Learning from elsewhere

In Part 1 the use of indicators that have been developed elsewhere was discussed. Chapman and Weir (2008) recommended the development of a comprehensive accessibility framework for New Zealand similar to what has been implemented in the UK.

The UK practice recognises that travel time is a necessary condition for access so national indicators based on travel time can identify gaps in transport systems, or in the location of facilities. Local authorities should also be able to identify other gaps because just having an achievable travel time is not the sole determinant (or condition) for access.

The UK national indicators were developed through extensive negotiations across sectors and between national and local government. This means the resulting indicators are a function of the process by which they were agreed. While no individual party would necessarily choose the indicators, it was an achievement to gain the agreement of such a wide range of organisations from health authorities to the national department of transport. However with 97 different indicators, each with multiple ways of representing travel time separation, resulting in 568 different indicators in total, the UK approach is inevitably a compromise. Ten years after the indicators were first put forward, only a few are being used. In 2010 the indicators were also renamed ‘accessibility statistics’ to clarify that they were a resource for use nationally and locally to help with accessibility planning, rather than an indicator of performance for any individual organisation.

It is of note that there has been an unhelpful debate between central and local government about thresholds. Central government was careful not to set travel time thresholds due to concern the thresholds could become a target. National government used continuous indicators for monitoring for five years but the Hansen style indicators were poorly understood. With a change of government in 2010 leading to greater autonomy for local government, the monitoring of local government largely ceased including monitoring accessibility. The national accessibility indicators are therefore published as a resource to assist with accessibility planning.

Across England, Scotland and Wales the only indicators in common are travel time to health, education and shops. These have been widely used for comparisons across the UK, including by campaigning groups, and are widely regarded as being the most established indicators. These indicators have the advantage of being both easy to understand and robust. For example, most people can understand a statistic for each house address that shows the total time it takes to walk to a grocer, school and GP, or the total public transport travel time to reach a major hospital, further education and a regional shopping centre. When looking at the competitiveness of modes, the ratios of these travel times have been most useful and the UK planning policy guidance recommends that ratios of car to public transport travel times are reported for all new developments.

One of the first decisions needed in New Zealand will be how wide a consensus is needed on the choice of indicators. For example, if the NZ Transport Agency (NZTA) sets indicators for access to employment without consulting other agencies, these agencies may see the chosen indicators as irrelevant or may oppose them. In the UK, access to health and education are statutory responsibilities of the health and education departments so if transport agencies set targets for these without first consulting the other groups then there could be complex consequences.

New Zealand local authorities may also have significant differences of opinion as was observed in the UK, and, for example, prefer indicators that better suit New Zealand’s rural character in some regions.

The UK national core indicators provide a useful snapshot of travel times. The primary focus is on destination access for two main reasons (DfT 2008):

1      The calculation is easier when it requires only the nearest destination of any particular land use type.

2      The indicators clearly differentiate the characteristics of the catchment population allowing opportunities for socially excluded population groups to be compared with the total population.

In the UK it has been particularly difficult to gain consensus on the origin indicators. How much retail floor space is available or the number of jobs accessible from each origin were initially published but the indicators proved to be too sensitive to variability in data quality. Over time the indicators have been allocated to broad bands representing the level of choice, but further work is needed to calibrate perceptions of choice with measured opportunities.

All countries using accessibility indicators change the indicators used in monitoring regularly so the lesson for New Zealand seems to be to maintain a flexible approach and use relevant indicators for each policy need. Common pitfalls from the UK analysis (DfT 2008) are reported as:

·         The indicators are poorly understood and used.

·         Each indicator should be optimised for planning and other needs.

·         There are currently too many indicators for the purpose of monitoring.

·         Further work is required to determine the strength and weaknesses of each indicator at their respective national, regional and local implementation level.

·         The chosen thresholds should relate to policy, monitoring or behavioural criteria, for example:

-         if monitoring then the lower threshold should relate to about a third of the population

-         if monitoring then the upper threshold should relate to about two thirds of the population

-         behavioural thresholds should relate to the population being considered – for example, NZHTS data should be applied in New Zealand

-         policy thresholds need to be developed with other government agencies – in New Zealand this would include targets developed with the Ministry of Education, Ministry of Health etc

·         Where data quality is high the indicator quality is also high, and conversely poor data quality leads to low indicator quality.

·         Car accessibility should be included in the accessibility calculation because it would be helpful for non-transport stakeholders to understand relative provision between modes and destination choices.

·         Ratios of accessibility of car and non-car modes would be useful to show locations where there are gaps in public transport/cycling and walking provision.

All of these issues will be important in designing a suitable indicator set for New Zealand.