The Scottish Government

« Previous | Contents | Next »

Listen

Chapter Seven Understanding the long distance commuter quantitative research

Introduction

7.1. This chapter aims at understanding the circumstances, characteristics, behaviour and attitudes of the long distance commuter. In particular, identifying statistically-significant differences between the long distance commuter and the average commuter recorded in the Scottish Household Survey ( SHS) data. Consideration was given to trips to educational establishments, as well as travel to work.

7.2. In order to explore the differences the following techniques were employed:

• SPSS Classification Tree-based analysis
• Analysis of Variance ( ANOVA)

7.3. The Scottish Household Survey data indicates that 24% of all respondents' (N=38,259) travel further than 15km when commuting to work.

7.4. SPSS Classification Tree-based analysis was used to detect which SHS attributes distinguish the long distance commuter from the short distance one. An initial analysis suggested that the following variables provide the most significant variation in the proportion of long distance commuting:

• mode
• employment status
• household car ownership
• sex
• home urban/rural classification
• home Local Authority area
• household income

7.5. Modes used for commuting provided the most significant variation in the proportion of long distance commuting. Table 7.1 shows the percentage of long distance commuters by mode. It may be seen that 56% of train-commuters travel further than 15km to work ( i.e. are long distance commuters) and 33% of car commuters travel further than 15km to work.

Table 7.1: Percentage of long distance commuters by mode ( SHS February 1999 - March 2005)

7.6. A more detailed analysis was then undertaken for those commuting by car. Within the car-based commuter category, the strongest Classification-Tree identified partition is the urban/rural classification of the respondents' home. For example, a much higher percentage of car commuters living in 'accessible rural' areas travelled long distances (48%) to their work than car commuters living in 'urban settlements' (18%).

7.7. The percentages of long distance commuting for each category of the standard Scottish Executive 6-fold urban/rural classification are shown in Table 7.2.

Table 7.2: Car long distance commuters by urban/rural 6 fold classification ( SHS 1999-2005)

Inter-relationships between variables

7.8. Preliminary Classification Tree analysis of the SHS data highlighted the need to investigate the inter-relationships between variables using ANOVA techniques. As part of this process, mean distances for a number of variables had to be determined. These values are considered next, prior to the reporting of the ANOVA analysis itself. Variables considered include: mode, employment status, car ownership, age, social class, urban/rural 6-fold classification, household income and household type. The main findings of the analysis are summarised below. Further details are provided in Technical Annex 3.

SHS and average commuting distance

7.9. Analysis of SHS data highlights the following differences in mean commuting length. It should be noted that the differences in the mean travel distance in each of the homogeneous subsets listed are significant (Tukey HSD, see Technical Annex 3):

• train-commuters commute further (average of 26km) than car commuters ( 15km) and bus commuters ( 10km)
• those employed full-time travel further to work (mean 13km) compared to part-time workers (mean 7km) and the self employed (mean 4km)
• those from two car households travel an average of 15km to work, compared with those from one car household who travel 11km and those from households without a car who travel an average of 7km
• those aged between 20 and 60 years travel, on average, 11km to work, slightly further than the under 20s (mean 9km) or the over 60s who are still working ( 8km)
• those in professional occupations travelled an average of 14km to work. Those in managerial and technical occupations travelled an average of 13km. Those in partly-skilled occupations travelled an average of 9km and unskilled occupations travelled an average of 7km
• on average, commuters in 'accessible rural' areas (mean distance 15km) and 'accessible towns' ( 14km) travel much further than those from 'large urban' ( 8km)areas and 'small remote towns' ( 8km)
• in general commuting distance tends to increase with increasing household income. Those from households with more the £40,000 annual net household income commute an average of 17km In contrast, those in households with income less than £10,000 commute an average of 8km
• small family households (mean commute 12km) and large family households ( 12km) tend to generate slightly longer commutes than single adult households which have an average commute of 11km and single parents commute an average of 8km

Analysis of variance

7.10. In the above section, we explored whether the explanatory variables are individually statistically different for mean distance travelled. Using Univariate Analysis of Variance, consideration was then given to the interactions between paired variables. These were analysed further with post-hoc tests. The results of the analyses are presented in Annex 3. These comparisons will only be mentioned here and throughout, when they are reliable, at least at the p <0.05 level.

7.11. Social Class was excluded from the full analysis of variance. This was because the ANOVA removes records where there is a 'missing' value for any of the variables included within the full analysis. A large proportion (31% N=45,299) of SHS records do not have a social class label associated.

7.12. A second ANOVA considered car commuters only. As a consequence main mode and car ownership were excluded. Employment status was also excluded to shift the focus to those who are commuting by car every day rather than those who may only commute once or twice a week.

7.13. It was noted that the interaction between the pairing of gender and other variables was significant. In addition, the relationship between rural/urban classification and household structure together with household income and household structure were also found to be important. These relationships are now explored in more detail.

Gender and age

7.14. Figure 7.1 details the mean distance travelled to work for both males and females by age. Females tend to commute shorter distances than males across all age groups. Males aged 35 to 39 commute the furthest (mean commute is 15km). For females, those aged 25 to 29 travel the furthest (11km).

Gender and income

7.15. Figure 7.2 details the mean distance to work for both males and females for different household income groupings. For all income bands, females travel a shorter average distance to work than males. As indicated earlier, there is a general trend for longer travel distances for higher earners. Males from households with incomes 'over £40,000' travel furthest (mean 21km). Similarly, for females those with incomes 'over £40,000' travel furthest (mean of 13km).

Gender and urban/rural area

7.16. Figure 7.3 details the mean distance travelled by urban/rural area for both males and females. Females travel shorter mean distances than males for each urban/rural classification. Males from 'small accessible towns' (17km) and 'accessible rural areas' (17km) commute the furthest. Similarly for females, those from 'accessible rural' areas commute the furthest (13km).

Gender and social class

7.17. For all Social class groupings, female commutes were shorter than male. Males from 'Managerial and technical occupations' travelled the furthest (mean commute 16km). Females from 'professional occupations' travelled further (13km) than females in other social class categories.

Gender and household structure

7.18. To further investigate the differences between male and females, the seven SHS household types were further aggregated into four groups using household groupings suggested by the initial (Tukey HSD) analysis (See the Technical Annex for details). Table 7.3 outlines household type definitions.

Table 7.3: Household type groupings

7.19. Males from families with children commute the furthest (mean 15km). In contrast, females from families with children commute an average of 9km. Males and females from households with no children commute an average of 13km and 10k respectively. For 'older/pensioners (no kids)' males travelled an average of 10km and females an average of 7km. There were no significant differences between the commuting distances of male and female single parents.

Income and household type

7.20. The general trend for all household types is an increase in travel distance with income, with the exception of households with children on incomes less than £10,000. It is also of interest to note that very low income (less than £10,000) households with children commute further than households with no children with a similar income.

Average commuting distance and single occupancy households

7.21. Figure 7.7 details the mean distance to work for single occupancy household for both males and females with and without a car available for different household income groupings. Figure 7.7 indicates that females with a car available commute shorter distances than males with a car available, with the exception of those from low income households (less than £10,000). Females without a car available commute shorter distances than males without a car available across all income groupings. This research suggests that a significant difference between male and female commuting distance still remains, once income and car ownership have been taken into consideration.

Long distance travel to education

7.22. SHS data includes responses to a number of questions about young people in the household. These include 'usual means of travel to school', 'reasons for choosing this mode' and 'availability of public transport'. This information was used to explore the nature of the variations in the distance travelled to educational establishments.

7.23. Road distance estimates were calculated for each response. These were based on the data zones of the child's' home and school. Using road based distance estimates, only 4% of children travel an equivalent to long distance commuters (further than 15km to school). Detailed distribution of distances travelled to school is provided in Table 7.4.

Table 7.4: School child: road-based distance estimates between home and school

7.24. An analysis was undertaken using Classification-Tree software to compare those who travelled further than 15km with those who travelled less than 15km. As may be expected, a higher percentage of children from households in rural areas travelled more than 15km, compared to those from towns or urban areas. For example, 19% of children from 'remote rural' area and 8% of children from 'accessible rural' areas travelled more than 15km to school. In contrast, only 0.4% of children in 'large urban settlements' and 0.4% of children from small remote areas travelled more than 15km to school.

Table 7.5: Percentage of children travelling 15+km to/from school by urban/rural classification ( SHS February 1999 to March 2005)

Additional comparisons between short and long distance commuters

7.25. The SHS asked parents a number of questions about a randomly-selected school pupil (if there is one) in the household. Responses to the question 'how satisfied or dissatisfied are you with the schooling your child is receiving?' may provide an indication of the importance of schools in family household location decisions for long and short distance commuters. However, there was no significant difference between long distance and short distance commuters in their satisfaction with schooling. Table 7.6 indicates that 94% of short distance commuters were satisfied (either 'very satisfied' or 'fairly satisfied') with schooling. Similarly, 94% of long distance commuters were satisfied with their child's school. This implies that moving house to be in a good school catchment area is not a major reason for long distance commuting.

Table 7.6: Satisfaction with schooling ( SHS February 1999 to March 2005)

*Asked of all SHS participants with a young person in the household aged 18 or under (who they or their partner are responsible for)

7.26. Previous research suggests that people who have lived at the same address for a long time may be more willing to commute further distances rather than moving home to be closer to their work place. However, Table 7.7 suggests that the distributions of time at the current addresses are very similar between short and long distance commuters. Table 7.7 shows that 15% of short distance commuters had lived at their current address for twenty years or more. 12% of long distance commuters had lived at their current address for twenty years or more.

Table 7.7: Years at current address for short and long distance commuters ( SHS February 1999 to March 2005)

7.27. The brief for this study also asked us to consider factors which prevent long distance commuting. This is particularly important for the vulnerable groups in society such as disabled people. In SHS Item RG5, the random adult respondent was asked, 'Could I just check, do you have any long-standing illness, health problem or disability that limits your daily activities or the kind of work that you can do?' Table 7.8 details responses to this question for short and long distance commuters. It may be seen that 7% of short distance commuters reported having a disability, illness or health problem. Slightly lower percentages of long distance commuters reported having a disability, illness or health problem (6%). This suggests that disability and ill-health creates a very small (but statistically significant) barrier to long distance commuting.

Table 7.8: Illness and/or disability and percentages long and short distance commuting ( SHS March 2003 to March 2005)

7.28. Interviewers asked the randomly selected adult how often they had used their local bus service in the past month. Table 7.9 shows that short distance commuters make more use of local bus services than long distance commuters. 37% of short distance commuters reported using the bus in the past month, compared to 21% of long distance commuters. This difference will be a combination of the different levels of bus use for long and short commuting trips and the increased proportion of rural residents who are long distance commuters.

Table 7.9: Frequency of local bus use for long and short distance commuters ( SHS February 1999 to March 2005)

7.29. In SHS, the random adult respondent was asked, 'What aspects of this immediate neighbourhood, if any, do you particularly like?' A greater percentage of short distance commuters (16% N= 29,272) liked the 'good public transport' in their neighbourhood, compared to long distance commuters (11% N=8,987). The random adult respondent was also asked, 'What aspects of this immediate neighbourhood, if any, do you particularly dislike?' Slightly more long distance commuters (8% N=8,987) disliked the 'poor public transport' compared to short distance commuters (5% N= 29,272).

Mode choice

7.30. Consideration is now given to some of the factors which may influence both commuting mode choice and distance. The choice of home and workplace location and the availability, cost and convenience of transport alternatives will clearly be primary factors determining commuting mode and distance.

7.31. The SHS interviewer asks the randomly selected adult a number of questions on why he or she chooses a particular method of transport to work. Participants could provide more than one reason. Figure 7.9 details reasons for current mode choice for both long and short distance commuters. The reasons for using his or her means of transport to work given most often by long distance commuters were 'most convenient' (58%) , 'only method available' (21%) and 'public transport unavailable' (14%). Short distance commuters most frequently cited reason for choosing their main mode of transport were 'most convenient' (58%), 'close, nearby, not far away' (18%) and 'only method available' (9%). Proportionately more long distance commuters choose their current mode because it was 'the only method available' or 'public transport was unavailable'.

Car commuters and mode choice

7.32. Car commuters were asked, 'Would it be possible for you to use public transport for the journey to or from work?' SHS (February 1999 to March 2005) indicates that long distance car commuters (37% N=7,293) were significantly less likely to claim to have the option of using public transport when compared to short distance car commuters (49% N=14,833).

7.33. Table 7.10 shows the percentage of long distance car commuters who have a public transport option available. The rows show the nature of the home urban/rural area, whilst the columns show the nature of the workplace area. It may be seen that long distance car commuters living and working in 'large urban settlements' (52%) were most likely to have a public transport option. Long distance car commuters living in 'remote rural areas' and working in 'towns and rural areas' were least likely (19%).

Table 7.10: Percentage of long distance car commuters with a 'public transport option' and home/workplace urban/rural classification ( SHS February 1999 to March 2005)

7.34. The interviewer then asked those who stated that they usually travel to work by car and could use public transport, the reasons why they cannot do so, or choose not to do so. If the SHS participant provided more than one reason, the interviewer asked the main reason. Figure 7.10 illustrates the reasons for not using public transport given by car commuters, who stated that they could do so. Proportionately, more long distance car commuters than short distance commuters stated that they did not use public transport because it 'takes too long', 'no direct route' or 'costs too much'.

Mode change

7.35. From April 2003, the interviewer asked randomly selected adults, who had specified a usual method of travel to work, whether they were in employment a year ago, and if so, what was their main method of travel to work a year ago.

7.36. The percentage of commuters who change mode during a year by the mode used in the initial year is shown in Table 7.11 for both long and short distance commuters. Clearly, mode choices were fairly stable. Table 7.11 indicates that only 8% of individuals had changed mode during the year. As might be expected, there were considerable difference in the proportion who changed for the various modes. Train passengers were most likely to change mode with 20% changing during a year. Car drivers were the least likely to switch modes (3% change in any year).

7.37. Slightly less long distance commuters (7%) changed mode when compared with short distance commuters (8%). All SHS participants who were walking to work a year previously and are now long distance commuters were no longer walking. Presumably, many of these participants had increased their commute distance (either by moving home and/or changing jobs) since the previous year.

Table 7.11: The percentage of commuters who change mode during a year by the mode used in the initial year for long and short distance commuters ( SHS April 2003 to March 2005)

7.38. For those who changed mode in the previous year (N=1,016), there was a definite trend towards greater reliance on the car for travel to work (32% previous year; 46% current year). Mode used a year ago and mode currently used are detailed on Figure 7.11 for both short and long distance commuters. The changes in mode share are detailed on Figure 7.12.

7.39. For SHS participants who had changed mode since the previous year, details of the mode used a year ago and the current mode are shown on Figure 7.13, for short distance commuters, and Figure 7.14 for long distance commuters. It may be seen that:

• there was an increase in the percentage of car drivers for both long and short distance commuters (short distance 6% increase; long distance 16% increase)
• there was an increase in the percentage of car passengers for both long and short distance commuters (short distance 6% increase; long distance 2% increase)
• there was a decrease in the percentage of short distance commuters using a train, in contrast to an increase in the percentage for long distance commuters. (short distance 2% decrease; long distance 11% increase)
• there was an increase in the percentage of short distance commuters who walked or cycled and there was a decrease in the percentage of walking or cycling for both long and distance commuters (short distance 1% decrease; long distance 23% decrease)

Reasons for changing usual method of commuting

7.40. The availability, cost and convenience of transport alternatives will clearly be related to home and workplace location. It might be expected that individuals who make changes in their home and/or workplace location will be more likely to change commuting mode and distance than those individuals who remain in the same homes and job. In the SHS those who had changed their usual method of travel to work were asked why.

7.41. More than one reason could be provided. Figure 7.15 shows the reasons for changing mode. The most frequently cited reasons were changing job (49% N=779) and moving home (23% N=779). There were no significant differences in reasons for changing mode between long and short distance commuters.

Figure 7.1: Average commuting distance by gender and age ( SHS February 1999 to March 2005)

Figure 7.2: Average commuting distance by gender and income ( SHS February 1999 to March 2005)

Figure 7.3: Average commuting distance by gender and urban/rural classification ( SHS February 1999 to March 2005)

Figure 7.4: Average commuting distance by gender and social class groupings ( SHS February 1999 to March 2005)

Figure 7.5: Average commuting distance by gender and household composition ( SHS February 1999 to March 2005)

Figure 7.6: Average commuting distance by household composition and income ( SHS February 1999 to March 2005)

Figure 7.7: Average commuting distance and single occupancy households (car availability and gender) ( SHS February 1999 to March 2005)

Figure 7.8: Usual method of travel to work for short and long distance commuters ( SHS February 1999 to March 2005)

Figure 7.9: Reasons for current mode choice for long and short distance commuters ( SHS February 1999 to March 2005)

Figure 7.10: Car commuters' reasons for not using public transport ( SHS February 1999 to March 2005)

Figure 7.11: Commuters who changed their usual method of travel to work ('mode one year ago' and 'current mode') ( SHS April 2003 to March 2005)

Figure 7.12: Commuters who changed their usual method of travel to work and percentage change in mode share ('mode one year ago' and 'current mode') ( SHS April 2003 to March 2005)

Figure 7.13: Short distance commuters who changed mode: modes before and after change ( SHS April 2003 to March 2005)*

Figure 7.14: Long distance commuters who changed mode: modes before and after change. ( SHS April 2003 to March 2005)*

*Figures do not add up to 100% as 'other' modes are not shown

Figure 7.15: Reasons for changing mode (long and short distance commuters) ( SHS April 2003 to March 2005)

« Previous | Contents | Next »