Guidance

Journey time statistics, notes and definitions: 2019

Updated 19 December 2019

Introduction

This guidance provides information on methodology, data sources, and definitions of key terms used in calculating the Journey Time Statistics 2019 published by the Department for Transport (DfT).

The Journey Time Statistics series has previously been published for 2014 to 2017, and 2019, with detailed data tables available online.

Previous notes and definitions are available for the Journey Time Statistics series for 2015 to 2017, and also for the related Connectivity series.

Overview

Outputs

The journey time statistics relate to theoretical journey times to local key services in England, covering:

• employment centres (small, medium and large)
• primary schools
• secondary schools
• further education
• GPs
• hospitals
• food stores
• town centres

The statistics are published in terms of the following core metrics:

• average minimum journey time
• origin indicators
• destination indicators

Core metrics definitions

Average minimum travel time (minutes). The shortest travel to a given type of service by a particular mode of transport, averaged over an area.

Origin indicators. These indicators measure the number of different services in a particular area that users can reach within a given time.

Destination indicators. These indicators measure the proportion of users that can access a service within a certain time. User populations relate to each key service.

Output geography

Statistics are published at a range of geographies covering national, regional, local authority, and lower super output area level.

Geography definitions

Lower Layers Super Output Areas (LSOA).

LSOAs are small areas designed to be of a similar population size, with an average of approximately 1,500 residents of 650 households. There are 32,844 Lower-layer Super Output Areas (LSOAs) in England. They were determined by the Office for National Statistics for the reporting of small area statistics and are derived from the 2011 Census.

Urban and rural definitions.

This report uses the Defra Rural-Urban Classification, based on 2011 Census Output Areas. The Rural-Urban Classification defines areas as rural if they fall outside of settlements with more than 10,000 resident population. See Defra’s Definitions and Local Authority Classifications for more details.

Calculation

The statistics are generated from journey times calculated using commercial software for 171,372 English origins to each key service. The calculations use information on road and path networks, traffic speeds and public transport timetables to calculate journey times to selected services.

The statistics are designed to represent idealised journeys which are completed at the following times:

• a Tuesday in the second week October of the year to which they relate
• and during the ‘morning peak’ between 7am and 10am

Figure 1: Outline of journey time statistics calculation process

Figure 1 shows schematically how the journey time indicator calculation process operates. The top 10 quickest journey to each destination that can be completed by walking, cycling, car, and public transport are calculated, with each transport mode having a different set of assumptions about the journey.

All journeys start from one of the 171,372 2011 Census output areas in England, before going to the nearest point on the road network. For walk, cycle and public transport calculations, this is on the road and path network, whilst for car journeys this is only on the road network.

The following describes the next stage of the calculation, which is dependent on mode of travel:

• journeys relating to walk or cycle proceed to a destination point on the road and path network at speeds which are dependent on the type of road link
• car journeys continue along the road network to a destination point using speeds determined by real traffic data
• public transport journeys can be either direct walks, or involve public transport, however the ‘walk only’ option is limited to 2km for journeys by public transport

Journey times are then aggregated from output area level to various geographies with the journey times weighted by the relevant populations using a weighted average shown in Figure 2. The minimum journey time is multiplied by the service user population for each individual output area and this value is then divided by the service user population for the appropriate high-level geography.

Figure 2: Weighted equation used for aggregating geographies

For example, 3 Output Areas with minimum journey times of 10 minutes, 8 minutes, and 16 minutes respectively, in addition to populations of 100, 25, and 75 minutes respectively give a Lower Super Output Area with average minimum journey time of 12 minutes and population 200 (Figure 3).

Figure 3: Example use of weighted equation for aggregating geographies

See the Model Parameters section for further details on the assumptions and parameters used in the model.

Origins

All calculations use the 171,372 Output Areas (OAs) from the 2011 Census as the locations of populations accessing each service. OAs are the lowest geographical area in the 2011 Census data and are represented by population weighted centroids which use the spatial distribution of the population in output areas as recorded in 2011 Census to represent the population as a single point.

To provide the actual journey start point in each OA, the population weighted centroid of the OA was shifted to the nearest link (road) on the road network. In a very small number of cases an alternative nearby point on the road network was used, usually the nearest road node (junction), due to issues arising from THE normalising (straightening out) OF the road network in the calculation.

Origins methodological change for 2019

Calculations for Journey Time Statistics 2019 use the nearest link (road) on the road network to the population weighted centroid of the OA as the origin point of the journey. Previous years have used the nearest node (junction) on the road network to the population weighted centroid of the OA as the origin point of the journey to avoid issues arising from the normalising (straightening out) of the road network, however a solution have now been found for this. The change is likely to have increased the accuracy of journey times for individual journeys, but made little change to overall journey times.

Key services and users

The number of key services used in producing the indicators by year is shown in Table 1.

Table 1: Number of key services by year

Key service	2014	2015	2016	2017	2019
Employment centres (small)	16,465	16,625	16,930	17,194	17,457
Employment centres (medium)	9,235	9,460	9,707	10,241	10,545
Employment centres (large)	645	676	719	785	843
Primary schools	16,463	16,484	16,655	16,927	16,948
Secondary schools	3,365	3,376	3,381	3,174	3,128
Further education	2,624	2,606	2,418	2,304	2,198
GPs	9,257	11,167	9,128	7,353	6,866
Hospitals	296	278	278	277	219
Food stores	19,549	19,746	21,665	20,987	23,161
Town centres	1,211	1,211	1,211	1,211	1,211

Service users

Service users in the modelling are specific to target population of the service. For educational and employment services, users are considered to be the relevant populations. For other services, the number of households is used.

Service users for educational and employment destinations are from the ONS (Office for National Statistics) mid-2019 Output Area population estimates, with the following age bands used at output area level:

• 16 to 74 for employment centres
• 5 to 10 for primary schools
• 11 to 15 for secondary schools
• 16 to 19 for further education

For all other services (hospitals, GPs, foodstores, town centres), service users are the number of households in each output area has been estimated from 2011 Census data (downloaded via Nomis) and the 2018-based ONS household projections for 2019.

Numbers of households for each output area are estimated using the 2011 Census distribution for households applied to the 2018-based ONS households projections for 2019.

Employment centres: small, medium, and large

3 types of employment centres were determined using the number of jobs in each English LSOA as specified by the 2019 Business Register Employment Survey (open access), downloaded via Nomis.

The types of employment centre were as following:

• large employment centres for LSOAs with 5,000 or more jobs
• medium employment centres for LSOAs with 500 to 4,999 jobs
• small employment centres for LSOAs with 499 or less jobs

The co-ordinates for each employment centre is determined by using the population weighted centroid for each employment LSOA mapped to the nearest link (road) on the road network. In a very small number of cases an alternative nearby point on the road network was used, usually the nearest road node (junction), due to issues arising from normalising (straightening out) the road network in the calculation.

The employment user population is the number of 16 to 74 year olds in each output area from the ONS 2019 mid-year population estimates.

Education: primary schools, secondary schools, and further education

The data source used for the primary school, secondary school, and further education destinations is the ‘All establishment data’ from Get Information about Schools (previously Edubase) published by Department for Education.

The dataset was filtered to include only educational establishments which were open on the 8 October 2019.

Educational establishments were then filtered on admissions policy. To be included, educational establishments were required to have an admissions policy field of non-applicable, non-selective or NULL.

Only educational establishments with the following establishment types were included:

• community school
• voluntary aided school
• voluntary controlled school
• foundation school
• further education
• academy sponsor led
• sixth form centres
• academy converter
• free schools
• free schools 16 to 19
• university technical college
• studio schools
• academy 16 to 19 converter
• academy 16 to 19 sponsor led

Additional filters were then used to separate out primary schools, secondary schools, and further education as follows.

Primary schools are defined as educational establishments with a phase of education field of primary, middle deemed primary, or all-through.

Secondary schools are defined as educational establishments which had:

• a phase of education of secondary, middle deemed secondary, or all-through
• a statutory low age is less than 16
• a statutory high age greater or equal to 16

Further education are defined as educational establishments which:

• had an Official Sixth Form (NULL and not applicable values were allowed) or FE college
• a phase of education of middle deemed secondary, secondary, 16 plus, or all-though
• further education type not Specialist Designated College, Land-Based College, or Art, Design and Performing Arts College
• a statutory low age of less than or equal to 16
• a statutory high age of greater than 16

Education filtering methodological change for 2019

Open and close dates.

Schools are now required to be open on 8 October 2019 (the second Tuesday of October). Previous Journey Time Statistics only included schools open the whole of the relevant academic year.

Boarders.

Schools are no longer excluded based on boarding status, as it is difficult to distinguish between specialist boarding schools and schools which are important to the local community and have only a small number of boarders. Schools with the following boarding types are no longer excluded, which are:

• children’s home (boarding school)
• boarding school
• college or FE residential accommodation

Secondary school establishment type.

Secondary schools with the following establishment types are no longer included, which are:

• foundation special school
• free schools special
• special post 16 institution
• city technology college

FE type.

The following specialist colleges are no longer included, which are:

• arts colleges
• land-based colleges
• specialist designated colleges

FE statutory low age.

A statutory low age of 16 is required for a further education establishment to be included.

FE phase of education.

FE establishments with a phase of education value of ‘non-applicable’ are now excluded as these relate to administrative locations only.

The primary school user population is the number of 5 to 10 year olds in each output area from the ONS 2019 mid-year population estimates.

The secondary school user population is the number of 11 to 15 year olds in each output area from the ONS 2019 mid-year population estimates.

The further education user population is the number of 16 to 19 year olds in each output area from the ONS 2019 mid-year population estimates.

GPs

The data source for the GP dataset was Locations of GP surgeries with registered patients in October 2019 published by NHS digital.

The ONS postcode directory was used to link the postcodes to co-ordinates.

The GP user population is the number of households in each output area, determined by the 2011 Census, in addition to Local Authority (LA) updates from ONS 2019 mid-year household projections.

Hospitals

The objective for the hospital dataset is to include all NHS ‘general’ hospitals, however no clear definition of a ‘general’ hospital is currently available. Consequently, any definition used is likely to be partly subjective and could either include or exclude a very small number of hospitals which might otherwise be considered to be in or out of scope.

Hospitals data source methodological change for 2019

There has been a change in the data source for the hospital destination dataset for the 2019 statistics. This has resulted in there being an overall smaller number of hospitals included in the analysis with a consequent impact on calculated journey times.

The hospitals used were from the Estates Return Information Collection (ERIC) Site Data for the year from 1st April 2019 to 31st March 2020 published by NHS Digital, using the below methodology approved by the responsible statistician for the ERIC data.

The dataset was filtered to only include hospitals with the following acute trust types:

• large
• teaching
• small
• medium
• multi-service

Additional filters were used to only include hospitals with site types of:

• general acute hospital
• specialist hospital (acute hospital)
• mixed service hospital

Specialist hospitals were removed using the following key word filters:

• birth centre
• maternity
• eye
• rheumatic diseases
• throat, nose and ear
• neurology or neurosurgery
• specialist emergency care
• orthopaedic
• heart hospital

In addition, a minimum gross internal floor space requirement of 8,000 metres squared for locations with a trust type of general acute hospital and 30,000 metres squared for locations with a trust type of specialist hospital (acute) or mixed service hospital.

The postcodes in the ERIC data were then joined to the ONS postcode directory to provide the location co-ordinates for each hospital.

The hospital user population is the number of households in each output area, determined by the 2011 Census, in addition to Local Authority (LA) updates from ONS 2019 mid-year household projections.

Food stores

A commercial data source was used for food stores. The food store destination data contains locations of chain grocery, chain supermarket or convenience stores open in October 2019. Independent stores were excluded.

The postcodes were then joined to the ONS postcode directory which provided co-ordinates for each food store.

Food stores methodological change for 2019

In previous years, food stores have been deduplicated at postcode level. For Journey Times Statistics 2019, duplicated postcodes are allowed.

The food store user population is the number of households in each output area, determined by the 2011 Census, in addition to Local Authority (LA) updates from ONS 2019 mid-year household projections.

Town centres

Locations of English town centres in 2004 from the DLUHC (Department for Levelling Up, Housing and Communities) Town Centres and retail planning statistics for England and Wales. The coordinates used were central focal points for each town then mapped to the nearest road.

The town centre user population is the number of households in each output area, determined by the 2011 Census, in addition to Local Authority (LA) updates from ONS 2019 mid-year household projections.

Road network

Road network methodological change for 2019

The 2019 release uses the OS Mastermap Highways Network (OS Highways), as the OS MasterMap Integrated Transport Network (ITN) Layer used in previous Journey Time Statistics releases has been discontinued.

The statistics have been compiled using OS Premium Data under the DfT OS License 100039241. Calculations for walk, public transport, and cycle use the roads and paths network from the OS Mastermap Highways Network, whilst calculations for car use the road network only.

Public transport network

The public transport timetables used are those in the National Public Transport Data Repository (NPTDR) compiled by Basemap. Timetables are compiled from the Traveline National Dataset (TNDS), Associate of Train Operating Companies (ATOC) and National Public Transport Access Node (NaPTAN) stop references.

The following public transport modes are included in the data. These are:

• bus
• coach
• National Rail
• light rail
• tube
• ferry
• tram
• metro

The NPTDR release used for the 2019 calculations is the second week of October 2019.

Road speeds

Road speeds methodological change for 2019

Road type classification. There has been a change in road type classifications with the change in road network, resulting in a slight change in the default cycle and car speeds.

Car speeds. Only speeds from cars are now included in analysis, whereas previously speeds from all vehicle types were used. Since speeds are higher for cars only compared with all vehicles, this has the effect of increasing speeds on the network.

Car speeds were calculated from traffic data for the year from September 2018 to August 2019, as the traffic data year is required to match the road network at the start of the traffic data year. Only speeds from 7am to 10am on weekdays for cars were included.

A sampling threshold of at least 10 samples for every month across all 12 months was required for individual link speeds for each direction. Where this threshold was not met, the speeds for the opposite direction was used. Where is was not possible to use either the speed from either of the 2 situations above, then default speeds were used for each road type. These default speeds were calculated as an average from all links (regardless of sample size) for each road type and are set out in Table 2.

The data is procured by DfT from a commercial organisation that has GPS boxes fitted to individual vehicles monitoring the location and speed of the vehicle at given time intervals when the vehicle ignition is on. More information is available in the background quality report for the road congestion and travel time statistics.

Table 2: Speeds (kilometres per hour) by road type and mode, 2019

Road type	Walk	Cycle	Car
Motorway	0	0	83.8
A Road	4.8	16	38
B Road	4.8	16	40.7
Minor Road	4.8	16	36.9
Local Street	4.8	16	25.6
A Road Primary	4.8	16	47
B Road Primary	4.8	16	31.7
Local Access Road	4.8	16	22.4
Restricted Local Access Road	4.8	4.8	24.5
Secondary Access Road	4.8	4.8	42.3
Restricted Secondary Access Road	4.8	4.8	45.6

Walk and cycle speeds are not derived from any particular source and are outlined in the Model Parameters section.

Model parameters

General parameters

Maximum journey time of 2 hours. Maximum journey distance of 100 kilometres. Maximum distance from both origin and destination each to road network (external connection distance) of 2000 metres. The off-road network speed across all modes is 4.0 kilometres per hour.

Road network parameters

Road network envelopes were used for all calculations. This creates a square buffer a specified outside the entire origin and destination set so only roads links within this boundary are used. The following values were used for each mode type. These are:

• 1 kilometres for walk only
• 2 kilometres for cycle
• 5 kilometres for public transport or walk
• 5 kilometres for car

Each road link was converted to a straight line (normalised) to speed up calculations.

Public transport parameters

Interval within which door-to-door journey must be completed is 7am to 10am on a Tuesday.

Maximum walk distance of 2 kilometres when walking directly from origin to destination without using public transport.

Maximum straight line distance between public transport interchanges (internal connection distance) of 500 metres

Public transport stops within 150 metres were grouped together at a single point (stop clustering). The individual timetables for each service are retained.

A minimum of 5 minutes was required between public transport interchanges.

A public transport network envelope of 20 kilometres. This creates a square buffer 20km outside the entire origin and destination set so only public transport stops within this boundary are used.

5 minutes were added to any journey time that uses any public transport to allow for catching first public transport service.

Car

5 minutes were added to journeys by car to allow for parking at the end of the journey.

Cycle

5 minutes were added to journeys by cycling to allow for storing cycles at the end of the journey.

Contact details

Journey time statistics

Email [email protected]

Media enquiries 0300 7777 878