Transportation

, Volume 39, Issue 1, pp 175–195

Weekly activity-travel behaviour in rural Northern Ireland: differences by context and socio-demographic

Authors

  • Julian Hine
    • Built Environment Research InstituteUniversity of Ulster
    • School of Urban DevelopmentQueensland University of Technology
  • Neale Blair
    • Built Environment Research InstituteUniversity of Ulster
Article

DOI: 10.1007/s11116-011-9322-4

Cite this article as:
Hine, J., Kamruzzaman, M. & Blair, N. Transportation (2012) 39: 175. doi:10.1007/s11116-011-9322-4

Abstract

Despite a wide variation in access to goods and services between rural areas, common policy interventions are often proposed in Northern Ireland. Questions remain as to the level and form of policy differentiation that is required, if any, both within and between different rural areas. This issue is investigated in this paper through the analysis of activity-travel patterns of individuals living in two rural areas with different levels of area accessibility and area mobility. Three focus groups, 299 questionnaires and 89 activity-travel diaries for 7 days were collected for individuals from these areas. Regression analyses were employed to explore the degree to which different factors influence activity-travel behaviour. The results indicate that individuals from rural areas with a higher level of accessibility are more integrated within their local community and as a result, are potentially less at risk of being excluded from society due to immobility. Differences, however, were also found between different groups within an area (e.g. non-car owning individuals who were more reliant on walking, and low-income individuals who made trips of a shorter distance). Based on the study findings and a review of existing policies, this research highlights the need to tailor policy responses to reflect the particular sets of circumstances exhibited in different areas.

Keywords

Travel behaviourTransport disadvantageRural formSocial exclusionNorthern Ireland

Introduction

Numerous studies have focused on the characteristics of the built environment in urban areas (commonly referred to as urban form) and in turn how these characteristics influence travel behaviour (Coevering and Schwanen 2006). In comparison, very little has been done to extend this concept to the study of rural form and its links with travel behaviour despite the heterogeneity found in rural areas (Gray 2000). Research on the contribution of urban form characteristics have been differentiated by subjective judgements concerning the availability of opportunities in different statistical divisions of a city such as the CBD (Buliung and Kanaroglou 2006). Other studies have objectively defined urban form e.g. using a zonal population density measure (Morency et al. 2011). This approach has been extended to the study of activity-travel patterns between rural and urban areas (Millward and Spinney 2011).

A wide range of indicators representing travel behaviour have been employed including number of trips (Kerr et al. 2007), the size of activity spaces (Buliung et al. 2008), travel distance (Morency et al. 2011), activity duration (Kamruzzaman et al. 2011; Kang and Scott 2010), satisfaction with public transportation (Ji and Gao 2010), and trip chains (Timmermans et al. 2003). Yet relatively little or no effort has been made to use this work to inform the development of inclusionary transport policy even though the reduction of social exclusion has been an integral part of some transport policies. As a result, the central question that this paper seeks to answer is: do activity-travel patterns differ between different types of rural areas and if so how does this relate to the development of transport policies in practice?

Social exclusion, a geographically relative concept, is generally agreed to be a dynamic process (e.g. societal systems, agencies) that leads to deprivations across multiple dimensions in life (e.g. income, employment—intermediate outcomes) at a certain point in time which together or individually prevent participation in all types of activities (ultimate outcome) (Burchardt et al. 1999; Higgs and White 2000). Transportation has been identified as an important policy area which influences social exclusion because it enables people to travel and to participate in activities (Social Exclusion Unit 2003). Social exclusion and transport disadvantage are linked through the concept of (in) accessibility. Therefore, transport disadvantage is a function of a lack of access to both transport and opportunities i.e. discrete spatial features where activities could take place e.g. buildings (Stanley and Lucas 2008; Stanley and Stanley 2004).

Studies have shown that the nature of transport disadvantage vis-à-vis social exclusion varies between and within rural areas primarily due to the differential level of access to both transport and opportunities (Gray 2000). However, transport policy interventions have treated all rural areas in similar ways irrespective of their spatial settings (rural form) (Banister 2008). Although disaggregated measures are highly desirable most of the previous research studies in the context of rural areas have used spatially aggregated accessibility measures to identify transport disadvantage (Higgs and White 2000). In the UK, policy has also been driven by the accessibility planning approach. This uses highly aggregate spatial data, and as a result, is not suitable for identifying the impacts of transport policies upon transport disadvantaged groups (Preston and Rajé 2007). Using disaggregated census data, Nutley (2003) did not find any consistent relationship between different explanatory factors and transport disadvantage in rural Australia. This study, therefore, called for a case study approach to identify transport disadvantage locally. Subsequently, Nutley (2005) collected data from two case study areas from rural Northern Ireland (NI) and found variations in activity-travel patterns both within and between the cases. Questions therefore remain as to the form of policy differentiation that is required, if any, both within and between rural areas to reduce transport related social exclusion. This is due to the fact that both accessibility and mobility are a relative concept and can be differentiated amongst individuals both spatially and temporally (Farrington 2007). This clearly suggests that the analysis of disaggregated data is needed to assist in the identification and reduction of transport related social exclusion (Department for Transport 2006).

This paper firstly, reviews existing rural transport policies for NI; and secondly, evaluates the efficacy of these policies in reducing transport related social exclusion in different rural contexts through an analysis of disaggregated activity-travel data. The paper goes on to discuss the methods employed in the analysis of activity-travel patterns of individuals by selecting two case study areas with differential levels of area accessibility and area mobility. The findings from this work are then used to assess the effectiveness of transport policy initiatives in improving accessibility and mobility.

Rural transport policy in Northern Ireland

Traditionally, policy making in Northern Ireland has followed approaches taken at the UK level (Nutley 2005). Targeting Social Need (TSN), the earliest policy initiative (1991) in NI, aimed at reducing social exclusion and targeting deprived neighbourhoods based on a multiple deprivation index (MDI) although transport was initially not considered under this programme (Research and Library Services 2001). Due to a lack of accountability and a proper implementation mechanism associated with this initiative, a new policy initiative ‘New Targeting Social Need (New TSN)’ was launched in 1998 as an overarching policy to address social exclusion. Transport was identified as a priority area in this strategy. An updated version of the MDI was developed in July 2001 incorporating a new dimension of deprivation based on ‘proximity to services’ which measures the shortest path distance to essential opportunities (e.g. hospital) from each of the administrative units (NISRA 2005a; Research and Library Services 2002). Using this measure, the most deprived areas have been found to be located in rural areas (EAFRD 2007). As a result, the Department for Regional Development (DRD) has been providing transport support in rural areas for people with reduced mobility options through the Rural Transport Fund (RTF) (e.g. subsidy for new rural Ulsberbus services, financial support for community transport) (Department for Regional Development 2008a). Parallel to this initiative, emphasis has been placed on both ‘transport’ and ‘opportunities’ in the Regional Development Strategy (RDS) for NI 2001 (Department for Regional Development 2001) and Regional Transport Strategy (RTS) (Department for Regional Development 2002). In order to provide goods and services accessible to rural communities, the RDS has proposed building up to 40,000 dwellings to facilitate services and amenities (i.e. opportunities) in the main towns (collectively known as hubs). These hubs are connected by transport networks to serve the towns as well as their rural hinterlands by public transport services (Fig. 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs11116-011-9322-4/MediaObjects/11116_2011_9322_Fig1_HTML.gif
Fig. 1

Spatial development strategy for Northern Ireland (adapted from Department for Regional Development 2001)

The integration between planning and transport policy has been enhanced by the publication of the Planning Policy Statement 13 (PPS13) in 2005 (Department for Regional Development 2005b). The PPS13 embodies 12 general principles, principle 10 applies to rural areas which states that ‘rural public transport schemes should be developed to link rural dwellers to essential facilities and larger settlements’. The issue of access to transport for the more vulnerable groups (e.g. disabled) has been taken into account in the Accessible Transport Strategy (ATS) (Department for Regional Development 2005a). The ATS aims to overcome physical, financial, and temporal barriers that impede access to the transport system for older people and people with disabilities through: redesigning bus stops, restructuring the concessionary fares scheme, introducing a new service standard (e.g. three return journey per day for villages and intermediate settlements) and introducing innovative public transport schemes (e.g. demand responsive transport—DRT). Although community transport services are currently operational in rural areas where mainstream public transport services are not geographically accessible, a DRT scheme has yet to be operationalised in rural areas despite this policy emphasis (Department for Regional Development 2005a).

In summary the main policy responses to the transport needs of rural communities in Northern Ireland are mobility orientated (e.g. concessionary fares, community transport services, 3 return journey policy, need for DRT in rural areas), very little emphasis, however, is placed on the need to provide proximate opportunities to rural dwellers to address their travel needs. These policies are aimed at the provision of these transport options thereby allowing rural dwellers to access goods and services located at the hubs where opportunities have traditionally been located.

Methodology

Selection of rural case study areas

All settlements in NI are classified into eight classification bands (A–H) based on population1 (NISRA 2005b). Band A to E are defined as urban whereas from band F to H are defined as rural. This classification of settlements was utilised in this research. For the purpose of this study, a case–control study was designed in order to assess the effectiveness of the policies in reducing transport related social exclusion. Spatial analyses were conducted and two case study areas, Moira and Doagh, were selected using criteria related to the differential levels of area accessibility (self-containment in terms of locally available goods and services, and proximity to urban areas) and area mobility (proximity to the motorway and train stations) options (Fig. 2).
https://static-content.springer.com/image/art%3A10.1007%2Fs11116-011-9322-4/MediaObjects/11116_2011_9322_Fig2_HTML.gif
Fig. 2

Criteria for the selection of case study areas (ad) and the selected case study areas with differential level of area accessibility and area mobility options (ef)

According to the rural–urban classification of settlements, Moira is classified as an intermediate settlement with a population size of 3,682 whereas Doagh is classified as a village comprising of 1,130 population (Northern Ireland Neighbourhood Information Service 2007). The calculated sizes of locally available opportunities based on non-residential building footprints were found to be 31657 and 4935 m2 in Moira and Doagh, respectively. The Northern Ireland Neighbourhood Information Service (2007) has reported that Doagh lacks retail, public administration, health, education, employment, and service centres locally. The motorway and train station are located around 10 km away from Doagh whereas these are located within walking distance from Moira (Fig. 2). Due to the variation in the nature of public transport services in rural areas (in terms of service frequency, spatial and temporal coverage) it was anticipated that areas close to the motorway were more likely to have good public transport services. Doagh is therefore a representative case associated with the policy objectives in NI that focus on the provision of goods and services in ‘hubs’ which will then in turn be connected by a limited number of rural transport services per day whereas Moira represents an ideal situation and was used as a control. The selected case study areas also meet the different criteria that have been used in the literature to differentiate types of urban form as discussed in “Introduction” section.

Data collection

Data was collected from respondents residing within the two selected case study areas in three phases. In the first phase, three focus groups were conducted (two in Doagh and one in Moira) in order to identify the problems associated with accessing transport and/or land use systems as well as to explore the reasons for choosing the selected areas as a residential location. The key issues identified in this phase were subsequently used to design a questionnaire which was operationalised in the second phase. The questionnaire survey, therefore, provided an opportunity to triangulate the findings from the focus groups in a quantitative way using a larger set of responses. Individuals’ socio-economic data were also collected through the questionnaire survey to use as explanatory variables in this research when combined with the later diary phase as shown in Table 1. In the questionnaire survey, respondents were also asked whether they would like to participate in an activity-travel diary survey in the third phase.
Table 1

Variables used in the empirical modelling and their definitions

Variable names

Coded categories and definition

Variable used as in the model

Data collected through

Area profile

1 = Moira (good area accessibility and area mobility options); 2 = Doagh (poor area accessibility and area mobility options)

Explanatory

Questionnaire survey

Gender

1 = Male; 2 = female

Explanatory

Questionnaire survey

Car-ownership

1 = No (no-car in household); 2 = Yes (one or more cars in household)

Explanatory

Questionnaire survey

Income

1 = Low-income (income level below the average income of rural NI); 2 = high-income (income level above the average income of rural NI)

Explanatory

Questionnaire survey

Age

1 = Young (18–59 years); 2 = older (60 years and above)

Explanatory

Questionnaire survey

Occupation

1 = Working (full/part time employed, business); 2 = non-working (retired, unemployed, household management, student)

Explanatory

Questionnaire survey

Home-ownership

1 = Household owning a house; 2 = otherwise

Explanatory

Questionnaire survey

Trip length

1 = Less than 2 km; 2 = between 2 km and 5 km; 3 = between 5 km and 10 km; 4 = more than 10 km

Explanatory

Activity-travel survey

Transport mode

1 = Driving a car; 2 = bus; 3 = lift; 4 = walk; 5 = taxi; 6 = bicycle

Dependent/(explanatory)

Activity-travel survey

Trip purpose

1 = Work; 2 = social; 3 = recreational; 4 = shopping; 5 = taking a meal; 6 = other; 7 = health

Dependent/(explanatory)

Activity-travel survey

Trip time in a day

1 = Mid-day (10:00–16:00); 2 = Morning peak (8:00–10:00); 3 = Morning (0:00–8:00); 4 = Afternoon peak (16:00–18:00); 5 = Evening (18:00–24:00)

Dependent/(explanatory)

Activity-travel survey

Trip day in a week

1 = Weekdays (Monday–Friday); 2 = Weekends (Saturday–Sunday)

Dependent/(explanatory)

Activity-travel survey

Trip destination

1 = Local trips (destination within 2 km from the home); 2 = Surrounding area trips (dest. between 2 km and 5 km from home); 3 = Medium range trips (dest. between 5 km and 10 km from home); 4 = Longer range trips (destination more than 10 km away from home)

Dependent

Activity-travel survey

Trip length

Continuous data type: geographic distance travelled in each trip

Dependent

Activity-travel survey

Activity duration

Continuous data type: time spent on undertaking different out of home activities associated with each trip

Dependent

Activity-travel survey

The respondents for the questionnaire survey were recruited via face to face interview in local neighbourhoods. A pre-designed questionnaire form with a postage paid return envelope was provided to those who consented to participate in the questionnaire survey and a total of 299 questionnaires were collected. The required sample sizes were determined and 153 questionnaires from Doagh and 146 questionnaires from Moira were collected (Cochran 1963). Amongst those respondents who completed the questionnaires, 85 and 96 individuals provided consent for the activity-travel dairy survey from Moira and Doagh, respectively. A 7 day activity-travel diary form was designed and delivered to these individuals with a postage paid return envelope. However, 89 individuals from the selected two case study areas (50 diaries from Doagh and 39 diaries from Moira) completed the survey (average returned rate 49%). Information about each trip such as trip day, trip origin, trip start time, trip destination, trip end time, trip purpose, travel mode, and roads names/routes travelled were collected. A total of 1,821 individual trips were reported by respondents in the 7 day survey. The socio-economic breakdown of the respondents in both the questionnaire and activity-travel survey is representative of the population found in these areas based on the 2001 census data (Table 2) (NISRA 2001). A rural–urban breakdown from NI Travel Survey data was obtained which showed a close match between the results presented in this research and the Travel Survey data across several categories (e.g. trip length, modal share, activity patterns) (Department for Regional Development 2008b).
Table 2

Correlation coefficients between the area profile variable and the other explanatory variables

Variables

Class

Distribution of samples (%)

Correlation coefficients between explanatory variables

Survey samples

2001 census

Questionnaire data

Activity-travel diary data

Moira

Doagh

Moira

Doagh

Area profile

Area profile

Gender

Male

42.5

36.6

48.8

48.8

−0.06

0.12

Female

57.5

63.4

51.2

51.2

  

Car-ownership

No

8.9

19.6

9.6

20.9

0.15a

−0.05

Yes

91.1

80.4

90.4

79.1

  

Income

Low

50.7

58.2

0.08

−0.17

High

49.3

41.8

  

Age

Young

74.0

64.1

−0.11

0.01

Older

26.0

35.9

  

Occupation

Working

72.6

54.2

−0.20a

−0.03

Non-working

27.4

45.8

  

Home-ownership

Owner

80.8

74.5

86.8

78.3

−0.08

0.20

Otherwise

19.2

25.5

13.2

21.7

  

aCoefficients are significant at the 0.05 level (2-tailed); and unmatched categories are not reported

Data processing

Using the activity-travel diaries, modal split, trip purpose, trip length, trip time in a day, trip day in a week, trip destination, and activity duration were derived and used as dependent variables in this research (Table 1). Previous research studies have identified a number of factors that potentially influence activity-travel behaviour including the characteristics of travellers (socio-economic variables e.g. age, sex, car ownership) (Kang and Scott 2010; Xing et al. 2010), contextual variations due to geographical heterogeneity (Páez 2006), and characteristics of the journey itself (e.g. trip purpose, travel distance, time of the day when the journey is made, and travel day in a week) (Buliung et al. 2008). As a result, and in addition to using respondents’ socio-economic and the area profile attributes, the derived transport mode, trip purpose, trip length, trip time in a day, and trip day in a week dependent variables were also used as explanatory variables in this paper depending on the chosen dependent variable for analysis.

Trip origin, trip destination, and travel roads/routes of the individual trips were geo-referenced and the lengths of these trips were derived using ArcGIS software. Trip length was used as a continuous variable when this was analysed as a dependent variable. However, when the trip length variable was used as an explanatory variable for analysis, it was categorised into four classes in order to fit with the other explanatory variables as shown in Table 1. Activity duration was calculated by subtracting trip end time of a trip from the trip start time of the subsequent trip of the chained trips. A ‘chained trip’ is referred to as at least two consecutive trips within a day. Time spent at home and overnight stays at activity locations for other purposes (e.g. social) were not considered as an activity duration. As a result, a total of 1,002 individual trips were considered in this analysis. The geo-referenced destinations were used to derive the spatial distribution of trips from the case study areas and were classified as local trips, surrounding area trips, medium range trips, and longer range trips (Table 1). These classifications were made in this way so that they matched with the spatial form of the case study areas (Fig. 2). Trips that were defined as home destinations were excluded from this analysis. As a result, a total of 1,014 individual trips were analysed that finished at locations other than homes (807 return home trips).2 The temporal distributions of trips were investigated by classifying individual trips into weekdays and weekends, and also between different times of day as shown in Table 1.

Data analyses

Researchers have analysed factors which can influence travel behaviour using parametric tests such as the multinomial logistic model, binary logistic model (Páez 2006; Xing et al. 2010). In this research the binary logistic regression model was used due to its computational interpretability and statistical goodness of fit (Eq. 1) (Kerr et al. 2007; Morency et al. 2011).
$$ ln(odds) = logit (P) = ln\left( {{\frac{P}{1 - P}}} \right) = b_{0} + b_{1} X_{1} + b_{2} X_{2} + b_{3} X_{3} + \cdots + b_{k} X_{k} $$
(1)
where, P is the predicted probability that an individual experiences the event of interest (e.g. use a car as a transport mode) given his/her set of scores on the explanatory variables X where there are k explanatory variables, b0 is the constant of the equation, 1 − P is the predicted probability of the other decision (e.g. use of alternative modes). As a result, the categorical dependent variables were coded into a binary form in order to fit with the model (Rose and Marfurt 2007). The binary logistic regression model computed the odds ratios (ORs) for each explanatory variable that indicated a measure of how much more likely one group (e.g. male) performed in one category (e.g. bus) over all other categories when compared to its counterpart (e.g. female), controlling for other variables in the model. All the explanatory variables were entered into the model using the block entry method. In addition, linear multiple regression analyses were conducted to investigate the influence of continuous dependent variables.

A correlation analysis was conducted to identify whether the collected data was significantly biased in any of the areas. Table 2 shows that no strong association exists between the socio-economic status of the individuals and the two case study areas. This indicates that the two samples (Doagh and Moira) are quite closely matched and that the results presenting comparisons between the case study areas in this research possesses the quality of a case–control study (Ornetzeder et al. 2008); and it was expected that any differences in activity-travel patterns that might exist are likely to be explained by differences in the accessibility and mobility options available in the respective case study areas. Although the self-selection issue has been viewed as inappropriate in the social equity literature which argues that a basic level of services is a merit good and should be available in all areas (Loader and Stanley 2009), the differences that could be found between the areas are not due to residential self-selection bias in this research (Handy 2005; Scheiner 2010). Analysis of focus group data shows that participants living in both case study areas had a higher level of expectation with regard to transport and opportunities in these areas; and as result, when their expectations were not met; they raised concerns about the existing transport and/or land use arrangements in these areas.

Descriptive statistics

A number of issues associated with using public transport services were identified in the focus group. These were also triangulated in the questionnaire survey in which 38% respondents indicated that temporal inaccessibility of public transport was their main concern followed by a lack of geographical coverage by these services in terms of the ability to reach their destinations (34%). Longer travel times (29%), due to frequent stopping in the peak hours, and a high rate of fare (25%) were also found to be a major concern with public transport services in both areas. 73% respondents felt no opportunity related problem existed in Moira, in Doagh this was lower at 14%. In terms of the availability of public transport services, 41% of respondents indicated a problem in Doagh compared to 16% in Moira. Traffic congestion in the morning was found to be a common problem in all areas.

92% of respondents, whose occupational status was described as working, indicated that they took a fixed route to and from work and no difference was found between the case study areas relating to this behaviour. Around half of the respondents also indicated that they thought of taking alternative routes mainly due to congestion on their regular travel routes. A significant difference was however found to exist in the patterns of grocery shopping between the two case study areas. Whereas around 89% of the respondents living in Moira had a fixed store location for grocery shopping, this was reduced to 81% for the respondents living in Doagh. Overall 85% respondents in rural areas were found to have a fixed store location for grocery shopping.

Regression analyses results

Activity patterns

The key outcome of social exclusion is a lack of participation in activities. In transport terms, such participation implies the number of journeys made in order to access goods and services. The only significant difference in activity patterns between the two case study areas was for health trips. Table 3 shows that respondents in Doagh were three times less likely to make a health trip compared to individuals who live in Moira. This does not necessarily mean that individuals living in Doagh experienced better health. The health deprivation scores of the NI Multiple Deprivation Measures show that individuals living in Doagh had a poorer health condition compared to individuals living in Moira (NISRA 2005a). This result, therefore, can be explained by the lack of health related facilities in Doagh. Table 3 shows that non-car owning individuals and individuals who lived in the rented housing sector made a significantly lower number of recreational trips, due to mobility and financial constraints. Non-working individuals though were found to be twice as likely to make trips for social and dining out activities. This behaviour can be explained by the fact that working individuals, who experience ‘time poverty’ and spend most of their time at work and as a result have little or no time to participate in other types of activity.
Table 3

ORs associated with different explanatory variables for undertaking different activities

Explanatory variables

Dependent variables: trip purposes

Work

Social

Recreation

Shopping

Taking a meal

Other

Health

Area profile (ref: Moira vs. Doagh)

1.208

1.176

0.854

0.823

0.767

2.210a

0.383a

Gender (ref: male vs. female)

0.817

0.837

0.720

0.913

0.549

2.727a

0.949

Car-ownership (ref: no vs. yes)

2.582

0.698

3.958a

1.253

0.798

0.133a

0.287

Income (ref: low vs. high)

0.732

0.792

1.311

0.974

0.447

1.584a

1.800

Age (ref: young vs. older)

0.628

1.174

0.991

1.346

0.535

1.270

1.523

Occupation (ref: working vs. Non-work)

2.031a

1.513

1.377

2.327

1.387

0.735

Home-ownership (ref: owner vs. rented)

1.244

1.013

0.463a

1.082

1.200

1.116

1.012

Mode: car (reference)

 Bus

8.604a

0.536

2.267

1.642

0.106a

1.604

 Lift

0.684

0.905

2.564

1.123

0.987

0.026a

0.948

 Walk

0.800

0.829

15.406a

0.631

0.473

0.181a

0.519

 Taxi

1.596

0.610

1.564

4.023

 Bicycle

2.870

0.830

12.088a

Trip length: Less than 2 km (reference)

 2 km–5 km

0.844

0.813

14.251a

0.378a

0.068a

0.761

0.380

 5 km–10 km

1.304

0.754

29.414a

0.252a

0.604

0.355a

0.261

 More than 10 km

2.927

0.864

6.882a

0.442a

0.137a

0.276a

0.595

Trip time: Mid-day (10:00–16:00) (ref)

 Morning peak (8:00–10:00)

10.445a

0.296a

.753

0.410a

0.749

0.512a

 Morning (00:00–8:00)

36.208a

.526

1.593

 Afternoon peak (16:00–18:00)

0.049a

1.387

1.177

0.468a

4.490a

1.613

2.635

 Evening (18:00–24:00)

0.863

2.164a

3.324a

0.363a

1.347

0.525

0.221a

 Trip day (ref: Weekday vs. Weekends)

0.025a

3.533a

1.999a

1.166

2.310a

0.343a

a Associated B coefficients are significant at the 0.05 level

The data also reveal a higher rate of bus use, compared to those for the car, for making work trips which indicates the potential of commuter services in enhancing the economic integration of rural people. The behavioural patterns of individuals as shown in Table 3 reveal that commuter services are needed on weekday mornings. Table 3 also shows that individuals were more likely to make social and recreational trips during the evening and on weekend periods whereas shopping trips were conducted mainly at mid-day throughout the week.

Modal split

Mode choice behaviour reflects the availability of alternative modes to different groups as well as the local availability of goods and services. A lack of mode choice options restricts the movement of transport disadvantaged groups who can become socially excluded as a result (Knowles 2006). In addition, Putnam (2000) has found negative links between car dependence and the development of effective social capital. Kawachi and Berkman (2000) indicated that high stocks of social capital also lead to socially inclusive societies. Evidence also suggests that individuals make more trips on foot where locally available goods and services are not limited, and individuals in these communities are less at risk of being excluded due to immobility. Kerr et al. (2007) have also shown that walking increases trust and social engagement and consequently increases the stock of social capital.

Table 4 shows that individuals in Doagh were around three times more likely to use a car when compared to individuals who live in Moira where individuals were more likely to walk. A higher level of area accessibility for individuals living in Moira has also possibly acted as a catalyst for building up trust and consequently social capital. The outcome of which is that these individuals were able to take advantage of such capital by taking more lifts which facilitated their participation in distant activities3 (Table 4). A significantly higher rate of bus use by older individuals clearly reflects the impact of the concessionary fare schemes for this group. Non-car owning individuals were found to make a significantly higher number of trips by bus, lifts in cars, and walking than their car-owning counterparts. Non-working individuals were also found to be more reliant on lifts. Table 4 also shows that taxi use increased when bus services were not available (e.g. evening). Focus group participants indicated that the taxi was costlier than the bus which meant that a lack of bus service put further financial pressure on already disadvantaged groups (e.g. non-car) to perform their required activities.
Table 4

ORs associated with different explanatory variables and the modes of travel

Explanatory variables

Dependent variable: travel modes

Driving car

Bus

Lift

Walk

Taxi

Bicycle

Area profile (ref: Moira vs. Doagh)

2.976a

0.736

0.499a

0.397a

0.953

0.261

Gender (ref: male vs. female)

1.083

0.612

1.375

0.619

0.716

0.032a

Car-ownership (ref: no vs. yes)

0.001a

0.112a

0.158a

0.167

Income (ref: low vs. high)

1.040

0.798

1.180

0.942

3.153

0.493

Age (ref: young vs. older)

0.275a

5.238a

0.749

4.750a

4.397

7.487

Occupation (ref: working vs. non-work)

0.960

0.370

2.262a

0.613

2.251

0.376

Home-ownership (ref: owner vs. rented)

1.330

0.216a

0.472a

1.409

0.678

0.089

Trip purpose: Work (reference)

    

No trip

 

 Social

0.839

0.465

1.137

1.011

1.098

 Recreation

0.194a

4.008

1.664

13.509a

5.270

 Shopping

0.935

9.568a

1.154

0.458

 Taking a meal

1.716

0.697

0.735

 Other

3.347

3.973

0.140

0.143a

 Health

1.362

4.729

1.162

0.172

Trip length: Less than 2 km (reference)

 

No trip

  

No trip

 

 2–5 km

61.194a

3.232a

0.003a

1.022

 5–10 km

56.584a

6.137a

0.001a

0.000

 More than 10 km

41.689a

4.246a

0.661

Trip time: Mid-day (10:00–16:00) (ref)

 Morning peak (8:00–10:00)

0.819

6.233a

5.508a

 Morning (00:00–8:00)

1.031

2.899

82.300a

0.643

10.362

 Afternoon peak (16:00–18:00)

1.643

0.290

0.875

0.591

12.672a

5.001

 Evening (18:00–24:00)

1.335

0.271a

0.784

0.873

35.144a

4.303

 Trip day (ref: Weekday vs. Weekends)

0.692

0.405

1.454

1.717

0.538

0.796

aAssociated B coefficients are significant at the 0.05 level

Spatial distribution of trips

Putnam (2000) found a reduced level of civic engagement for individuals who spent more time travelling and who travelled longer distances daily. A consequence of this is that highly paid commuters can be spatially excluded from their local neighbourhood precisely because of their high mobility (Cass et al. 2005). This implies that an analysis of the spatial dimension of activity participation can complement the assessment of transport related social exclusion. Cass et al. (2005) have mentioned that thinking about the spatial and mobility related aspects of social exclusion is important.

In this study a significant difference was found to exist in the spatial distribution of trips between the case study areas (Fig. 3). Individuals from Moira were found to be thirteen times more likely to make a trip locally than their counterparts in Doagh. Whereas respondents from Doagh were found to be about seven times more likely to make trips to surrounding areas than respondents living in Moira. No significant difference exists in the number of trips that were completed at a destination located between 5 and 10 km from the case study areas, although respondents from Moira were found to be more likely to make trips that finished further away. These differences are due to the fact that most of the essential opportunities are located within Moira whereas for Doagh essential opportunities are located between 2 and 5 km away from the village (Fig. 2). Individuals from both areas travelled to urban areas for accessing higher order goods and services which are located between 5 and 10 km away from the two areas; and as a result no significant difference was observed within this range. The data clearly indicates that the location of available opportunities dictate the spatial distribution of trips.
https://static-content.springer.com/image/art%3A10.1007%2Fs11116-011-9322-4/MediaObjects/11116_2011_9322_Fig3_HTML.gif
Fig. 3

Spatial distribution of trip destinations from the case study areas

The mobility constraints of females, non-car groups, and non-working groups were reflected in their behaviour as they made a significantly higher number of trips to local and surrounding areas. This was primarily to participate in all types of activities other than work. Behavioural patterns also show that local opportunities were accessed mostly on foot and equally at different times in a day and also on different days in a week. This signifies the importance of local opportunities in enhancing social inclusion for disadvantaged groups.

Temporal distribution of trips

Due to the variation in bus schedules and the opening hours of opportunities (activities) between weekdays and weekends, and between peak hours and non-peak hours, it is important to examine whether the temporal distribution of trips differs significantly in these periods between different groups and between the two case study areas. Previous research studies have shown that due to a lack of public transport services at different time periods (e.g. weekends, evening), the mobility of transport disadvantaged groups is highly constrained (Wu and Hine 2003). In the case study areas analysis shows that a significant difference exists between them in terms of accessing goods and services at different times of day. Individuals in Doagh were found to be more likely to make trips at mid-day and less likely to make trips in the evening compared to individuals living in Moira. These differences can be explained by the availability of public transport services. In both areas no bus service was available after 6:30 pm. As a consequence, the non-car owning group from Doagh had to finish their trips before this time, whereas in Moira the non-car owning group were able to participate in activities late at night because opportunities were also located within walking distance.

The trip making behaviour of females was not only found to be restricted spatially as discussed earlier, it was also found to be restricted temporally particularly in the evening. Safety, scarcity of local opportunities, difficulties associated with organising out of home and in-home activities together with the availability of transport probably dictated this behaviour. Non-car owning individuals were more likely to make trips in the early morning than their car-owning counterparts, because non-car owning individuals travelled by bus and as a result had longer travel times ("Descriptive statistics" section), and as a result had to leave their homes earlier in order to reach their activity locations.

Trip distance

Trip distance by activity type

A multiple regression analysis was conducted using the distances associated with the 1,821 individual trips as a dependent variable (Table 5). On average people from Doagh travelled marginally longer distances per trip (13.3 km) than the people from Moira (12.9 km). Trip distances were found to be significantly longer for males, older people, and for those in work, those with a car, and those with a higher level of income (Table 5). However, the beta coefficients associated with these explanatory variables reveal that occupation has a larger effect on trip distance followed by income. The only significant difference in trip distance between the case study areas was for the purpose of work (Table 5). Analysis shows that individuals from Moira travelled significantly longer distances for work (18.3 km) than individuals from Moira (15.9 km). Income was found to have a larger impact on trip distances associated with work; with higher income individuals making significantly longer work trips. Working individuals made longer distance trips for the purpose of recreation, whereas car owners and those in the owner occupied housing sector were found to make longer distance shopping trips.
Table 5

Multiple regression analyses results showing the impacts of the explanatory variable on trip distance

Explanatory variables

 

All

Trip purpose

Transport mode

Work

Social

Recreation

Shopping

Food

Health

Car

Bus

Lift

Walk

Taxi

Bicycle

Area profile

t

0.01

−2.29a

1.38

−0.60

1.48

0.22

0.92

−1.56

−2.49a

−1.03

0.22

−0.11

−3.38a

 

Beta

0.00

0.15

−0.10

0.06

−0.10

−0.09

−0.18

.043

0.44

0.15

−0.02

0.03

0.83

Gender

t

−3.27a

−0.12

−1.74

−0.08

0.21

0.46

0.49

−2.94a

−1.57

0.26

−0.37

0.06

0.03

 

Beta

−0.08

−0.01

−0.12

−0.01

0.02

0.14

0.09

−0.08

−0.18

0.03

−0.03

0.01

0.01

Car-ownership

t

2.87a

−1.21

1.92

−0.93

2.44a

0.00

1.05

−2.25a

1.97

2.42a

 

Beta

0.07

−0.09

0.15

−0.08

0.18

0.00

0.29

−0.42

0.30

0.23

Income

t

3.72a

3.21a

1.66

−0.27

−0.21

0.34

0.21

3.75a

1.97

−0.95

−0.65

6.68a

0.69

 

Beta

0.10

0.23

0.14

−0.03

−0.02

0.18

0.05

0.11

0.36

−0.14

−0.06

1.03

0.18

Age

t

2.66a

0.02

0.94

1.60

0.45

0.15

−1.29

2.86a

−2.13a

1.34

−0.85

1.31

−0.00

 

Beta

0.08

0.00

0.09

0.17

0.04

0.13

−0.29

0.11

−0.32

0.20

−0.07

0.26

0.00

Occupation

t

−5.27a

−1.06

0.09

−3.32a

0.08

0.55

0.97

−4.82a

−1.43

2.66a

0.59

 

Beta

−0.16

−0.07

0.01

−0.37

0.01

0.43

0.27

−0.18

−0.24

0.24

0.14

Home-ownership

t

−1.01

0.26

1.56

−0.67

−2.32a

0.09

−0.79

−0.28

−1.04

−0.09

−0.20

0.21

 

Beta

−0.03

0.02

0.12

−0.06

−0.17

0.05

−0.19

−0.01

−0.16

−0.02

−0.02

0.05

F-coefficient

 

13.09a

2.06a

2.97a

2.25a

3.08a

0.60

1.03

7.76a

1.98

2.94a

2.23a

8.86a

10.44a

a Coefficients are significant at the 0.05 level

Trip distance by transport mode

For those using a car, a significant number of shorter distance trips were made by those respondents who were female, young in age, had non-working occupational status, and had a lower level of income (Table 5). Table 5 shows that respondents from Moira made a significant number of longer distance trips by bus. Older respondents made a significantly higher number of trips by bus, although the trip distances using the bus were found to be significantly shorter than in the case of bus trips made by those in younger age groups (Table 5). Car-owning individuals not only made significantly fewer trips by bus but their trip distances using the bus were also found to be significantly shorter. Car owners, although making fewer walking journeys were found to make trips on foot which were significantly longer than their non-car owning counterparts. Although an equal number of trips were made by bicycle, the bicycle trip lengths of individuals living in Moira were found to be significantly longer than individuals who live in Doagh (Table 5). Further analysis indicates that most of the bicycle trips in this case were associated with recreational trips.

Travel time and activity duration

Measuring participation in activities by only counting the number of trips made to reach different activity locations does not indicate the magnitude of participation in these activities, and as a result, researchers have used activity duration as an indicator of the level of engagement in these activities (Burchardt et al. 1999; Kamruzzaman et al. 2011). A correlation analysis was conducted using travel time and trip distance from the 1,821 recorded trips and a significant correlation was found to exist between these two measurements. As a result, travel time was excluded from the analysis. Previous studies have indicated that a significant correlation exists between travel time and activity duration although such a correlation was not confirmed in this study (Dijst and Vidaković 2000). As a result, a multiple regression analysis was conducted using activity duration as a dependent variable to investigate the impacts of different explanatory factors on this variable.

Average activity durations were found to vary significantly between the areas; for instance, individuals from Moira spent significantly more time (189 min) undertaking a single activity compared to their counterparts in Doagh (154 min). Average activity durations were found to be higher for the purpose of trips to work (412 min) and lower for undertaking other types of activity (28 min) amongst all individuals. No significant differences in activity duration were found to exist between those in work and those not in employments for the remaining activity categories. Non-car owning individuals were found to spend significantly more time undertaking a work activity. It seems in this situation that restrictions on personal mobility meant that they spent longer periods of time at work.

Key findings and implications for policy

This paper has identified and explored the linkages between different factors which influence the differences in adult travel behaviour in rural areas. Two rural case study areas were identified from NI based on different area accessibility and area mobility criteria. The null hypothesis of this research was that no significant difference could be found in the activity-travel patterns of individuals living in these two areas; and that as a result, generalised inclusionary transport policies could effectively be applied to all rural areas. The findings of this research show that a significant difference exists in activity-travel behaviour patterns which can be grouped into contextual differences, and socio-economic differences.

Contextual differences

Results from the analysis show that individuals from both areas made an equal number of trips to undertake different types of activities. The travel patterns associated with undertaking these activities, however, were found to be significantly different between the areas. In Moira where more goods and services are available within the settlement, all individuals were found to make a significantly higher number of trips on foot. Research has shown that walking increases trust and social engagement (Kerr et al. 2007), and certainly the evidence from this work would suggest that individuals in Moira are more integrated in their local community (e.g. a higher number of trips by taking lifts). In Doagh, on the other hand, due to a lack of proximate opportunities, individuals made a significantly higher number of trips using the car and a significantly lower number of trips on foot. This higher level of car dependence can potentially impede the development of effective social capital and also result in residents becoming excluded from their local community. Temporal inaccessibility of public transport services was found to exacerbate the situation for non-car owning individuals living in Doagh. As a result, an implementation of the three return journeys policy for areas having similar profile like Doagh would further impose temporal limitation to access goods and services for non-car owning individuals. This is due to the fact that currently one bus per hour runs through Doagh due to its geographical location on the inter-urban bus route (e.g. Ballyclare-Belfast). A higher level of area accessibility combined with mobility means that significantly more time can be spent in activities. This means that individuals from this area (Moira) have potentially a greater opportunity to extend their participation and access to local goods and services. This finding has a serious policy implication for the transport disadvantaged groups living in areas with a poor level of accessibility and mobility.

Socio-economic differences

In addition to identifying differences between the case study areas, this research found a significant difference in activity-travel patterns between the different socio-economic groups living within the case study areas. Although both males and females made an equal number of trips using different transport modes and participated in different activities equally, the travel behaviour of females was found to be constrained both spatially and temporally. As discussed earlier in the research findings, females were found to be less likely to make trips that ended further away from their neighbourhood and which were in the evening (Scheiner 2010). This research found that non-car owning individuals made a significantly higher number of trips using the bus, taking lifts, and on foot than their car-owing counterparts. This group was also found to undertake less recreational activities. Analysis also showed that their odds of making local trips were increased. As a result, trip distances were found to be significantly shorter. Therefore, in an area with lower levels of accessibility to goods and services non-car owners are clearly at risk of not being able to participate fully in society due to their immobility.

A higher level of income and car ownership enabled individuals to make longer distance work trips. This means that these individuals were able to search for a job located further away. In comparison, lower income individuals, such as those living in the rented housing sector, and the unemployed with a lower level of car-ownership, and those living in areas with a low level of access to job opportunities will have fewer job opportunities due to this financial constraint. Respondents from both areas raised concerns, in focus groups, about the expense of using public transport services, although concessionary fares for older people were felt to stimulate their demand for public transport. It is therefore reasonable to assume that similar policy interventions if developed for the transport disadvantaged would assist them in accessing goods and services.

This research has utilised the development of disaggregated measures of travel behaviour to identify transport disadvantaged groups. The travel behaviours exhibited by the disadvantaged groups in this research (e.g. female, low-income, non-car owning, non-working) are in line with those found in other studies both in the context of NI and elsewhere (Department for Regional Development 2001; Department of Agriculture and Rural Development 2003; Hine and Mitchell 2001; Nutley 2005). These groups were found to have different travel experiences compared to the more affluent and mobile groups in the population. The findings also support the argument of previous research studies for disaggregate measures to identify transport disadvantage (Department for Transport 2006; Farrington 2007; Preston and Rajé 2007). The legitimacy of traditional zone based measures is also questionable given the variety of rural area contexts. As a result there is a need for the development of an approach which identifies the variety of transport experiences but also which at the same time reflects the different types of rural contexts in which this behaviour takes place and the need for non-standardised approaches to policy development.

Footnotes
1

These include: Band A—Belfast metropolitan urban area, Band B—Derry urban area, Band C—large town (population between 18,000 and 75,000), Band D—medium town (population between 10,000 and 18,000), Band E—small town (population between 4500 and 10000), Band F—intermediate settlement (population between 2,250 and 4,500), Band G—village (population between 1,000 and 2,250), and Band H—small village, hamlet, and open countryside (population less than 1,000.

 
2

Note that this classification is different from the classification of the route length variable. For this analysis, a categorisation of the trip destinations was made based on a network distance from the population weighted centroid of each case study areas and also based on trips of individuals from the respective case study areas. On the other hand, the route length classification was made irrespective of case study areas and also irrespective of the origin and destination of a trip.

 
3

Train journeys were not considered in this analysis due to the fact that the train was not accessible for the respondents living in Doagh. In addition, analysis shows that the train played a minor role in facilitating travel in rural areas. Only 0.7% of trips were made by train by the respondents from Moira.

 

Acknowledgment

The authors would like to thank the Editor-in-Chief (Martin G. Richards) of this journal and the three anonymous reviewers for their insightful comments. Base map: Land and Property Services, Permit No. 110017, © Crown copyright 2011.

Copyright information

© Springer Science+Business Media, LLC. 2011