.1% usually Detailed statistics on time cycled for between 30 minutes and 1 hour, and a further 6.8% usually usually spent walking and cycling, including at local cycled for one hour or more. Of those people walking or cycling for more authority level, can be found in tables CW0104 than half an hour, a considerable majority do so for recreational purposes. and CW0105.

Definition Chart 9: Time usually spent cycling: England 2014/15 (table CW0104) Time usually spent is not % the same as average Any cycling time. For example if 5 someone cycled for half Recreational cycling an hour each weekday but for 2 hours during weekends, then usual 4 time would be half an hour. Time usually spent 3 walking/cycling for recreation is available from the Active People Survey, but not the time 2 spent for utility purposes A small number of people did not state how long they usually walked or 1 cycled for, which is why the figures do not add up to the proportions who 3.4 1.0 4.1 2.4 4.1 3.4 2.7 2.8 walked/cycled at least 0 once a month (86.3% and < half hour half to <1 hour 1 to <2 hours 2 to 17 hours 14.7%). Time per day usually spent cycling Chart 10: Time usually spent walking: England, 2014/15 (table CW0105) Distance The APS does not ask respondents how far % Any walking they travel, but according 35 to the National Travel Recreational walking Survey there has been a long term increase in the 30 distance people cycle, up 26% in 2014 compared 25 to 1995/7. However for walking, there has been a 9% decrease in the 20 distance travelled over the same time period. 15 Further statistics 10 Further statistics on distance travelled when 5 walking and cycling are available from the 29 10 30 18 17 15 9 7 National Travel Survey. 0 < half hour half to <1 hour 1 to <2 hours 2 to 17 hours Time per day usually spent walking

Statistical Release - Local Area Walking and Cycling Statistics - Page 8 of 19 Factors influencing walking and cycling: purpose and age Walking There are clear age differences depending on whether people walk for recreation or utility. For utility purposes, walking prevalence is highest among the youngest adults and declines steadily with age. Prevalence of walking for recreational purposes, however, tends to increase with age up to the 55-64 age group before declining afterwards. These patterns hold for both general and frequent walking (at least once per month, and at least 5 times per week).

Chart 11: Percentage of adults recreational and utility walking at least once a month, by age: England, 2014/15 (table CW0203) Recreational Walking % Definitions 90 Utility Walking 80 Recreation purposes For the pleasure or 70 value of the activity, or enjoyment of the 60 surroundings. 50 Utility purposes: 40 Getting from A to B, which 30 might be commuting, but would also include 20 purposes such as shopping, going to the 10 library, college or hospital, 47 84 53 73 58 64 61 56 62 53 52 49 or visiting friends. 0 16-24 25-34 35-44 45-54 55-64 65+ Age Cycling Cycling shows a different pattern by age and purpose from walking. Firstly, among cyclists generally, recreational cycling is more prevalent than utility cycling. For both purposes, though, there is a decline in prevalence between 16-24 and 25-34, followed by a rise in the 35-44 age group. Cycling prevalence then steadily declines with age from that group onwards.

% Recreational Cycling 16 Utility Cycling

14

12 Chart 12: Percentage of

10 adults recreational and utility cycling at least 8 once a month, by age: 6 England, 2014/15 4 (table CW0203) 2 13.110.5 11.3 7.9 14.6 8.3 12.6 7.0 8.4 4.8 3.1 2.2 0 16-24 25-34 35-44 45-54 55-64 65+ Age Statistical Release - Local Area Walking and Cycling Statistics - Page 9 of 19 Factors influencing walking and cycling: age and gender

Walking Walking levels between men and women tend to be quite similar at all ages. On average, both 86% of men and women walk at least once per month. Walking levels between age groups are also similar although they do decline gradually with age.

Chart 13: Percentage of adults walking at least once a month for any purpose, by age and gender: England, 2014/15

% 100

86% 86% 80

60

40

20

92 94 90 91 90 91 88 88 85 86 76 74 0 16-24 25-34 35-44 45-54 55-64 65+ Age Cycling There is a much greater difference between men and women and across age groups for cycling. Cycling levels peak between ages 35-44 before declining thereafter. A lower proportion of women cycle than men in all age groups. Overall, 20% of men cycle compared to 10% of women.

Chart 14: Percentage of adults cycling at least once a month for any purpose, by age and gender: England, 2014/15

% 30

25 20% 10%

20

15

10

5

26.0 14.5 22.4 11.6 26.9 14.3 23.2 11.9 16.1 8.0 7.6 2.6 0 16-24 25-34 35-44 45-54 55-64 65+ Age

Statistical Release - Local Area Walking and Cycling Statistics - Page 10 of 19 Factors influencing walking and cycling: urban and rural areas

Urban and Rural Areas Urban and rural Walking and cycling for recreational purposes is more prevalent in rural classifications areas, but more prevalent in urban areas for utility purposes. For walking, The Department for Environment, Food & these two tendencies cancel each other out, so that there is not any Rural Affairs 2011 Rural- noticeable overall difference in prevalence between urban and rural areas. Urban Classification of local authority districts However overall cycling prevalence levels are higher in rural authorities. and other high level geographies can be found here. Chart 15: Percentage of adults walking and cycling once a month in rural and urban authorities: England, 2014/15 (table CW0203) 1 Based on ONS 2015 Mid Year Estimates for ages 16+

61 7 Utility Utility 54 6

55 10 Recreational Recreational 63 Urban authorities 12 Rural authorities

86 15 Any Any 87 16

0 20 40 60 80 100 0 5 10 15 20 % % While rural authorities have a higher percentage of adults that walk recreationally than urban areas (63% vs 55%), there are more people who walk recreationally in urban areas (5.9 million vs 19.2 million) because urban areas have higher populations1. Chart 16 demonstrates this, with the majority of rural authorities having a higher recreational walking prevalance levels but lower populations compared to many urban authorities.

Chart 16: Population and recreational walking rate: local authorities in England, 2014/15

% walking at least 1x a month 100

95

90

85

80 Urban authorities 75 Rural authorities 70 0 100,000 200,000 300,000 400,000 500,000 600,000 Population aged 16+

Statistical Release - Local Area Walking and Cycling Statistics - Page 11 of 19 Factors influencing walking and cycling: ethnicity

Walking Ethnic group Overall walking prevalence is similar across all ethnic groups (between Further walking and cycling statistics by 86%-90%) although there are differences between groups in the purpose ethnicity and other for walking. White British adults walk for recreational purposes more than demographics are available in Table most ethnic groups but walk for utility purposes the least. Conversely, CW0203. Black or Black British adults walk the least for recreational purposes but 1 Census data from one of the most for utility purposes. NOMIS reveals that White British residents Chart 17: Percentage of adults walking at least once a month, by ethnicity make up 95% of the rural and purpose: England, 2014/15 (table CW0203) population but 77% of the urban population. Utility Recreational

Chinese 83 White-Other 59

Black or Black British 76 White-British 58

Mixed 74 Other ethnic group 53

Asian or Asian British 73 Mixed 48

White-Other 72 Chinese 47

Other ethnic group 71 Asian or Asian British 40

White-British 60 Black or Black British 35

0 20 40 60 80 100 0 20 40 60 80 100 % % Cycling Unlike walking, there are noticable differences in overall cycling prevalence between ethnic groups. Chinese adults have the highest prevalence rate cycling at least once a month (20%), whilst Asian or Asian British adults have the lowest (9%).

Chinese 19.6

White-Other 18.3

Mixed 15.8 Chart 18: Percentage of adults White-British 15.0 cycling at least once a month Other ethnic group 12.2 by ethnicity: England, 2014/15 Black or Black British 9.4 (table CW0203) Asian or Asian British 8.6

0 5 10 15 20 % It is important to remember that these differences are not necessarily all due to cultural differences. For instance, there are higher proportions of White British residents living in rural areas relative to urban areas than other ethnic groups1. As rural residents are more likely to walk or cycle recreationally and urban residents for utility purposes, this may be impacting the results we see here. Other factors, such as differences amongst ethnic groups in employment, income, or car ownership, might also have an effect. Statistical Release - Local Area Walking and Cycling Statistics - Page 12 of 19 Factors influencing walking and cycling: elevation Map 2: Elevation map of England (source: Ordnance The difference in walking and cycling Survey terrain data) with selected local authority elevation patterns between urban and rural ranges and once per month utility walking rates areas may be partly influenced by their elevation levels. Urban authorities tend to be flatter whereas rural areas can have more varied and sometimes hilly terrains. One way of measuring variations in the elevation is by using the range between the highest and lowest land level points in each local authority, which makes it possible to investigate the relationship between terrain and walking and cycling levels.

Elevation above sea level

0m 300m

Walking Methodology for Overall, walking prevalence had no relationship with the elevation range of elevation analysis local authorities. However, recreation walking rates are positively correlated Ordnance Survey Terrain 50 data are available for with elevation range and utility walking rates are negatively correlated with 50m2 grids in England. Each grid indicates how elevation range. This means that the greater the elevation range, and so far the land data point is likely more hilly, recreational walking rates tend to be higher. above/below sea level (and is freely available here). These data were Chart 19: Elevation range and percentage of adults walking once a month mapped against local for recreation: England, local authorities, 2014/15 authority boundaries and made it possible to identify the range % recreational walking between the highest and 100 lowest land level points for each local authority 90 Data points below sea 80 level were counted as 0 rather than negative 70 as there were some authorities with artificially 60 low minimum land 50 levels which were not representative of the 40 terrain residents typically walk/cycle along. For 30 example, Charnwood 20 has a quarry which goes to 129 metres below sea 10 level but it is unlikely residents walk or cycle in 0 the quarry. 0 200 400 600 800 1000 Elevation range (metres)

Statistical Release - Local Area Walking and Cycling Statistics - Page 13 of 19 Factors influencing walking and cycling: elevation

The opposite is true for utility walking: the greater the elevation range the Regression lower utility walking levels tend to be. analysis % utility walking Chart 20: Elevation range and A linear regression 100 percentage of adults walking revealed terrain range is significantly able to 90 once a month for utility: predict walking and cycling levels. 80 England, local authorities, We can use the standard 70 2014/15 regression formula to work out to predict how 60 much terrain range affects 50 walking and cycling levels: 40 30 Recreational walking For every 100m increase 20 in the elevation range, a 10 1.7% recreational walking increase would be 0 predicted. 0 200 400 600 800 1000 Elevation range (metres) Utility walking This is similar to the rural and urban walking patterns and so could be a For every 100m increase in the elevation range, sign that elevation levels do partly influence walking patterns. It may be that a 1.4% utility walking decrease would be in rural areas varied elevation levels may be more scenic and enjoyable for predicted. recreational walking, whereas urban areas with flatter elevation levels may be more practical for walking for utility purposes such as walking to work. Utility cycling For every 100m increase Cycling in the elevation range, a 0.8% utility walking The same, but weaker, negative correlation exists for cycling between utility decrease would be predicted. purposes and elevation range however there is no relationship between terrain levels and cycling for recreational purposes. Whilst these formulas are better at predicting Chart 21: Prediction of how much elevation affects walking and cycling rates walking and cycling levels than by a random (at least once a month) in local authorities: England, 2014/15 guess, there are many other influencing factors and therefore should be treated as a general Recreational Utility Utility walking Walking Cycling indicator. Other factors Other factors Other factors 83% 89% 83%

Elevation Elevation Elevation 17% 11% 7%

It is important to point out that the scale of these relationships, where they exist, are not particularly strong and there are many other factors that influence walking and cycling levels.A regression analysis revealed that elevation explained 17% and 11% of the patterns in recreational and utility walking levels across local authorities. This means that 83% of recreational walking and 89% of utility walking patterns are explained by other factors.

Statistical Release - Local Area Walking and Cycling Statistics - Page 14 of 19 Factors influencing cycling: bicycle ownership and safety perceptions

Bicycle ownership In England, 42% of people aged over 5 own a bicycle. Bicycle ownership is most prevalent amongst people aged under 17 years old. Amongst adults, bike ownership peaks at ages 40-49 with 51% owning or having access to a bicycle.

Chart 22: Bicycle ownership or access, by age: England, 2014/15 % 90 80 70 Bike Ownership 60 Bike Ownership statistics from the 2014 50 National Travel Survey 40 are available on Table NTS0608 30 20 10 84 73 45 35 44 51 43 21 0 5-10 11-16 17-20 21-29 30-39 40-49 50-59 60+ Age

Safety Perceptions British Social Safety perceptions may partly influence cycling level differences between Attitudes Survey men and women, as the British Social Attitudes Survey found that women The British Social Attitudes Survey is conducted were more likely than men to believe cycling was too dangerous for them annually with a sample of 3,000 people and the (71% vs 57%). survey includes transport related questions. Drivers: 63% 71% 57% Non-drivers: 66%

Chart 23: Differences in “It is too dangerous for cycling safety perception, me to cycle on roads” Great Britain, 2014

Cyclists: 51%

Non-cyclists: 70%

Another possible reason influencing gender differences could be that on average women tend to undertake different journeys to men, such as being more likely to take children to school, according to the National Travel Survey. These types of journeys may be less suitable by bike. However we cannot be certain whether these are reasons behind the difference and if so how much of an influence they have, and there are also likely many other factors involved.

Statistical Release - Local Area Walking and Cycling Statistics - Page 15 of 19 Walking and cycling safety

Comparison with other modes Walking and cycling are less safe relative to some other modes of transport. However at an absolute level they are a safe way of travelling and serious accident are rare. In 2015, there was roughly 1 Killed or Seriously Injured (KSI) casualty per 1 million miles cycled and 1 KSI casualty per roughly 100,000 hours of cycling.

Chart 24: Casualty and fatality rates per billion passenger Road Safety miles by road user type: Great Britain 2015 Statistics The road safety statis- Vulnerable road user groups tics presented on pages 7,000 140 16 and 17 are from the Reported Road Casual- 6,000 120 ties Great Britain 2015 report, Road Accident

5,000 100 Safety Dataset and Road Safety Factsheets. Pedal cycle traffic is available 4,000 80 here. Further information can be found in these 3,000 60 releases.

2,000 40 ‘Other’ in Chart 25 consists of KSIs on Fatality Fatality rate per billion passenger miles

Casualty Casualty rate per billion passenger miles 1,000 20 motorways and that have unknown locations. 273 1.8 188 0.2 76 0.5 72 1.9 2,108 35.8 5,824 30.9 6,676122.3 0 0 1 1

Walking The number of pedestrians killed or seriously injured has generally been decreasing since the 1980s. Urban roads account for more than four fifths of pedestrian KSI casaulties across all road types.

Chart 25: Pedestrians killed or seriously Chart 26: Percentage pedestrian KSIs by injured, Great Britain: 1979-2015 road type, Great Britain: 2015

Other 1

25,000

20,447 Rural 20,000 Roads 15

15,000 Pedestrian KSIs in 10,000 5,349 2015: 74% 5,349 5,000 since 1979 0 1979 2015 Urban Roads 84

Statistical Release - Local Area Walking and Cycling Statistics - Page 16 of 19 Road Safety Cycling Although the number of cyclists killed on the roads in 2015 was the lowest figure on record, the 100 fatalities is very similar to the figures for each year since 2008. Since 2008, the number of deaths has been between 100 and 118. In statistical terms, there has been no change in the number of fatalities over this period. Chart 27: Number of KSI cyclists compared with cycle In contrast, the number of cyclists traffic, Great Britain: 2000-2015 seriously injured has generally been Index: 2010-14 average = 100 140 2015 14-15 % rising since the lowest point in 2004. figure change

Although there was a fall of 5 per 120 Seriously injured

cent to 3,237 serious injuries in 3,237 4.8% 100 2015, this is still the second highest Pedal cycle traffic year since 1997. This long term 80 3.2 bvm 6.1% increase is likely caused by the 60 Killed

long term increase in cycle traffic, 100 11.5% reflecting an increase in the number 40

of cyclists exposed to potential 20 dangers. 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Chart 28: Cycle traffic and casualties on rural and urban roads, percent, Great Britain 2015 Urban Rural Urban and Rural Casualties Cycle traffic Despite carrying only 33% of cycle traffic, nearly half (49%) of cyclist deaths occur on rural roads. Accidents that occur on rural 67 33 roads are more likely to be of a fatal nature compared to those on urban roads. This is because rural roads have a much higher Killed (99) average speed than urban roads. Rural roads are often also much more sinuous and narrow in nature, with blind bends, dips 49 51 and other distractions. However the pattern differs for non fatal casualties. Accidents on urban roads are more likely to to result in All casualties (18,838) slight or serious injuries. Urban and Rural 81 19 casualties Overall, the majority of cyclist casualties The urban and rural Seriously injuried (3,234) occur on urban roads (81%) which is figures do not include motorway casualties or casualties where the 70 30 unsuprising given that the majority of location of the accident cycling takes place here and that there was unknown. Therefore are many more opportunities for cyclists the sum of urban and Slight injuries (15,505) rural casualties will not to interact with other vehicles. quite add up to total 84 16 casualties.

Statistical Release - Local Area Walking and Cycling Statistics - Page 17 of 19 Background information Users and uses of these statistics Within the Department for Transport, we anticipate these statistics being used in the evaluation of local area interventions to encourage sustainable travel (for example, the Local Sustainable Transport Fund), as background information in the development and targeting of policies, for ministerial briefing and to answer public enquiries. Other users include local authorities, campaign organisations, Parliamentary Groups, researchers and individuals with an interest in walking or cycling.

Strengths and weaknesses of the data The figures in this release unless otherwise stated are based on theActive People Survey (APS), administered by Sport England and used to derive official estimates of participation in sport and active recreation. The APS has a sample size of over 160,000 persons in England, thus enabling analysis at local authority level. Statistics at this level are not available from the National Travel Survey (NTS), which has an annual sample of around 16,000 persons. Results from the APS are grouped by the area where survey respondents live, which may not be the same as the area where they walk or cycle, particularly for urban areas where there are multiple local authorities in a relatively small area. Although the APS has a standard sample size of at least 500 persons per local authority, because the numbers of those cycling are small (only 15% nationally), some of the measures relating to cycling are based on only a few people per local authority and may not be robust. The tables accompanying this release include 95% confidence intervals for the estimates derived from the survey, to demonstrate the reliability of the estimates and the likely range of values for the true value1. Of nearly 370 authorities and regions, this means that some intervals will include errors, but it is not possible to specify which ones: the confidence interval may not contain the true value for the population, or a change may show as statistically significant when it is not. The tables also highlight whether the change in walking or cycling activity since 2013/14 is statistically significant, rather than simply due to random variation in the survey sample. More details of the statistical test used to assess significant change can be found in the accompanying Notes and Definitions document. Caution should be taken when interpreting these changes because of factors that include small sample sizes and the inherent uncertainty in doing any statistical test. The APS sample is weighted to ensure that the results are representative of the population. However, it is exclusively a telephone survey and only covers households with a fixed landline. The 15% of households in England that are mobile-only2 are excluded from the survey, which may introduce bias into the sample. From 2015-16, the Active People Survey will be suceeded by the Active Lives Survey. The walking and cycling questions will remain exactly the same however the methodology is changing. Rather than contact participants by telephone, a letter will be sent with a unique link inviting them to take part in the survey online. This is hoped to reduce the sample biases existent in the APS. Sport England will be running both methodologies simultaneously for a period of time to investigate the effects of this change, which will be accounted for during analysis.

1The confidence intervals used are Wilson Score intervals. For more details, see the accompanying Notes and Definitions document. 2 See paper commissioned from the ONS Methodology Advisory Service, available at: http://www.sportengland.org/research/ active_people_survey/consultation.aspx Statistical Release - Local Area Walking and Cycling Statistics - Page 18 of 19 Background information

Further information The web tables give further details of the results presented in this statistical release. They are available here: www.gov.uk/government/organisations/department-for-transport/series/walking-and- cycling-statistics Guidance on the methods used to compile these statistics, including the calculation of confidence intervals, is available in the “Notes and Definitions” document, which can be found here: www.gov. uk/transport-statistics-notes-and-guidance-walking-and-cycling Details of ministers and officials who receive pre-release access to these statistics up to 24 hours in advance can also be found at the link above. Further information about the Active People Survey and published sports participation measures for APS9 can be found on Sport England’s website: www.sportengland.org/research/active_people_ survey.aspx

Request for feedback We welcome any feedback on these statistics, to ensure future releases best meet user needs. Feedback can be provided by email to [email protected].

Next update The next release in this series is due to be published in Spring 2017 and will contain statistics on walking and cycling from the Active Lives Survey, the successor to the current Active People Survey. It will cover the period October 2015 to October 2016. Other sources of information on walking or cycling • The National Travel Survey: https://www.gov.uk/government/collections/national-travel-survey- statistics • Road Accidents & Safety statistics: https://www.gov.uk/government/collections/road-accidents- and-safety-statistics • Road Traffic statistics: https://www.gov.uk/government/collections/road-traffic-statistics • The 2011 UK Census contains detailed information on mode of travel to work (which includes walking or cycling), available via NOMIS: https://www.nomisweb.co.uk/census/2011 • The Travel in London report provides walking and cycling statistics for London http://content.tfl. gov.uk/travel-in-london-report-8.pdf • The Propensity to Cycle Tool is an interactive tool which maps the cycling rate potential at local levels http://pct.bike/ • The British Social Attitudes Survey includes numerous questions about perceptions of walking and cycling https://www.gov.uk/government/uploads/system/uploads/attachment_data/ file/481877/british-social-attitudes-survey-2014.pdf • The Scottish Household Survey contains walking and cycling statistics for Scotland http:// www.transport.gov.scot/statistics/j389989-08.htm • The National Survey for Wales contains walking and cycling statistics for Wales http://gov. wales/statistics-and-research/active-travel/?lang=en • Cycling and walking to work statistics for Northern Ireland https://www.gov.uk/government/ statistics/cycling-and-walking-to-from-work-in-northern-ireland-201415

Statistical Release - Local Area Walking and Cycling Statistics - Page 19 of 19