Age friendly communities: promoting health and well-being and easing inequalities¹

In the next 15 years the number of people aged sixty-five and over in the UK will increase from 13m to 17m, one in three of whom will live alone. Whilst some will benefit from living in purpose-designed accommodation the vast majority will continue to remain in their family homes.

The infrastructure around them will be similar but a key change will be the way they spend their time in retirement. They will be less inclined to travel, spend more time at home and in the local area whilst longer trips will be fewer, and their average geographical radius will shrink.

One way to formulate this shift is at a neighbourhood level focusing on local amenities and services but there is a deficit in how to conceptualise and measure progress towards age friendlier living. It is not only about spending more money but making better use of existing assets and opportunities like meeting places and green space.

A problem faced by many areas is that where services are located reflects an earlier demography, and in some cases former industrial landscapes. These have led to increased separation and creeping inequality of access between where people live and the services, they rely upon, for which ageing is a factor.

But the world is changing in other ways. More people work from home rather than commuting into the office, adding to the daytime population and local footfall and social interaction. More people are choosing to extend their working lives and are active volunteers. Among the questions is how transactions work: should services or work come to them or vice versa?

Children and families accessing schools and childcare also have a local footprint, but their service needs also overlap with older people, health care being an obvious example. Our methodology copes with all ages to give an inter-generational perspective and so ascertain whether an area is both age friendly and family friendly depending on the focus of interest.

A popular model gaining wide traction for the functional organisation of neighbourhoods is the concept of a 15- or 20-minute city. In addition, the World Health Organisation (WHO) has created a global age-friendly city network with over 1300 members tasked with promoting active ageing. Membership includes Kirklees, a large English local authority, which is a focus of centre of attention here.

In the UK, the pedestrianisation of town centres is well established but the emphasis on neighbourhoods is new. For example, the 10-year NHS plan wants to introduce a neighbourhood health service, whilst the recently published National Planning Framework (NPF) supports built forms that minimise the number and length of journeys. It emphasises the health and well-being benefits of access to green space which promises to become more important in an era of climate change.

Direct support for age friendly communities is to be found in the report of the Older Peoples Housing Task Force⁴ which is feeding into the Governments housing strategy and sets out the principles for age friendly housing and communities. Amenities such as shops, libraries, leisure facilities, places of worship and parks, it says, should be close-by for ‘senior citizens to be actively involved in their communities’.

However, whilst there is a keen sense of what needs to be done there is much less agreement on the practical details, e.g., what makes an age-friendly area and what distinguishes one from another? At the same time there are huge data resources ready to be used or combined with other data to answer such questions, but it is being under exploited.

The methodology and case studies in this research are intended to fill that gap by providing a practical set of tools and insights with which to evaluate and compare communities in several dimensions – such as their proximity to each other, and the walkability of services. These are transferrable to other settings and can be used in multiple ways at a level of granularity based on point locations rather than areas.

The concepts are based on the widely accepted premise that proximity to services like a GP, library, shop or to green space confers a social value which is measurable in financial and non-financial terms. Evidence shows that there are tangible financial benefits like living near a bus stop which increases mobility but also reduces motoring costs. Other benefits are improved health and well-being leading to greater community cohesion.

The research uses publicly available data to evaluate age friendliness at local level. It draws on a detailed study of Kirklees in West Yorkshire, one of the largest local authorities in the country, and Stoke-on-Trent, a large conurbation in the West Midlands. Kirklees is a good test with its complex mix of urban and rural landscape in which challenges include awkward terrain, beautiful countryside, and an ageing population.

We enumerate assets in Kirklees from bus stops to libraries and their proximity to homes and older people. We investigate gaps, overlaps and opportunities for improvement and introduce novel concepts such as the walkability index and an age friendliness score. We focus on access to common services including health providers, post offices, shops, community facilities, places of worship and other meeting places.

The emphasis is to provide detailed insights that can be used as a starting point for conversations with local communities, transport providers, home builders, planners and others. Local authorities who have an oversight and direct responsibility for services like social care will find it particularly useful as they transition to meet the needs of a changing demography and technology.

However, this is not exclusively about local authorities or older people. Others will benefit – whether in the private or third sector. It is possible to match their organisational data to open data in many creative ways to the benefit of their clientele and their outreach. They will need to develop new analytical skills to fully capture the potential of our approach but will stand to benefit enormously as a result.

The methods and case studies showcased here can be applied anywhere in the UK. For further information contact:

Email: Lesmayhew@googlemail.com

Notes

<forward to come

<acknowledgments to come>

<references are incomplete>

Age sixty-five and over is a frequently used benchmark for older people but this age bracket covers a multitude of person types, ranging from the fit and healthy and still working or volunteering, to those with mobility and health problems which are life changing in many instances.

What they lack in similarity is compensated for by a sense of identity and community coupled with a preference for independent living. In the next 15 years the number of people in this age bracket in the UK will increase from 13m to 17m. Some will move into purpose-designed accommodation like retirement villages, but the great majority will remain in the family home and of these one in three will live alone.⁵

The infrastructure around them will be similar but a key change will be the way they spend their time. If they are retired, they will be less inclined to travel, spend more time at home and in the local area whilst longer trips will be fewer, and their average geographical radius of action will shrink. But even if they continue to work, their home will partly replace the need to commute.

One way to formulate this shift is at a neighbourhood level focusing on local amenities and services but there is a deficit in how to conceptualise and measure progress towards age friendlier living. It is not only about spending more money but making better use of existing assets and opportunities.

We know that older people spend more time in their surroundings, which encourages greater social interaction and thus is an antidote to loneliness. Similar benefits extend to other age groups, e.g., young families who rely on local health services, education, and childcare but these are subject to separate study.

Proximity to amenities and services also means that they are more likely to be used. Interaction between locations, such as a home or a shop, decreases as the physical distance between them increases. In other words, the closer together the less is friction of distance there is and increases scope for multi- purpose trips and meeting up with friends.

However, whilst there is a keen sense of what needs to be done there is much less agreement on the practical definition and metrics to measure age friendliness. Methods need to be transferrable and scalable to be able to compare neighbourhoods or settlements with one another and to facilitate closer working between authorities, providers and local people.

A problem faced by many areas is that where services are located reflect an earlier demography, past decisions and in some cases former industrial landscapes. These have meant increased separation and creeping inequality of access between where people live and the services they rely on. The ideas in this study are intended to reverse that.

There are choices such as whether services come to them or vice versa, but the balance is in flux, driven by the on-line revolution. More people work from home, which adds to the daytime population and footfall whilst fewer commute every day which over time will reshape the built environment.

Building homes for older people driven primarily by an ageing population. It is therefore an investment opportunity, and a potential boost to age friendliness, although it depends what type of housing is needed and where it is located. Retirement villages are a blended communities with services in situ, but other solutions piggyback on services available locally.

This paper is intended to fill these gaps by providing a practical set of tools and insights with which to evaluate and compare communities in several dimensions – such as their proximity to each other, the services provided and choice, and their size and geography. If combined these metrics can be applied in multiple ways to improve efficiency and reduce inequalities.

A popular model gaining wide traction for the spatial and functional organisation of neighbourhoods is the concept of a 15- or 20-minute city. A former Paris mayor, Anne Hidalgo, famously made creating a “15-minute city” a key pillar in her re-election campaign in 2020, a policy also strongly supported by the present mayoral incumbent.⁶

Meanwhile the World Health Organisation (WHO) founded a global age friendly city network with over 1300 members which are designed to promote active ageing. Its guide to age friendly cities⁷ covers eight domains, five of which resonate with our research: outdoor spaces, housing, social participation, community support, and transport. Similar dimensions are embedded in the Centre for Better Ageing’s engagement tool kit, which is concerned with the implementation process. ⁸

There is also a string of influential academic publications from then right up to the present exploring the dimensions of age friendliness from different perspectives. Two may be highlighted due to their emphasis on geography. One is the article by Van Hoof et al (2021) which makes the case for proximity to services to keep people fit and connected to opportunities.⁹ The other is Rémillard-Boilard and Doran (2024) who emphasise the need for a ‘spatial lens’ to understand older people’s behaviours and experiences.¹⁰

Another foundational WHO publication is its guide to measuring the age-friendliness of cities in 2015 in which walkability is a key indicator.¹¹ The global reach of the concept of walkable cities and neighbourhoods is therefore undoubted and active experiments are underway in cities around the world. But there is also a deeper history dating back to the garden city movement pioneered by Ebenezer Howard (1850-1828) which is the antithesis of Abercrombie’s 1944 car dominated post-war vision for London (P. Abercrombie, 1879-1957).

Despite the surge in interest in walkable and age friendly cities, there is a shortage of analytical tools and methods to exploit data at the levels of accuracy needed. Some of the work is theoretical and refers to new-builds or smaller scale developments with an architectural emphasis. Suggested metrics are mentioned but are stronger on the ‘what’ than on the ‘how’. We wish to fill this void with spatial-based analytical tools ranging in scale from regions down to household level.¹²¹³

A specific example is cities with hub-spoke arrangements lend themselves to transformations into walkable spaces in which the car is subservient to walking and cycling and pedestrian areas are commonplace. The city of Huddersfield which is part of Kirklees is one such example. Surrounded by a 1km diameter ring road, it is quicker to reach destinations via the ring than going through the city centre which benefits from quiet streets and local attractions.

Spatial accuracy is critical for capturing people’s behaviours and willingness to walk – we argue at scales down to the nearest metre. In the UK, Local Super Output Areas (LSOAs) typically form the basis for applications in planning, public health, education and transport to give a few examples.¹⁴ The average LSOA contains about 1500 people or 640 residential properties and is 6.4 km² in area. However, this scale is far too coarse and insensitive to how people perceive distance.

Walking speeds vary but for an older person 3.42 kms per hour is an often-quoted average which equates to 855 metres per 15-minute journey. At this speed, a typical LSOA could take 51 minutes to traverse diametrically, equating to 2.85 kms. Census Output Areas are smaller but are also very variable in size and shape. The willingness to walk to a bus stop is reported to be only 300 metres for example.

Attempts to measure whether access to services is within 15-minute radius at LSOA level is therefore likely to be approximate at best given the size, shapes, and dimensions of LSOAs. However, evidence shows that even relatively short distances can alter behaviour, not only for older people but also for young families and people with disabilities.¹⁵ The net result is that it is almost impossible to evaluate access accurately.

The argument put forward in this paper is that if we could put a value of location it would lead to better social outcomes and greater efficiency, but for this we need better spatial analytics and accounting methods. Whether a town, city or settlement satisfies the 15-minute criterion requires data with exact x, y coordinates to obtain the required level of accuracy.

It would then be possible to group any points of interest within any natural or administrative set of boundaries and measure the distance/time to walk from any point such as a residential address to any other point such as doctor’s surgery, school, or other often accessed services or facilities. But not all services places of interest need to be nearby, which is why public transport is important.

Proximity to bus services is partial compensation for lack of walking access but only if bus stops are accessible and bus services are frequent. Even so it is important to bear in mind that door-to-door journeys involving buses and trains will almost always exceed 15 minutes once connection and waiting times are considered and so are best reserved for certain types of journeys and functions.

The Ordnance Survey data we use is available for any part of the country which means any of the examples in this paper can be replicated in other locations, whether towns, rural areas, local authorities, or regions.¹⁶ However, applications are enhanced if local data sources are added, drawn from any of the services managed and delivered locally. A good example might be whether people living in council housing stock have better or worse access to a GP on average.

For practical purposes we define an age-friendly neighbourhood as one in which most residential addresses can access Tier 1 services within a defined walking time. We call them Tier 1 services because they tend to be local and are highly frequented and valued by the community. Higher tier services have greater reach and could include major retail centres, transport hubs and so on.

Our approach focuses on convenience of access to daily needs, reducing the need for long journeys and equitable access to everyday services that matter and contribute to well-being. In the longer term it encourages mixed-use development combining residential, commercial, and leisure spaces rather than separating them. It also results in quieter traffic-free environments.

What is included in locally oriented Tier 1 services and is of a suitable walking standard is a matter of definition and is presented below and is informed by local opinion and in some cases regulation. If not walkable, there should be a bus network with accessible stops, supported with cycle paths and pavements to optimise convenience.

The purpose of this paper is fourfold:

1. To evaluate the practicality of measuring whether existing settlement patterns meet the 15-minute criterion using a range of regularly accessed services, classified by function, by the general population and older people.
2. To consider whether individual services satisfy the 15-minute or similar criteria but also individual services where there may be deficits or gaps – for example, out of five services three satisfy the criterion but two do not.
3. Use spatial accounting methods to quantify areas with good or poor access based on the number and location of residential addresses and their proximity to services and provide results in the form of easily interpretable tables and maps.
4. Suggest uses of spatial accounting for the purpose of evaluating service delivery, siting of hospitals and allied services, community hubs, shops and land use and transport planning. Summary measures of access and spatial efficiency are also proposed.

On their own the results could be useful in diverse ways - for example addressing gaps in provision and quantifying unmet access needs across a mixed population and land use. Organisations work independently and locational decisions might overlook the synergy between co-located activities in terms of footfall, and traffic reduction and land management.

We explore other uses which link data from various sources for the purpose of identifying catchments or sub-groups with special needs (e.g., low-income households with below average access to services). With practice, users should find the results helpful and accurate from designating sites for new housing or enumerating gaps or surpluses in current provision.

The list of potential applications is long and could include care homes, nurseries, social care providers, primary schools, places of entertainment, sporting activities, access to green space, and community hubs. The data held in common enables closer collaboration between health and social care, and other sectors such as education, retail, and community safety.

By providing a clear, data-driven picture of community strengths and gaps, the ideas put forward here enables partners to target resources where they will have the greatest effect. This supports more equitable service provision, improved well-being, and stronger, more connected neighbourhoods and localities with clear separation of the strategic from the local.

The methodology informs a wide range of knotty problems which authorities and investors struggle with including policy through to service design:

• Policy: Walkable cities, age-friendly communities, neighbourhood health centres, health needs assessments, transport planning, crime and safety, and service provision.
• Investment: Site selection for older people’s housing, care homes, primary care facilities, community centres, retail centres, service delivery hubs, and repurposing assets such as failing high streets.
• Planning: Locating community assets and infrastructure relative to population needs and environmental constraints, aligned with wider policy aims such as walkability and improved well-being.
• Service design and delivery: Developing spatially efficient service models located close to demand, reducing travel times, and ensuring services are fit for purpose and appropriately targeted.

It follows that there are a range of potential users who will find the methodology useful in their day-to-day work. These include Local authorities responsible for social care, housing, and planning, Government agencies involved in place-based economic planning, new towns, and affordable housing.

Policy documents emphasising the importance of neighbourhoods in fostering wellbeing appear with increasing regularity which together show strong alignment and complementarity. Four are shown in Box1 which demonstrate the direction of policy nationally and internationally.^17181920

The beneficiaries will include both residents and providers through joined up services, information sharing and a commitment to efficiency which builds on a consumer as well as a provider perspective. Spatial efficiency is key, and throughout this paper we give many examples of why location matters and what can be done.

The beneficiaries will include both residents and providers through joined up services, information sharing and a commitment to efficiency which builds on a consumer as well as a provider perspective – whether a service is provided online, the service comes to you, or you go to the service.

Our case study centres on Kirklees in West Yorkshire, which is a member of the Age-Friendly Communities National Network and is affiliated with the World Health Organisation (WHO). This makes it ideally suited as a case study for testing out innovative ideas.

Kirklees has a resident population of just over 440,000 people covering an area of 409 square kilometres. Of these around 79,000 or approximately 18% of the resident population is aged 65+ although their geographical spread varies widely.

The overall age structure of Kirklees is broadly similar to that of England and Wales and has a median age of around 39 years. The population is growing gradually and is projected to continue increasing modestly over the coming decades to around 470,000 by 2040.

Kirklees covers a mixed urban and rural area of approximately 409 square kilometres, with several densely populated towns alongside more sparsely populated countryside. The main towns include Huddersfield, Dewsbury and Batley, alongside smaller settlements, with variation in housing types, terrain, and transport connections.

Kirklees experiences notable levels of deprivation and is ranked 70th among local authority areas in England. Kirklees performs below the national average on several key health indicators, including healthy life expectancy, which is estimated to be 3.2 years shorter compared with the England national average.

Our second area is Stoke-on-Trent which we compare in places with Kirklees on specific metrics. It has a resident population of just over 256,500 people, of which 43,000 or 17% of the resident population is aged 65+. The overall age structure is like that of England and Wales but with lower median age of thirty-eight compared with one year more in Kirklees.

It is smaller in area than Kirklees covering a mixed urban-rural area of 93 square kilometres with pockets of high density. It consists of six townships aligned on a north south axis and includes outlying former coal mining settlements which are part of the north Staffordshire coal field. The population is growing slowly and will reach about 300,000 by 2040.

Stoke-on-Trent is more deprived than Kirklees and is currently ranked 13th most deprived local authority area in England. According to its corporate strategy ²¹ Stoke on Trent also fares badly on several key indicators such as healthy life expectancy which is 6.8 years shorter compared with the England average and 3.6 years less than Kirklees.

Cities are composed of neighbourhoods which would be expected to enjoy similar levels of access to amenities and services. Gaps can be identified using maps in which household addresses are colour-coded and enumerated according to the number of services accessible within a given walk time or radius. In this section we will explain how it works.

We are interested in Tier 1 services used extensively by older people. Examples used in our case studies are doctors’ surgeries, pharmacies, dentists, post offices, and supermarkets. These are among the most accessed facilities by older people but there are others like clubs, places of worship and community hubs which are highly frequented.

Other public facilities are also considered such as libraries, leisure facilities, or green space but may require the use of public transport or other modes of travel. We ignore schools, nurseries and facilities for young people which are out of scope for this application, but they can be easily accommodated using the same techniques.

Ordnance Survey Points of Interest (POI) data provides the locations (x and y co-ordinates, i.e., British National Grid eastings and northings) of businesses, public services, places of worship and many more. For example, there are 81 doctor’s surgeries, 97 pharmacies, 55 dentists, 65 post offices, and 410 supermarkets within the Kirklees boundary. Our task is to compare access to them from residential addresses and by older people.

Tier 2 services are spread more thinly and usually require other modes of transport to reach. They include larger shopping centres, places of entertainment or learning or sport like cinemas, museums or art galleries, swimming pools, and attractions further afield. The list is very long, opening many obvious extensions to the examples presented here.

Every address in the UK has a unique property reference number known as a UPRN which is classed as either residential or non-residential.²² For current purposes we are interested only in residential UPRNs which we use as a proxy for households, although the comparison is not exact as some UPRNs may contain more than one household. There are 205,000 residential UPRNs in Kirklees inside the local authority boundary.

Each UPRN is provided with x and y coordinates which enable them to be mapped alongside other information such as roads, green space and the services contained in Tier 1 or other Tiers. There is no limit to what can be included if they have coordinates to identify their locations. We can also identify areas such as parks or open green space as well as point locations.

For greater accuracy we include services located within the borders of Kirklees plus services within a 4km buffer beyond the boundary to be sure of capturing the access experience of residences straddling neighbouring administrative jurisdictions. It means a service situated outside the city boundary is deemed accessible to an address inside the boundary if it falls within a 15-minute walk-time.

Table 1 shows 22 types of Tiers 1 and 2 assets in Kirklees. Around 40% of accessible Tier 1 assets fall outside Kirklees and shows the importance of looking beyond the confines of local boundaries. After bus stops the most numerous attractions are supermarkets and places of worship. Least in number among Tier 1 assets are libraries and children’s activity centres.

Higher tiers congregate in city centres alongside lower tiers. An analysis of attractions inside Huddersfield’s partly pedestrianised area located within a 1km diameter ring road shows high concentrations of legal, financial, and other commercial services. It includes a large retail presence comprising specialist shops, cafes and eating places, and transport infrastructure.

Together these account for 80% of the 970 Points of Interest located there with public infrastructure accounting for a further 9% and health and education 5%. It turns out that there is hardly any manufacturing whilst sport and entertainment comprise mainly betting shops.

We use a walkability index to evaluate the age friendliness of any area or neighbourhood. Output includes maps, tables, catchment areas, walkability (e.g., how many community centres are within walking radius). We analyse choice among different services and attractions but also isolated areas with no access.

Access is crow fly distance, but actual access may involve circumventing barriers such as a main road or railway. Crow-fly distance is a good approximation as distances are small with access times inferable from walking speeds. Longer distances (e.g., for Tier 2 services, like cinemas or leisure facilities) we can combine walking time with use network travel times on public transport or by car which are purchasable independently.

There are 205k residential properties in Kirklees each shown as a dot in the map in Figure 1. Each is colour-coded based on how many services are accessible in a 15-minute walking radius. Because there are so many they morph into a continuous area in urban areas but not in rural locations.

Five everyday Tier 1 services are included – a supermarket, GP, pharmacy, dentist, and post office. Colour-coded addresses form a continuum ranging from zero to five. If a property can access all five it is cream-coloured yellow; if none is accessible, then it is darkest green. Lighter green is for intermediate cases between two and four services.

As we are interested in older people, red contours show relative concentrations of older people so we can do a deep dive into hotspots. A superimposed grid of 2.5 x 2.5 kms allows locations to be speedily identified, and comparisons made e.g., Denby Dale is J9, while Huddersfield is borderline cells F6 and G6.

Figure 1

Open Figure 1 full screen

The pattern shows that built-up areas have better access to services than rural areas, but some small towns and villages do better. Note that there would be a considerable loss of accuracy if we had used area-based data to show the same effect. This is caused by the averaging that always occurs using area-based data, the varying size and shape of area-based units, and the positioning of centroids from which access times are measured.

In our case the concept can be applied for any chosen access time, individual or grouped points of interest, or for any shape or area, making it useful for summary comparisons or economic appraisals to give two examples of applications. We also know that the shape and size of the underlying statistical units will not distort the findings and affect comparisons.

The availability of assets impacts both on choice as well as nearness – e.g. how many shops are in a 15-minute radius compared with post offices say; why are libraries harder to access than GPs?

Choice allows users to select service providers they value most in a world where quality of service may matter more than convenience. However, more choice for some often results in others having no access or at greater disadvantage.

In Kirklees we find access and choice tend to be correlated. We see this in Figure 2 which shows access to range of typical tier one services ranging from places of worship to pubs – how many have no access, single or multiple access in a 15-minute radius.

A standout finding is that supermarkets are the most numerous among local services for which 84% of household have access to two or more in a 15-minute radius (these are the green pie slices in Figure 2).

Shops beat everything else, but quality can be very variable – how many sell fresh fruit and vegetables for example versus just basic convenience foods and other products?

Places of worship give very similar levels of access as supermarkets and so very good for spiritual health but how good are they for social interaction? In principle they should be major hubs and meeting places but are they under utilised? Community centres also perform this role but significantly 22% of UPRNs have lesser access compared with only 6% for places of worship.

Post offices carry out many Government functions, but the days when pensioners collected their pensions once a week are long gone so queues are shorter. Many post offices have closed in recent years with some services going online so they are now scarcer, but even so 69% have 15-minute access.

This is because they designed to be accessible and do not compete for business in the same way as shops. Despite the geographical constraints, site sharing is an option for spreading convenience. We already see that many post offices are co-located in shops - which is called stacking in the jargon - and is a concept that could be more widely applied to other services.

Public libraries are the fewest among the services we have analysed and there is a strong case for suggesting that they are really Tier 2 services as only 23% of UPRNs are within a 15-minute walking radius and so users must use public or some other form of transport. For this reason, we analyse libraries in more detail later as part of a case study.

One way to overcome library shortage is to create pop-up libraries in cafes, community centres – even telephone kiosks- and this probably occurs at an informal level. The same applies to other services like farmers’ markets which can generate footfall in temporarily used locations on specific days of the week and month.

Table 2 provides a summary for each service above plus dental services for comparison. Additional metrics are included as well as counts of the number and percentages of UPRNs with 15-minute access. Choice is expressed as the percentage of UPRNs with two of more services points within a 15-minute radius.

The results show considerable variation with supermarkets at one extreme giving most choice and libraries giving least. Accessibility is sensitive to the number and distribution of service points whilst improvements in access occur at a diminishing rate as service points increase. Higher densities increase choice proportionately as is evident with supermarkets.

Median time of access gives different information based on the absolute time of access for 50% of all households without reference to any benchmark like 15 or 20 minutes. We note that all are less than 15 minutes except for library services where median access is 26 minutes. It would take at least a doubling in their number to reduce this to 15 minutes.

The IQR or Inter-Quartile Range is the difference between the bottom and top quartile of the distribution of walk times and is a measure of dispersion. Ideally it should be compact to minimise inequalities in access with larger values indicative of greater inequality of access. For example, there is greater dispersion among dentists compared with pharmacies.

The ratio of the IQR to the median indicates the spread of the middle 50% of access times is relative to its central value. A value of less than 1 indicates tighter clustering around the median meaning that access is more homogenous, whereas a value greater than 1 indicates the opposite. Places of worship, pubs and post offices do well on this measure compared with dentists and GPs.

A column to the right of the table ranks each type of services based on its walkability in a 15-minute radius. This gives the interesting result that places of worship are the age friendliest of activities and possibly the least used in this list and supermarkets are the second. Lagging in 7th, 8th and 9th places are GPs, dentists and libraries. Ranking is highly correlated with choice where there is an almost 90% statistical association.

A general rule of thumb is that around 90% of households will have 15-minute access assuming asset densities of 1 per square km, 99% if it 2 per square km and only 70% if it is only one every 2 square kms. This breaks down if the underlying distribution is highly skewed between urban and rural areas, but it suffices in sub-areas where distribution is more even. This may be a useful guide in applications of the 15-minute principle.

Do assets tend to cluster? Our mapping suggests they do but to what degree and which assets cluster more varies. An analysis of meeting venues such as places of worship, community centres, pubs, restaurants and cafes found that two-thirds of all UPRNs had 15-minute access to at least four different meeting places, but 8.8 thousand had access to none or only one.

Although there are more pharmacies than GPs, we find they have similar spatial footprint which is convenient for patients; for example, 75% of pharmacies are located within a 3-minute walking radius of a GP. However, a 15-minute walk or greater is required in the case of 5 GP surgeries. This needs to be set against the finding in Table 2 that 63.8 thousand households are without 15-minute access to a GP and 54.4 thousand to a pharmacy.

We can summarise accessibility to points of interest using the walkability index. This a general measure which calculates the percentage of households or UPRNs within a specified walking radius. In this illustration, the index concerns just health providers, a GP, pharmacist or dentist. Results are presented in chart form as shown in Figure 3 (a) for Kirklees and (b) for Stoke-on-Trent.

It shows the percentage of households or UPRNs within walk times of 5, 10, 15, 20 or 25 minutes. Comparing (a) and (b) we find that access in Stoke-on-Trent is generally better than in Kirklees. For example, based on a 15-minute radius, only 50% of Kirklees households or UPRNs have access to all three providers compared with 55% in Stoke-on-Trent.

This concept can be applied in multiple contexts for groups or asset combinations where proximity to households and assets collectively add to social value and therefore community. A good example is meeting places offering a variety of choice – this could include pubs, cafes, restaurants, community centres, places of worship and others. Other examples might include sport and leisure including parks and green space.

With a little effort we can focus down on specific issues. One of the more important of these is identifying age friendly areas – something Kirklees is keen to do. Each can be a kind of magnet – e.g. for attracting age- appropriate services like care hubs and inward investment like housing, care homes, leisure opportunities or volunteering.

But they need a critical mass to make them attractive for investors for example and a nice environment helps. The map in Figure 4 divides Kirklees into deprived, somewhat deprived and affluent areas – grey, yellow and green based on the IMD (Index of Multiple Deprivation). It is noteworthy that areas in grey tend to fall in more urban areas in the north and centre of Kirklees.

Figure 4

Open Figure 4 full screen

Overlaid are red contours showing 50% or higher concentrations of older people. These areas are mainly small in area – one can think of them as hot spots which have above average concentrations of older people, sometimes known as naturally occurring retirement communities (or NORCs).

The superimposed square grid with each cell 2.5 x 2.5 sq. kms in area gives an age friendliness score based on the number of services that are within walking access to people living there. There are five services in this example as before (shop, a GP, dentist, post office and pharmacy).

A value of 5 in a cell means 100% of households or UPRNs are in a 15-minute radius, a value of 4 means 80% and so on. It also exposes spatial inequalities. In rural areas, the value frequently falls to below one. The map also shows whether older people hotspots occur in more or less deprived areas.

It is evident that the hotspots are widely dispersed and variable in size. The map enables us to ask the question whether older people are more likely to reside in deprived areas and if so in what respect are they disadvantaged, how many live there and are services adequate. Again, the picture appears to be varied.

Which is the most attractive area to live is unsurprisingly a question we are asked many times. A deeper dive reveals that Lindley in cell E5 is particularly attractive for example, but there are others in the suburbs close to Huddersfield. Figure 4 a and b demonstrate that Lindley’s access to services is far better as compared with the whole of Kirklees

It is situated in an affluent area. It has 6,800 residential properties (UPRNs) and is well stocked with shops, cafes pubs and places of worship and essential services like health care. It shows that 92% of households or UPRNs in Lindley have 15-minute access to a pharmacy, 95% to a GP, and 84% to a dentist.

It also has 15-minute access to at least two supermarkets and 55% of addresses to two community centres. Analysis shows it is better provided than the whole of Kirklees which is seen in the chart on the right and is confirmatory evidence that Lindley is age friendlier than the Kirklees average.

Kirklees covers a mixed density area; however, Lindley also scores highly on neighbourliness. The average time of access to the nearest neighbour is only 16 seconds as compared with 2.5 minutes for Kirklees as a whole, which can range as high as 22 minutes in peripheral, more rural areas. Such information can be used for example to map community isolation or as a proxy for vulnerability for persons living alone with health or mobility needs.

Lindley is also the location of Huddersfield Royal Infirmary, the main hospital serving Kirklees, reinforcing its age friendly status. The general point, however, is that the Lindley example is illustrative of how to evaluate age friendliness at both small and large spatial scales and the ability to pinpoint small advantages or disadvantages.

We can see that household access is important, but we can also flip the question to see how it looks from a provider perspective – what are the gaps and opportunities for example?

Earlier in Figure 5 we saw that 31% of UPRNs do not have 15-minute access to GP which could be of concern. We also know that car ownership declines with age disadvantaging older people.

Providing a local neighbourhood health service is the strategic aim of the NHS and we have already compared GPs, pharmacies and dentists. GPs are gateways to other NHS services and remain the first port of call for many health conditions, so where they are located is important.

Although there are several ways to contact a GP including home visits, these are increasingly infrequent, whilst telephone or email tend to be more concerned with appointments, prescriptions and general queries and are no substitute for in person examinations, tests and clinical procedures.

Evidence shows that distance from a GP is a barrier in non-urgent cases, potentially resulting in delayed diagnosis and treatment and worse outcomes. A study in the London Borough of Tower Hamlets found that locating eye tests in GP surgeries resulted in a 16% increase take up in free eye testing for the 65+ population compared with locating them in optometrist outlets.

This simple screening measure saves the NHS much more than it costs and plays especially effective role in prevention of age-related eye disease.

Where GPs locate is hence crucial for matching resources to needs and no more so than in Kirklees. Figure 6 shows their locations including branch practices. UPRNs located within a 15minute walking radius of a practice are coloured green and the rest pink.

Overlaid are contours showing concentrations of old people so that we also can observe the extent to which some are disadvantaged. We find that pink areas straddle remote locations and villages and are understandably more difficult to provide for. Overall, only 68% of UPRNs within a 15-minute walking range of their nearest GP which could be considered concerning.

The result is a mixed picture of very good access in built up areas such as Huddersfield but also much reduced access in others especially in the northeast between cells H4 to K4 where there appear to be gaps and which is replicated in various of the outer suburbs around Huddersfield and in rural areas.

Figure 6

Open Figure 6 full screen

Where GPs locate influences the locations of co-providers of health services like pharmacies. These are remarkably similar in access terms and are sometimes co-located. GP visits may include a trip to a pharmacy to collect a prescription, other shopping and a visit to a cafe. Multi-purpose round trips such as this can also be analysed by this method.

Our walkability index found that 50% of Kirklees households had 15-minute access to a GP, pharmacy and a dentist and 70% to any two of the three providers. Health care must work for everybody but especially the older population. Based on the existing configuration 35% of older people’s addresses do not have 15-minute GP access compared with 30% of under 65 households that do.

In its ten-year plan the NHS talks about developing a ‘neighbourhood health service’ but is silent on what it means exactly – rather it talks vaguely about joined up working. What is needed is a statement of how this will work but this analysis shows that geographic proximity must remain at the core of any solution.

One approach is to produce a comprehensive all-provider configuration including social care building on current assets. Since frailty among the elderly is a key driver of need, households could be graded accordingly to inform the most efficient spatial organisation. A good starting point is to identify candidate locations which would bring the majority of the 35% of households without access to within a 15-minute walking radius.

Catchments measure the reach of a service and the user population located within a given travel radius. Among closely packed services, they tend to overlap and whilst this creates greater choice among UPRNs lucky to be within walking distance, other UPRNs may feel disadvantaged if they have trouble accessing providers not in their vicinity.

Here we use the example of public libraries which are also credited with having a high social value in the community. Benefits include free or largely free services, they support literacy in the local community especially among children, act as meeting places or even provide free office space for home workers and researchers.

With 25 libraries covering an area of 409 sq. kms. they are relatively scarce compared with other local services. Earlier we found that 76% of addresses without 15-minute access compared with 37% in the case of GPs. It means that user access is more likely to involve the use of bus or car. A 30-minute radius is hence a more realistic benchmark in which case we find that 77% of addresses have access but that 23% are still excluded (48,000).

We distinguish between a gross and a net catchment. The first is simply the number of UPRNs located with a 30-minute travel radius on foot while the second adjusts for other UPRNs that fall within the same travel radius of one or more other libraries.

For example, if a UPRN shares two libraries a value of one half a UPRN is assigned to each library, if three libraries then it is a third of a UPRN and so on. This occurs if gross and net catchments are equal and is indicative of service with only limited reach and a well-defined user population and hence a local monopoly of users.

If a gross catchment is much larger than the net catchment it could be an inefficient location that duplicates the services of another library in the same travel radius. By the same token, it may indicate a library that provides a different level and more specialised service as is often the case if the library is in an urban centre – for example, reference facilities such as historical texts.

Figure 7 shows this for public libraries in Kirklees each numbered for later use. Libraries with near identical gross and net catchments include Denby Dale (7), Honley (12), Holmfirth (11), Marsden (18), Meltham (19) and Shepley (23) all of which are in remote or rural locations with small catchments.

We can also see that gross catchments vary in size from a few thousand UPRNs to 20,000 or more. For example, Huddersfield (13) and Kirklees libraries (16) are both located in the centre of Huddersfield itself which benefits from good transport connections and large day-time population.

However, we also see that their net catchments are considerably smaller indicating overlap with other nearby libraries. One exception in an otherwise built-up area is Cleckheaton library which also has a small footprint and hardly any overlap.

Also included in Figure 7 is a tall column next to the origin which shows that over 48,000 UPRNs or 24% have no access to a library within a 30-minute walking radius. This is just 2% less than the percentage with access to 2 or more libraries within the same radius showing an asymmetry between choice on the one hand and having no access on the other.

How might this information be used? One way is to review individual locations to see who does and does not have access. Libraries mapped in Figure 8 use the same numbering as in Figure 7 whilst libraries in neighbouring authorities are shown in the key below the map. The maps show UPRNs having poorer access of greater than 30 minutes in pink and those with access within 30 minutes in green.

Figure 8

Open Figure 8 full screen

It is striking that large areas do not appear to have access. Overall, 25% of UPRNs are outside the radius or 52,000 UPRNs and so worryingly high. Cells like H2 and H3, F6 and F7 and J5 and K5 are notably deficient as are rural locations south of row 8. Further analysis shows that 82% of addresses in older people’s hotspots have 30-minute walking access compared with 73% elsewhere so older people are a bit better off.

One might argue that 30-minute access is an unreasonably high expectation. For this reason, libraries should be classed as Tier 2 rather than Tier 1 service because they are not easily walkable, but this is partly the result of poor location planning. Further analysis also reveals a disparity in 30-minute access between older and younger households – 54% versus 60%

In mitigation Kirklees runs a mobile service in conjunction with the Royal Voluntary Service that delivers books to the door 4-weekly and is open to everybody. This is an example of where the service comes to you rather than you go to the service. The trick is to make sure areas with the least access are prioritised but not ignore the need for equitable access.

It is widely accepted that convenient access to green space is beneficial to well-being and an escape from hot weather. This can deliver opportunities for social interaction, fresh air and health and well being - whether it is exercising the dog, going for a stroll or sitting on a bench chatting with a friend.

Obviously, there is a balance between the size of green space versus its proximity to homes and here we find a major difference between Stoke-on- Trent and Kirklees.

Figure 9 shows that green space in Kirklees divides into many small parcels across the local authority. We do not consider public footpaths or cycle ways, but these are plentiful too.

It also denotes previously identified age friendly areas. Households or UPRNs with 10-minute walking access to green space are coloured grey and those more than 10-minute are purple.

These are typically in rural areas, although public rights of way in the form of footpaths and bridleways will compensate poorer access to a degree. Nevertheless, it turns out that access is very good with 98% of all UPRNs having 10-minute or better walking access.

Figure 9

Open Figure 9 full screen

The contrast with Stoke-on-Trent is striking where only 70.4% of UPRNs have access even though it is proclaimed to be one of the ‘greenest’ cities in the UK. The reason for this is because green space in Kirklees is scattered among numerous public spaces over wide area, whereas in Stoke-on-Trent it is concentrated in large parks.

In Kirklees, good access extends to older people’s hotspots with 99% of UPRNs having 10-minute access compared with 98% generally. However, some areas benefit more than others as Figure 9 shows. There are local solutions available such as greening areas or providing public benches but preferably away from traffic

The difference between Kirklees and Stoke-on-Trent can be seen in Figure 10 which shows that 25% of UPRNs in Kirklees are less than 2.5 minutes from the nearest green space compared with only 10% in Stoke-on-Trent – so over twice the percentage. Overall, this advantage should provide Kirklees with opportunities to improve its ‘green offer’ in imaginative ways.

Kirklees contains challenging terrain which constrains walking speeds and may impact on access to services in some areas. Kirklees borders the Pennines but even within some of the built-up areas walking can be challenging for residents.

We devised a hilliness index adapted from William Naismith’s rule (1892) which adjusts walking speeds for gradient at a property level and can be averaged over small areas as required. We consider it useful for local adjustments to walkability in especially hilly terrain to inform the potential re-siting of hard-to-reach services.

We sampled the elevation of thousands of points in Kirklees and calculated the slope gradients between them. A slope is defined as the change in elevation between two points divided by distance x 100. The results are shown in Table 3 which shows slope gradients at cell level based on the grid used in all the maps with an overall average of 3.4%.

It shows that gradients can vary from a low point of 1% in the northeast (cell L2) to 8.3% in the southwest (cell A10). On average gradients decline from west to east (cols A to L) whereas they are more or less constant from north to south (rows 1 -11). The basic idea would be to assign a small correction to walking speeds to every address.

The rule of thumb is that for every 1% increase in slope gradient walking speed reduces by 0.27 kms per hour. A key finding is that about 56% of the Kirklees area experiences 10% lower average walking speeds based on the index. We suggest therefore that terrain should be factored into planning decisions.

One example of how this can impact a local area is Slaithwaite in cell D7. A picturesque town with approximately 20% of its population aged 65+, it was one of the locations featured in the popular TV comedy ‘Last of the summer wine’.

Slaithwaite scores 3.5 for Tier 1 services based on our age-friendly index (i.e.70% of addresses have 15-minute access) although it does much better on meeting places like pubs and places of worship with 90% access.

However, average gradients in Slaithwaite are 4.8% compared with half that in built up areas. The effect is a lowering walking speeds by around 15% which means access is slightly lower than the headline figure of 70% access would suggest.

Because they are scarcer the ability to access Tier 2 services like leisure centres, museums and cinemas is dependent on other means of travel such as car or public transport.

Proximity bus services are vital for people without cars and those who rely on public transport for work, shopping, secondary school as well as access to Tier 2 facilities. Good access can also reduce motoring costs, congestion, and road accidents.

In Kirklees, the older population (aged 65+) has lower car ownership rates than middle-aged groups, with 22% of older households having no access to a vehicle. Those more likely to be affected are live in rural areas but built-up areas are also affected.

Data for Kirklees show that that the number of concessionary bus journeys by older and disabled people has fallen by around half since 2010, albeit the decline by slightly less than the national trend.

It is not clear whether this means they are travelling less or are using other means of travel, but data show a big drop during the COVID pandemic which has not yet recovered to previous levels of usage.

The issue we consider here is whether older people are disadvantaged from using buses because of where they live, whilst acknowledging that patronage can depend on the level of service provided and not just nearness to bus stops.

Reliability and accessibility are extremely important. Even with good service any journey will always take more than 15 minutes, after taking walking and waiting time into consideration.

We are interested in access to bus stops according to where users live, which makes a difference to people’s mobility. Various sources recommend that they should not be more than 250 metres and never exceed 400metres and this is the benchmark we use here.

We consider the extent to which this rule of thumb is met in Kirklees, whether older people are disadvantaged and whether there are bus stop deserts that could be easily remedied.

With 3,400 bus stops in Kirklees, coverage is extensive, as the map in Figure 11 map shows. Stops are shown as square blue symbols which snake their way throughout Kirklees and into adjacent LAs.

Figure 11

Open Figure 11 full screen

Our analysis finds that bus stop access is better in older people’s hotspots (shown as red contours) than in other areas (93.5% versus 91.4%), and so not by very much.

Whilst it is to be expected that UPRNs in isolated areas will have more limited access to bus routes, close inspection reveals bus-stop deserts. These occur as dark pink clusters of properties in built up areas whereas in rural areas they are isolated points.

Examples of clusters are found in cells like J3, and J4 near Bately, and E8 around Mirfield. We conclude that this map and the accompanying analysis are likely to be a useful tool for conversations with local groups and transport providers to see what can be done to improve access in the areas affected.

Bus stops should be accessible and close to where people live but also to destinations, whether it is a library, leisure centres, and other attractions, as an analysis of visitor access to care homes shows.

Long-term residential, or nursing care accounts for about half of total support among people aged sixty-five and over.²⁵ It turns out that care homes are more common than GP practices and are closer on average to where people live.

As a result, walkable access to a care home is good by comparison to some other services. This benefits care home residents that are located close to where they used to live or to their relatives also living locally, although not if they must travel by bus.

Our analysis found that of the care homes, only nine had no bus stop within 250ms and 3 had only one. All others had two or more bus stops from which to choose.

The least accessible were in rural areas. However, we also find that a slightly greater percentage of care homes are located within older people’s hotspots and that 15-minute access to care homes is 76.5% compared with 73.1%.

We have demonstrated how the concept of 15-minute communities can be brought to life, evaluated and how individual services match up. This required new spatial accounting methods and previously unavailable metrics. This is thanks to the use of address level data which provides more options in terms of spatial accounting and completely flexible geography.

A range of applications demonstrated how to identify gaps and opportunities which could be identified with greater precision, whilst the tiering of services provided a basis for measuring age friendliness from an address level upwards. Methods allowed the matching to other data sources such as demographic information to research efficiencies and inequalities by area and sub-groups rather than a silo-based approach.

We are confident that the method appears to do what it was meant to in terms of proof of concept. We are not aware of other similar studies and so comparing Kirklees and Stoke on Trent with other areas of the country will be fascinating. A key benefit is that it forces us to review previously hidden detail at both a micro and macro level including inequalities.

A major difference with existing methods is that these rely on area-based analysis using ONS data and pre-ordained boundaries. In our approach, the user has greater control over granularity through their ability to focus on point data rather than areas. Results can be aggregated into areas of any size or shape from postcodes to the entire country.

A second advantage is that the analysis is based on publicly accessible Ordnance Survey data. Unlike ONS, OS data provides information on points of interest, including everything from shops, transport infrastructure, health providers to industry. This makes possible studies linking people, places and transport which is useful in planning, place making or walkability analyses.

A third advantage is the data are linkable to other sources of data owned and maintained by public authorities, businesses, voluntary sector and survey-based data. For example, we can use data matching techniques to link activities across organisation and administrative boundaries (subject to data protection considerations).

A fourth advantage is that we can combine address level data with ONS data at output area level. These are the smallest geographical units used by ONS to collect and publish census data, ranging in size from 40 to 250 UPRNs. This enables us to focus on specific age groups, household types and poverty for example with reasonable accuracy.

A fifth and promising application is putting social value on proximity to services. This made possible and potentially more accurate using x, y co-ordinates rather than area-based data and is largely undeveloped as a result. For example, studies indicate that bus stops beyond 400 m are perceived as too distant with a consequent drop off in usage by older people who walk more slowly and may have mobility problems.

More generally, the scoring of local areas for their age friendliness is new. It is helpful both to both current and prospective residents who will benefit from knowing that neighbourhoods are well provided with walkable Tier 1 services and access to green space, bus services and green space. The concepts are extendable in other ways, for example to analyses of family friendly or mixed generational areas.

The use of organisation-specific data extends the range of possibilities - in publicly owned and operated services like adult social care, in commercial applications such building homes and job creation, or supporting third sector organisations. Greater granularity means more accuracy in uses ranging from small interventions such as flexing GP opening hours to the re-development of town centres or re-generation projects.

A workshop involving officials in the planning directorate, social care, policy, public health and housing were asked for their views on potential short- and long-term applications in health, housing, library services and public transport. This produced forty-four different suggestions and twenty-six extensions to other community assets like public toilets, and applying the methodology to cohorts including children and families, the neuro-divergent and even pets.

Suggestions ranged from installing seating and greening streets to make local areas and natural meeting places age friendlier, the re-development of town centres like Dewsbury, and planning for the impact of climate change. Among other suggestions were working more strategically with the voluntary and private sectors, access to leisure, sport and meeting places and helping older people to stay in work if they want to.

Housing received a lot of attention including:

• Using the methodology to inform housing need and allocation of sites for housing development.
• When working up specifications for new housing sites, identifying priorities for new services, design features, etc.
• A better understanding of the reasons for older tenants not downsizing in the context of the age-friendliness of where they live
• Overlaying age friendly data with other useful data e.g. on the adaptability of homes for later living
• Matching UPRN data to social housing stock and resident ages to analyse age friendliness and scope for improvement
• Generally, supporting local plans and housing policy

In a related development using property listings, we used artificial intelligence to rate properties based on their location and physical characteristics such as size, whether the have step free access, a garden and access to local amenities like a bus-stop or GP, and value for money including Council Tax and Stanp Duty.

Intended for use by individuals such as last time buyers the App is also being used as a tool in conversations with social care clients where downsizing is an option. There are parallel commercial applications such as site identification for investment in retirement housing or villages, nursing and residential care where there is advantage in co-location with existing services and infra structure.

Whilst the concept of a 15-minute city or its equivalent at town or neighbourhood levels is compelling, we found that necessary metrics and analytical tools were lacking, such as how to measure age friendliness. We also found an under-appreciation of location as a driving force and contributor to sustainable communities, greater well-being and reduced inequalities.

In part we believe this is due to the limitations of how national statistics are reported, how planning works, and data that lack granularity or are out of date. For example, traditional research into poverty emphasises characteristics such as low income, poor housing and health, but never considers the public good and wider benefits of an advantageous location such as living near green space or health provider.

Our approach using property level data and points of interest can provide previously inaccessible levels of intelligence to assist in locating new council stock, filling service gaps, making better use of green space, improving public transport or making better use of dormant assets. Expertise is needed in geographical information systems, but analytical methods are the glue that holds everything together.

We appreciate that not all users are focused on the same things. There is a difference between publicly accessible data and data owned by organisation and between personal identifiable data and statistical data. Recognising this difference makes it easier to share data across different geographies and organisational boundaries. Additionally, knowing how people behave faced with alternative choices is where surveys can help.

The resulting data base captures activity across a range of local authority functions, enabling more strategic and joined up working (Figure 12). A practical question concerns how to use the tools to make a difference rather than just listing users in different sectors like health, housing education, transport and planning.

Table 4 is intended as guidance on how to get the best from this research and how to position it to inform policy and bring about change. It divides applications into four use classes: The first two are about strategy and operational effectiveness; the last two are about improvements well-being and partnership working, both the private and wider public sectors.

Use classes A and B focus on the major drivers which include demographic change, housing and the local economy. Key outcomes are alleviating poverty, regeneration and improvement to the coverage and efficacy of key services and the environment.

Use class C is concerned with improvements to well-being which is a core element of Kirklees vision for the future, which is to be a district that combines a strong, sustainable economy with thriving communities, healthier lives, growing businesses, high prosperity and low inequality.

Concepts such as social value apply here which is concerned the with the broader positive of investment decisions and age friendliness. Evidence shows that age friendlier communities which are safe, well-connected and sustainable, leads to longer, healthier lives which have a social value.

Use class D is about leveraging resources through closer working with neighbouring authorities and sharing services and with business. Better transport services are key for connecting homes higher tier activities like leisure, retail hospital or transport and employment hubs are examples of this.

Table 4, final column, is intended to provide general principles rather than specific recommendations. It is important that access is golden thread that runs through feasibility studies, in which regeneration and alleviation of poverty.

One way to do this is create geography of age friendliness which is used as a benchmark for mapping people to services and vice-versa. Such an index would be helpful for identifying service gaps, be incorporated into housing needs assessments, investment profiles used by business and transport providers as we have done here.

The same principles can be extended to health data sets for use in resource allocation, public health initiatives and others through better targeting of health promotion, GP opening hours and access to urgent care.

Adult social care is also a beneficiary if data are shared between primary care and care providers to enable better joined up working.

At a macro level, planners will be interested in applications in for example zoning and regeneration and environmental projects at a strategic scale but also micro applications such as siting of public toilets or bus stops, the creation of quiet areas with street furniture, lighting and green space, traffic calming, and speed limits.

Running through the research there has been a presumption the localisation of services is age friendlier. Overall demand for travel has stabilised over time and fell markedly during the pandemic. More people working from home, and online shopping is ever more popular, or other factors. This suggests localisation is happening naturally and that society needs to adjust.

The main message from this research is that fits neatly with the grain of current policies in health, planning, transport and education, and housing including building new towns. By using the same sources and combining data on people and places at an address it contributes to joint working across organisation boundaries. This stands it apart from present methods because enables fresh perspectives and joined up working.