Identifying Through Accessibility Planning how Sustainable Growth can be achieved in the Compact City: A case study of Edinburgh

This paper describes how measures of accessibility have been used in Edinburgh to support sustainable land use planning decisions. The paper explores the relationship between accessibility and travel demand by each mode and considers the circumstances in which improved accessibility by public transport and walking can reduce car use. This paper was first presented to the European Transport Conference, Strasbourg, October 2005, by Duncan Smith and Derek Halden.

  1. Introduction
  2. What is accessibility and is better accessibility sustainable?
  3. The Structure Plan Accessibility Assessments
  4. Are measures of accessibility by public transport, walking and cycling an indication of sustainability?
  5. Conclusions
  6. References

1. Introduction

If we are to ensure that development is sustainable, we need to consider all aspects of the development including the consequential transport impacts. Changes to the built environment are by their nature long term, and once a development is complete, and travel patterns established, they are difficult to alter. This paper describes how measures of accessibility have been used in Edinburgh to support sustainable land use planning decisions. The paper explores the relationship between accessibility and travel demand by each mode and considers the circumstances in which improved accessibility by public transport and walking can reduce car use.

2. What is accessibility and is better accessibility sustainable?

At its simplest level, accessibility is the ease of reaching opportunities or the ease of being reached (Jones 1981). Accessibility is an attribute of people and goods rather than transport modes or service provision, and describes integrated systems from a user viewpoint.

Some have argued that if better accessibility leads to more travel, and if more travel is unsustainable, then better accessibility must be unsustainable. Such arguments selectively confuse many concepts, so to gain a clearer understand the actual relationship between accessibility and sustainable development, it is important to start with a definition of accessibility that clarifies all the dimensions of accessibility and then use these concepts consistently to identify the key relationships.

As a starting point it is important to be clear about the focus of attention, specifically whether accessibility is being considered from the viewpoint of people or places. When considering people, accessibility is about “the ease with which any individual or group of people can reach an opportunity or defined set of opportunities”; this is often referred to as origin accessibility. When considering places, accessibility is “the ease with which a given destination can be reached from an origin or set of origins” (Simmonds et al 1998); this is usually referred to as destination accessibility, catchment accessibility or facility accessibility.

All definitions of accessibility, whether explicitly or implicitly, include three key elements:

  • The category of people or freight under consideration – Each section of the population has specific needs and desires to be involved in defined activities.
  • The activity supply point – Opportunities are defined in terms of the land use supply which would allow any individual to satisfy their desire to participate in the activity under consideration.
  • The availability of transportation or other communication – This defines how an individual could travel to reach the relevant facility. In assessing the transport options it needs to be recognised that all stages of each possible journey by each available mode must be taken into account.

The level and detail of the analysis of each of these issues needs to be commensurate with the decision being considered and the detail with which particular problems need to be understood. It should also be noted that sustainable development takes place within a context where the boundary between transport and other communications, as a means of gaining access, is becoming increasingly blurred.

Considerable confusion has resulted from differences in treatment about what is implicit and what is explicit. Organisations specialising in providing services, such as employment agencies, health services, supermarkets, etc. tend to use very simple proxies (such as distance) to represent transport and communication, and concentrate on the facilities available and the population characteristics within the catchment.

In contrast, transport planning has concentrated in greater depth on the transport system operation, looking in great detail at movement patterns rather than giving detailed consideration of why they are travelling, and where they are going (DHC 2000). People and opportunities have therefore been considered within the planning of improved transport only to the extent that the characteristics of the people (e.g. physical disability or car ownership) or of the places (e.g. pedestrianised area) affect mobility and the demand for travel.

Most practical accessibility planning in Europe is carried out at one of three main geographical levels:

  • Local accessibility to facilities in the neighbourhood
  • Regional accessibility often for cities and their hinterlands
  • Interregional accessibility to measure the connectedness of a region or country

In practice these typologies need to be linked to the decisions required at each level: within communities, neighbourhoods, and larger administrative areas. This means that applied accessibility measures need to be closely related to the relevant policies and required decisions.

If accessibility is improved at one level, e.g. the inter-regional level, but this is not balanced by changes at the local level, or if accessibility changes are made by one mode, e.g. car but not by other modes, then the imbalance, like other economic imbalances (e.g. between production and consumption) will be unsustainable. Improved accessibility does not therefore always lead to sustainable development but improved information about accessibility change, if this information is comprehensive and representative of real behaviour, will always enable decision makers to make sustainable choices in land use planning and transport investment.

Perhaps the greatest challenge is therefore to ensure that analysis is sufficiently comprehensive and representative of real behaviour. There are two main approaches to ensure representativeness:

  • The normative approach
  • The behavioural approach

The normative approach is generally based on some assessment from a non-transport sector which for land use planning might be a specific economic or planning objective e.g. to zone land for business development where the catchment population exceeds a threshold value. The behavioural approach seeks to assess the value of these travel needs as they are expressed through observed behaviour – whilst recognising that ‘revealed’ behaviour may not always be a good measure of ‘preferred’ behaviour.

Although some have suggested that land use planning can be used to influence travel behaviour (Cervero 1989), it is not yet clear whether changing the built environment actually leads to changes in travel behaviour, or simply that people are choosing the type of lifestyle they want (Handy 2004). For example, people who want less car dependent lifestyles choose to live in locations where there is good local non-car accessibility, and those who want more car dependent lifestyles operate within much wider boundaries of space and time.

The evidence is, however, unambiguous that market forces acting in isolation will tend lead to a decline in accessibility for many people, since markets operate within narrower boundaries than the broader social and economic needs which characterise sustainable development (Halden 2005).

Within any population group, travel demand and accessibility by any mode are strongly correlated (Ecotec 1993). The ability to reach clearer conclusions about these relationships requires more studies setting transport and accessibility within wider lifestyle choices (Lyons 2004).

Handy and Niemeier (1997) adopted a consumer welfare approach to measuring accessibility, by linking individual travel behaviour with journey-to-work mode and destination choice. The approach joined a transport model and a compensating variation model based on econometrics. The aim was that accessibility could be assigned a value i.e. the monetary “worth” of accessibility for the journey-to-work trip. This approach was adapted and simplified for the Structure Plan option appraisal in Edinburgh and the Lothians as discussed below (DHC 2000).

3. The Structure Plan Accessibility Assessments

Development Options

Traffic leaving Edinburgh.

Traffic leaving Edinburgh.

A consultation on major issues for the Structure Plan was undertaken in 1999. The population of Edinburgh and the Lothians is over 750,000 and is experiencing the fastest population and employment growth in Scotland. Analyses of anticipated population and employment growth and the demand for new housing, allowed five development scenarios to be identified as illustrative of the options for land use changes which could be included in the Structure Plan. The five scenarios were:

  • Committed Development – This included population and employment change based on the latest audits of housing and employment including development under construction.
  • City Expansion – A development approach which emphasises the development of brownfield sites reinforcing Edinburgh as a compact city with a high density core.
  • Green Belt Development – Development within the long standing continuous green belt around the city but linked to existing transport networks.
  • Development of the Landward Towns – Major expansion of existing towns in the Lothians, particularly towns in need of regeneration.
  • Development in New Settlements – New communities in the Lothians with a balance of housing and employment.

In practice these five scenarios were not mutually exclusive, and about 80% of the housing and employment growth was common to all five scenarios, being based on proposals from previous Structure Plans.

Transport Options

Studies associated with the development of transport strategies and plans for the South East Scotland Transport Partnership and the City of Edinburgh Council had identified three broad expenditure scenarios for transport linked to practical funding of transport as follows:

  • Committed Minimum – This included only those schemes Council funds were already committed to promoting.
  • Low Spend – This included the schemes which may be fundable with no increases in resources other than those available through Council transport funding from national initiatives such as the Public Transport Fund.
  • High Spend – This included those schemes which would only be fundable if a road charging scheme were implemented in Edinburgh.

For each scheme in these packages, the time and cost impacts had previously been estimated and these were coded up for each scenario within the accessibility model. Although several options were being considered for road charging in Edinburgh, the option included in the high spend scenario included a charge at the charging cordons and where a trip crossed either or both cordons the charge was taken as £2.

The flexibility of accessibility analysis means that it is very easy for an analyst to produce too much information and therefore hamper clear decisions by policy makers. It is important to clarify what policy questions need to be answered and confine the analysis to looking at these. There were many potential combinations of land use, people and transport factors but the test programme was restricted to 125 tests.

Accessibility indices were reported for the combinations of origins, destinations, time of day, people, trip purpose and mobility of greatest interest to policy development. 125 tests may seem like a very large number, but it had been found in previous accessibility analysis for the area (Simmonds 1998) that more aggregate analysis of accessibility to reduce the number of tests obscured the important changes for particular people groups, trip purposes, and times of day.

Access to employment

The zones with the best car available accessibility to jobs are located in Edinburgh and the west side of West Lothian. As congestion grows, car available accessibility decreases by about 10% outside Edinburgh, but in central Edinburgh the equivalent fall in accessibility is nearer 30%. The predictions illustrate that if action is not taken to curb city centre congestion then the pressures for decentralisation will be substantial.

For non-car available accessibility there is a significantly different picture. Peak accessibility is achieved in central Edinburgh and accessibility then declines radially from central Edinburgh, although the influence of public transport corridors is clearly visible.

Each of the development scenarios is compared with the base level of development. Committed development has a relatively small impact outside Edinburgh. The New Settlement and Landward Towns scenarios spread the impacts of the development most widely particularly for non car available accessibility.

Access to Shopping

Car available access to retail is greatest for residents of Edinburgh and West Lothian. The influence of the west of Scotland is significant on the west side of West Lothian. Car available accessibility is good throughout but non-car available accessibility is much greater near to major shopping centres. The ratio of distance to generalised travel time for public transport is much lower by public transport than by car, so city centre locations with a large supply of shopping result in city centre residents experiencing much higher levels of accessibility than those outside the city.

Access to Education

The location of universities and colleges accounts for the main differences between areas. For car available accessibility, Central and West Edinburgh perform best due to the location of Edinburgh, Heriot Watt and Napier Universities. The poorest access to education for both car and non car available trips is in East Lothian.

Away from the universities and colleges much of the area has similar, and not particularly good non car accessibility to education but West Lothian performs much better for car available accessibility. This imbalance is likely to create pressure for car dependent patterns of access to Education in West Lothian.

Access to Healthcare

As for Education, the difference in non car available accessibility to healthcare between the best and worst performing areas is much greater than for car available accessibility. In each case the south side of Edinburgh City centre has the best accessibility, but for car available trips most of Edinburgh, and all but eastern East Lothian and western West Lothian have good accessibility. For non car available accessibility the proximity to major hospitals is a much more significant factor.

Access to population

Calculations of car available accessibility for the population show that eastern West Lothian has the greatest catchment with a population index of 1.1 million people. This compares with eastern East Lothian with the lowest population index of 0.37 million people. Zones on the west side of West Lothian also perform very well due to the large population centres in West Central Scotland. For non car available trips, population density becomes much more important with the south side of central Edinburgh performing best with Dalry having the highest index at 125,000 people.

Impacts of transport and development changes

The introduction of road pricing and the associated public transport improvements reduces car available accessibility in the city centre by about 40% but increases non-car accessibility by over 55%.

Adding the committed development to any of the transport options reduces non car accessibility by up to 7% and car accessibility by up to 3%. In contrast each of the four proposed development scenarios has a positive impact. For non car accessibility, even with low spend transport options, accessibility is maintained or improved throughout the area with increases of between 2% to 45%. For car accessibility, each of the development scenarios increases accessibility throughout the area by between 4% and 10%. In both cases the largest percentage increases are in eastern East Lothian and the largest absolute increases are to the East and South East of Edinburgh.

Table 1 summarises the impacts of the different development scenarios.

Table 1 – Summary of Development Impacts

Scenario Summary Comments
City Expansion Maximises car and non car accessibility within the city but has the least impact outwith the City. Nevertheless, destination accessibility changes outside the city, particularly to the south and east are still substantial.
Green Belt Maximises car and non car accessibility for zones around the City Bypass but the effects spread more widely to the Lothians for car accessibility, particularly destination accessibility.
Landward Towns Lower level of both car and non car accessibility within the city, but higher level of non car accessibility in some Landward Towns.
New Settlement Similar pattern as for the Landward Towns scenario, but the impacts are greater in Mid and East Lothian. The largest impacts for car and non car destination accessibility are in East Lothian but origin accessibility for these zones is only marginally different from other options.

It is also possible to consider accessibility in aggregate for the study area. Such analysis provides a measure of how well land use and transport are integrated within the Structure Plan area. As part of the accessibility model calculations, population weighted utility indices were developed for each zone for a range of trip purposes. These can be used to develop composite indicators of the overall value of the scenario for the affected population as shown in Table 2. The composite indicators include access to work and population and seek to provide a reasonably robust indication of the sustainable development impacts for each alternative development scenario.

Table 2 – Comparison of Overall Accessibility Change

Committed City Expansion Green Belt Landward Towns New Settlement
Generalised Minutes x106
Change from base development level -1.46 -4.87 -5.22 -5.15 -5.04

The lowest values indicate the highest degree of integration between land use and transport since there are fewer generalised minutes associated with travel. Overall the best option in accessibility terms was the Green Belt development scenario which combines the benefits of development near the city with positive impacts for areas outside the city. In moving to a preferred option the choice of development scenario therefore needed to combine features from each of the tested scenarios.

4. Are measures of accessibility by public transport, walking and cycling an indication of sustainability?

If sustainable development aims are to minimise travel demand within any development scenario, then the measures which best represent travel demand will have a key role in future analysis. The limited analysis reported below was undertaken since it was suggested that simple public transport, walking and cycling accessibility measures for travel to work might be an adequate proxy for sustainable transport. This contrasts with the rather than the more complex utility based measures covering multiple trip purposes reported above. There is a need for day to day decision making on development planning to use simple measures so the analysis can be repeated regularly to assess the impacts of even quite small developments.

It was important to adopt readily available data. The analysis therefore considered work trips, and the accessibility measures used were accessibility of workplaces by public transport and walking, based on journey times and the location of jobs. Journey cost and other softer factors affecting travel decisions have not been included since data are harder to obtain and the aim was to identify whether or not accessibility measures using readily available data could predict the demand for car travel.

The first point of note is that the city of Edinburgh starts from inherently sustainable land use patterns. Development is focussed around the high density city centre, with the main transport links forming a radial structure around this. There is an unusually high degree of business centralisation in Edinburgh, based around the New Town and West End areas and this combined with relatively good public transport leads to a baseline low modal share for car travel.

Table 3: Journey to Work Modal Split

% Train Bus Car Cycle On foot
Scotland 3.7 15.1 62.7 1.5 15.2
Edinburgh Residents 1.2 26.1 41 3 28.7
Employees in Edinburgh 3.8 24.18 47.6 2.4 22

However as the city expands, businesses are increasingly attracted to sites outwith the congested city centre. Leith and the Gyle are the two main business growth areas. City centre destinations have by the far the fewest proportion of car trips. This sustainable pattern extends to destinations in central eastern wards, and drops off more quickly in the west. Close to the city bypass and particularly junctions, car accessibility is highest and the most unsustainable modal splits are found.

The Gyle ward on the western outskirts of the city combines an unsustainable mode split of 70% car, with a high number of work trips due to the recent business development. The Gyle/Edinburgh Park is a large scale planned business park begun in the early 1990s, and has been highly successful in attracting the financial services industry (CEC, 2002).
Regression analysis was used to quantify the strength of the relationship between accessibility by public transport and car travel demand. The results for the city centre and the Gyle/Edinburgh Park sites are shown in Tables 4 and 5 respectively.

Table 4: PT accessibility as a predictor of city centre car modal share

R squared values for linear regression analysis, with trips by car to city centre as the dependent variable.

Accessibility by PT and walking Households with 1+ Cars Households with 2+ Cars Accessibility and 1+ Cars Accessibility and 2+ Cars
0.31 0.53 0.67 0.65 0.74

Table 5: PT accessibility as a predictor of Gyle car modal share

R squared values for linear regression analysis, with trips by car to city centre as the dependent variable.

Accessibility by PT and walking Households with 1+ Cars Households with 2+ Cars Accessibility and 1+ Cars Accessibility and 2+ Cars
0.35 0.58 0.56 0.73 0.67

Despite public transport accessibility often being used within sustainability appraisals as a transport indicator it can be seen that it is actually a very poor predictor of travel demand. It is also notable that it is an equally poor predictor for both the City centre and the Gyle/Edinburgh Park.

This is perhaps not surprising. The normal laws of supply and demand would suggest that public transport availability and travel time could be expected to be a predictor of public transport travel demand and this was found to be the case in the analysis. However, there is no inherent reason why public transport supply should be highly correlated with car travel demand, although assumptions about modal shift often make this suggestion.

Car availability may be a slightly better predictor of car travel demand even though car travel time is not included in simple analysis using readily available data. The results show that even car availability is a relatively poor predictor of car travel demand. More detailed analysis using a transport model to consider travel time and congestion would be needed to obtain a stronger relationship.

The analysis therefore demonstrates that the public transport accessibility indicators commonly used in development planning to assess the sustainability of transport solutions would be very poor measures of sustainability in the case of Edinburgh. This echoes similar conclusions from other more comprehensive studies (e.g. CBP 2001).

5. Conclusions

The City of Edinburgh is facing rapid expansion and new development locations need to be found in inherently less accessible locations. Travel demand management through sustainable land use planning is now even more pertinent since economic mechanisms such as road pricing are no longer being viewed as short term options.
Previous work has shown that imbalances in the economy, including accessibility imbalances are inherently unsustainable. Assessing the sustainability of development plans therefore requires accessibility analysis to consider the spatial distribution of resources and activities.

Accessibility analysis shows that accessibility measures by any particular mode or modes for any specified relevant people group gives a good indication of the demand for travel by that group using these modes.
However accessibility analysis also shows that some assumptions about public transport supply and its impacts on modal shift may be seriously flawed.

6. References

  • Cervero, R., (1997) ‘Paradigm Shift: From Automobility to Accessibility Planning’, Urban Futures: Issues for Australian Cities.
  • City of Edinburgh Council (2003), Local Transport Strategy, CEC: Edinburgh.
  • City of Edinburgh Council (2002), Major Developments: Edinburgh Park/South Gyle, CEC: Edinburgh.
  • Colin Buchanan and Partners (2001), Planning for Mode Share in New Development,
    Scottish Executive: Edinburgh.
  • Derek Halden Consultancy (2003a), Accessibility: Review of Measuring Techniques and their Application, Scottish Executive Social Research: Edinburgh.
  • Derek Halden Consultancy (2003b), Barriers to Modal Shift, Scottish Executive Social Research: Edinburgh.
  • General Register Office for Scotland (2001), Scottish Census 2001, http://www.scrol.gov.uk/scrol/common/home.jsp.
  • Halden, D., (2000b) ‘Using accessibility measures to integrate land use and transport policy in Edinburgh and Lothians’, Transport Policy, Vol.9, pp 313-324.
  • Halden, D (2002), Using accessibility measures to integrate land use and transport policy in Edinburgh and the Lothians, Transport Policy 9, pp 313-324.
  • Halden (2005) Accessibility Planning as a Tool for Transport Delivery. Scottish Transport Research and Applications Conference.
  • Halden, D., McGuigan, D., Nisbet, A., McKinnon, A., (2000) Accessibility: Review of Measuring Techniques and their Application, report to the Scottish Executive, Central Research Unit.
  • Hagerstrand T (1973). The Domain of Human Geography. New Directions in Geography. Methuen. London
  • Handy, S (1992) ‘Regional Versus Local Accessibility: neo-traditional development and its implications for non-work travel’, Built Environment 18, pp253-267.
  • Handy, S and Niemeier, D (1997) ‘Measuring Accessibility: Exploring issues and alternatives’, Environment and Planning A, 29, (7), pp1175-1194.
  • O’Sullivan D, Morrison A, Shearer J (2000) Using desktop GIS for the investigation of accessibility by public transport: an isochrone approach, International Journal of Geographical Information Science, 14 pp 85 -104.
  • Ortuzar, J D and Willumsen, L G (1994), Modelling Transport, Wiley and Sons: Chichester.
  • Openshaw, S (1984) The Modifiable Areal Unit Problem. Geo Books/Headley Brothers Ltd: Kent.

Leave a Reply

Your email address will not be published. Required fields are marked *