The new draft London Plan, published for consultation on 16 July 2026, appears to place sustainable accessibility at the heart of its approach to accommodating growth, and whilst it is acknowledged that there remains more to digest, this article reflects early observations of two specific elements of sustainability in the transport context due to current projects leading me to invest some more thinking around them.
The draft London Plan introduces a new Sustainable Access Measure (SAM) and a separate framework for determining whether Green Belt development is in a “sustainable location”. Together, they represent a significant change in how transport accessibility may shape development density, parking and the location of growth across London. But do the measures genuinely define sustainability—or primarily rank and select locations for development?
These proposals have only just been published, and the underlying methodology may evolve through consultation. Nevertheless, they raise important early questions about what is being measured, how the results should be interpreted and whether the emerging framework provides a sufficiently robust basis for describing a location as “sustainable”.
From access to the network to access to opportunities
For many years, PTAL has been London’s principal measure of public transport accessibility. It measures the ease with which someone can reach and enter the public transport network, principally through walking time, waiting time, service frequency and route availability.
Its limitation has long been clear: PTAL says relatively little about where those services actually go.
A location may have several frequent bus services and therefore a reasonable PTAL score, but those services may not provide fast access to a large range of jobs or facilities. Conversely, a location slightly further from a railway station may have a relatively modest PTAL score, even though the station provides rapid connections to a very large employment and service catchment.
SAM has been designed to address that issue. Rather than measuring access to the transport network alone, it considers the destinations that can be reached within particular journey times by public transport and walking. TfL describes its objective as measuring the ability of a typical Londoner living in a given location to meet regular travel needs using sustainable, space-efficient modes.
That is an important and, in principle, welcome development. Accessibility is ultimately about the opportunities people can reach—not simply the distance to the nearest bus stop.
How SAM is constructed
SAM is calculated for each 100-by-100-metre grid square across Greater London. It combines three separate layers:
access to employment within 60 minutes by public transport;
access to a wider range of destinations within 30 minutes by public transport; and
access to those destinations within a ten-minute walk.
The employment layer also acts as a proxy for access to specialist or city-wide destinations. The shorter public transport and walking layers include education, shopping, health, leisure, community and other facilities, weighted broadly according to the frequency with which Londoners travel for different purposes.
Each layer is standardised using Z-score normalisation and then combined using weightings derived from statistical analysis of car ownership. The resulting weights are:
43 per cent for city-wide public transport access;
41 per cent for walking access; and
16 per cent for other public transport access.
TfL’s rationale is that locations offering better access by public transport and walking should, after controlling for income, housing type and other relevant factors, generally be associated with lower car ownership.
The final underlying score is then converted into a simplified band between 1 and 5, with additional A and B subdivisions at the highest and lowest ends. The draft Plan proposes using SAM to inform matters including housing capacity, development density and parking standards.
A significant improvement—but what does a SAM band mean?
SAM is more sophisticated than PTAL in several respects. It captures end-to-end journeys, differentiates between destinations and gives walking a substantial role. It should therefore provide a better indication of whether a person can reasonably undertake regular activities without relying upon a car.
However, the way the underlying score is translated into policy bands creates an important interpretive issue.
The main SAM bands are based on quintiles. In other words:
SAM 5 represents the top 20 per cent of London grid squares;
SAM 4 represents the next 20 per cent;
SAM 3 represents the middle 20 per cent;
and so on.
This means SAM is primarily a ranking system. It tells us how a location compares with other parts of London, but it does not, by itself, establish whether that location meets an absolute definition of sustainable accessibility.
There would always be a bottom 20 per cent, even if every part of London had excellent public transport and local facilities. Equally, there would always be a top 20 per cent, even if accessibility were poor across the city.
A SAM 2 location is therefore not necessarily “unsustainable”. It is less accessible than approximately 60 per cent of the London land area included in the assessment. Similarly, a SAM 5 location has not been proved sustainable in the round; it is simply among the most accessible locations under the factors measured by the model.
That distinction will be important once the scores are used to influence planning decisions.
The problem of variation within bands
The SAM methodology shows that the distribution of underlying scores is long-tailed. Most of London sits on a relatively smooth curve, while a small number of highly connected central locations rise sharply above the rest.
This raises questions about how much variation exists within each band.
Two locations may both be identified as SAM 5B, but one could sit just above the 95th percentile while another may have several times the underlying level of access. At the other end, SAM 1 includes both weakly connected urban locations and, through the 1A subdivision, areas such as parks and Green Belt land with little or no public transport provision.
For the measure to be properly understood, TfL should publish more than the final band. It would be helpful to see:
the continuous underlying SAM score;
the percentile position;
the minimum, maximum and median score within each band;
the individual walking and public transport layer scores; and
the sensitivity of the result to reasonable changes in journey-time thresholds and weightings.
The immediate issue is within-band dispersion: how different can two places be while appearing identical on the published map?
Would equal score bands be better?
An alternative could be to divide the range between the lowest and highest underlying scores into equal numerical intervals. That would mean the number of locations in each category could vary, but every category would represent a consistent score range.
This has some intuitive attraction, particularly if the objective is to communicate an absolute difference in accessibility.
However, because the SAM distribution has a pronounced upper tail, equal intervals based on the raw minimum and maximum could result in most of London falling within one or two lower categories. A handful of hyper-connected central locations would determine the scale for everywhere else.
A more useful approach may therefore be to report three things separately:
the continuous underlying SAM score;
the London-wide percentile ranking; and
an absolute policy classification based on defined accessibility outcomes.
An absolute classification might consider, for example, whether a location provides access to a specified mix of essential facilities within walking distance, a defined scale of employment opportunities by public transport and a credible ability to support low-car living.
That would allow a location to be described as both meeting a minimum sustainable-access standard; and
ranking in a particular percentile compared with the rest of London.
Those are related but fundamentally different pieces of information.
SAM is not a complete measure of sustainability or liveability
The title “Sustainable Access Measure” needs to be interpreted carefully. SAM measures access using modes considered sustainable and space-efficient. It does not assess the full sustainability or liveability of a location. It does not directly measure matters such as:
public transport crowding, reliability or affordability;
actual travel behaviour or carbon emissions;
walking-environment quality, road danger or severance;
air quality and noise;
housing affordability or quality;
access for people requiring step-free journeys;
biodiversity or climate resilience; or
the quality of public space and community infrastructure.
The methodology expressly acknowledges that cycling is not yet included and that walking calculations do not currently account for factors such as crossing delays, traffic conditions and topography. TfL also identifies step-free accessibility and cycling as areas for potential future development.
SAM is therefore best understood as a strategic sustainable-mode accessibility indicator. It may be an important component of a liveability assessment, but it cannot substitute for one.
Car ownership as the calibration outcome
The regression-based weighting is one of the most technically interesting parts of SAM.
Using observed car ownership gives the composite index a connection to an identifiable real-world outcome. That is particularly relevant where SAM is being used to inform parking standards and the potential for car-light development.
However, car ownership is not synonymous with sustainable travel.
Lower-income households may have low car ownership because they cannot afford a car, rather than because their location provides excellent alternatives. Some higher-income households may own cars while undertaking most regular journeys by public transport. Car ownership is also affected by housing type, historic parking provision, family composition and the legacy of earlier planning decisions.
The methodology controls for several of these factors, but the published summary does not provide sufficient detail to test the strength and stability of the model fully. It would be helpful for the supporting statistical work to publish matters such as model fit, confidence intervals, residual mapping, sensitivity testing and out-of-sample validation.
This is especially important if the results are to influence housing capacity, density and parking across the whole of London.
Destination quantity is not necessarily destination quality
SAM’s walking and 30-minute public transport layers use a large database of destination points. These are weighted by broad journey purpose, so a frequently used type of destination can contribute more than a less frequently visited one.
This provides a workable strategic proxy, but individual points do not necessarily represent equivalent opportunities.
A small convenience shop and a large supermarket may both count as food-shopping destinations. A minor medical service and a major hospital may fall within the same broad health category. Capacity, quality, opening hours, cost and eligibility are not necessarily reflected.
Observed trip frequency may also understate the importance of destinations that are essential but visited less often, such as specialist healthcare.
The measure therefore captures the availability and concentration of destination types, but not necessarily the usefulness or quality of each opportunity.
Hard journey-time boundaries
The use of ten-, 30- and 60-minute catchments also creates inevitable cliff edges.
A destination reached in 29 minutes contributes to the 30-minute layer, while one reached in 31 minutes does not. A transport improvement that reduces a journey from 36 to 31 minutes may therefore produce little or no benefit in that layer, despite being meaningful to passengers.
A continuous time-decay function—or several overlapping journey-time bands—could better represent the fact that destinations generally become less attractive as journey time increases rather than abruptly ceasing to matter at a fixed threshold.
Clear thresholds are useful for policy and mapping, but they should not create a false impression of precision.
Enter Policy PV7: defining sustainable Green Belt locations
The draft London Plan takes a more definitive approach when considering major development in the Green Belt.
Policy PV7 seeks to ensure that Green Belt release is focused in the most sustainable locations and supported by comprehensive planning, infrastructure, appropriate density and environmental enhancement. The policy responds to changes in national Green Belt policy and to the draft Plan’s conclusion that London’s development needs cannot be met entirely within the existing urban area.
The framework is important because national policy requires relevant Green Belt development to be in a sustainable location. PV7 attempts to define how that concept should be applied in the particular circumstances of London.
The PV7 sustainable-location routes
For major residential-led development, Box PV7A provides several routes by which a proposal may be treated as being in a sustainable location.
Broadly, these include:
identification through a London Plan-compliant Local Plan allocation or masterplan;
inclusion within an appropriate government-designated New Town; or
locating the vast majority of homes within the specified walking distance of a qualifying railway station or close to a bus corridor meeting the detailed requirements in PV7C.
The station-based route is tied to the stations shown in the Plan and identified in Annex C. The bus-based route is not simply proximity to any existing bus service. It requires a high-quality corridor with demanding standards concerning journey time, service frequency, capacity, dedicated right of way, operating speed, stops and passenger information.
Associated provisions in PV7B address matters including:
comprehensive masterplanning;
mixed and complementary land uses;
infrastructure delivery;
density;
parking;
access to green space; and
enhancement of the wider Green Belt.
This is significantly more than a transport-accessibility score.
PV7 addresses one of the main limitations of SAM: it sets identifiable policy thresholds rather than relying solely on a comparative percentile. A proposal either meets the relevant station or bus-access requirements—or it does not. It must also satisfy broader expectations concerning form, density, infrastructure and green-space delivery.
This gives the term “sustainable location” a clearer policy meaning. It effectively defines the minimum circumstances in which major Green Belt growth may be considered capable of supporting a well-connected, car-light neighbourhood. That is more defensible than saying that any site above a particular SAM band is automatically sustainable. However, PV7 is not an independent alternative to SAM.
The relationship between PV7 and SAM
The Green Belt Strategic Approach evidence assessed stations close to the Green Belt using factors including service frequency and the number of jobs accessible within a defined public transport journey time. The job-access analysis was informed by the same broader analytical tools underpinning SAM. The station results were then combined with walking-distance catchments and wider environmental and strategic planning considerations.
PV7 therefore uses some of SAM’s underlying logic but applies it differently.
SAM is a composite, London-wide measure based on three access layers and expressed principally as a relative band. PV7 is a specific Green Belt gateway based on station or high-quality bus accessibility, supported by design, density and infrastructure requirements.
The two measures are not formally contradictory because they serve different policy purposes. But they do reveal something important about the limitations of the SAM banding system.
When the draft Plan needs to decide whether Green Belt land is sufficiently sustainable to accommodate strategic development, it does not simply require a particular SAM band. Instead, it introduces absolute service, journey-time, walking-distance and infrastructure criteria.
That implicitly recognises that a percentile ranking alone is not enough to define a sustainable location.
Potential differences between the two systems
A location could conceivably have a relatively strong SAM score but fail the PV7 test because it is not within the catchment of an identified station or qualifying bus corridor.
Equally, a Green Belt location could pass the PV7 strategic test while currently having a modest SAM score. This may occur where the area has good strategic rail access but few existing local facilities, or where a new neighbourhood is expected to provide those facilities as part of a comprehensive development.
This is not necessarily a policy conflict. The questions are different:
SAM asks: how accessible is this location, under the modelled conditions, compared with the rest of London?
PV7 asks: is this Green Belt location part of a strategic development proposition capable of supporting sustainable, car-light growth?
The draft Plan would benefit from stating that distinction explicitly. The terms “sustainable access” and “sustainable location” otherwise risk being treated as interchangeable when they are not.
PV7 is also a spatial-strategy test
Another important point is that the qualifying station locations do not appear to have been selected solely through transport analysis.
The Green Belt evidence applies further considerations including environmental constraints, development scale, infrastructure requirements, opportunities for comprehensive growth and the preferred spatial distribution of development.
PV7 is therefore doing two things. It is testing whether a location has sufficiently strong transport characteristics, but it is also determining whether that location forms part of the Mayor’s preferred strategic pattern of Green Belt growth.
This is legitimate in a strategic plan. However, it means PV7 should be understood as a plan-led sustainable-location framework, not as a neutral or universal transport metric.
A site close to an apparently well-connected station may still fall outside the PV7 framework because its wider location has not been selected for strategic growth.
Areas requiring further clarity in pV7
PV7 raises several questions that should be explored through the consultation.
What does “the vast majority” mean?
The policy should specify the proportion of homes expected to fall within the relevant station or bus catchment. A phrase such as “the vast majority” could be interpreted differently by applicants, boroughs and decision-makers.
For a test intended to determine whether Green Belt development is sustainably located, a clear numerical expectation would improve consistency.
Why were the thresholds selected?
Walking distance, employment accessibility, service frequency and bus journey-time thresholds may all necessary if the policy is to operate in practice. However, the evidence should clearly explain whether these represent demonstrated behavioural tipping points or strategic planning judgements.
Sensitivity testing would show how many locations and potential homes move into or out of the framework if a boundary changes modestly. This would help determine whether the outcomes are robust or overly dependent on cliff-edge classifications.
Walking distance is not walking quality
A route of less than 1,200 metres may still be unattractive or inaccessible if it involves poor crossings, heavy traffic, severance, gradients or an unwelcoming public realm. An audited and deliverable route should therefore be required. Network distance alone is not sufficient evidence that most residents will regard the station as realistically walkable.
How certain must future infrastructure be?
National policy is concerned with locations that are sustainable or can be made sustainable. PV7 rightly allows planned transport improvements and new neighbourhood infrastructure to be considered.
But there must be a clear distinction between:
existing accessibility;
funded and committed improvements;
improvements secured through the development; and
aspirational proposals with no confirmed delivery mechanism.
Should there be an equivalent-outcomes route?
PV7 is intentionally demanding, particularly for the bus-based route. However, a rigid test could exclude proposals that achieve equal or better outcomes through another combination of transport, land use and demand-management measures.
A tightly controlled equivalent-outcomes route could allow an applicant to demonstrate, with compelling evidence, that a proposal would achieve at least the same:
public transport accessibility;
sustainable mode share;
employment and service access;
car-ownership outcome;
density;
parking restraint; and
infrastructure certainty.
Such a route would need to avoid weakening the strategic framework, but it could provide resilience against unforeseen transport solutions or site-specific circumstances.
What should SAM and PV7 be used for?
The emerging framework makes most sense if the tools are treated as complementary rather than interchangeable.
SAM can provide a London-wide comparison of sustainable-mode accessibility and help diagnose the relative contribution of walking, local public transport and city-wide employment access.
PTAL can continue to explain local access to the public transport network and variations within individual sites.
PV7 can operate as a higher, plan-led gateway for determining where major Green Belt growth should be considered and what infrastructure, density and masterplanning conditions should accompany it.
None of these should be treated as a complete assessment of liveability. For an individual development proposal, the final judgement should also address:
transport frequency, capacity and reliability;
walking and cycling quality;
inclusive and step-free access;
access to essential services;
delivery of local facilities;
forecast mode share and car ownership;
parking demand;
transport emissions;
public realm and road safety; and
broader environmental and community outcomes.
Early conclusion
SAM is a thoughtful and potentially valuable evolution in London’s approach to accessibility. It responds to genuine limitations in PTAL and provides a much richer picture of what people can reach without travelling by car.
Its principal weakness is not necessarily the underlying accessibility analysis, but the risk of overinterpreting the final band.
A SAM score is a relative position within London. It is not an absolute certification that a location is sustainable, nor is it a measure of liveability in the round.
PV7 goes further. It establishes clearer thresholds and combines transport accessibility with density, infrastructure, mixed uses and Green Belt enhancement. PV7 is also selective and plan-led. It reflects the Mayor’s chosen spatial strategy as well as the transport characteristics of individual locations. Meeting the policy should demonstrate that a proposal has passed an important minimum gateway; it should not remove the need to test how the development will work in practice.
Perhaps the most revealing early insight is this:
When the draft London Plan needs to make its most consequential judgement about whether land is in a sustainable location, it moves beyond SAM’s quintile bands and relies on more explicit accessibility, infrastructure and place-making requirements.
That suggests that based on the current suggested use SAM can be a powerful comparative and diagnostic tool; but not as a single definitive answer to the much broader question of whether a place is sustainable. Despite this, it does hold the potential to provide more clarity on this matter in the transport context subject to development of the understanding and interpretation of the location specific accessibility scoring output.
