Transport Affordability vs. Housing Affordability: An Indicator to Highlight the Economic Efficiency of Sustainable Modes of Transport

Abstract

Background: The rising costs in the metropolitan real estate market are compelling individuals to relocate outside of the city. The anticipated financial savings, however, may be undermined by long and costly commutes; raising the question of whether this trade-off is a worthwhile proposition. This paper uses a digital model of workplace commutes, income levels and house prices in England as well as Wales, to evaluate the trade-off between (i) moving to the city centre and cycling to work versus (ii) continuing to commute by car from a residence on the periphery. Methods: An indicator has been introduced that unifies the transport and housing affordability by expanding the concept of the ‘effective speed’ to include housing costs. The effective speed itself is commonly defined as the travel distance divided by the time dedicated to the transport activity (i.e., travel duration and time given to earn the money to pay for the costs incurred). Results: If only the associated fuel and mortgage costs are considered, residing on the periphery can—for those already living there—be a cost-effective option specially in cities like Cambridge and Oxford. Yet, accounting for the total ownership costs of cars or external effects, this advantage shifts in favour of relocating to the city centre. Conclusion: This study does not negate the existence of an affordable housing crisis in urban environments, though it demonstrates that strategies to cut transport emissions can produce economic gains.

1. Introduction

In response to soaring housing costs, urban residents are increasingly relocating to suburban or rural regions [1]. While the initial move may seem a choice of economy, factoring in the commute burden—both in terms of time and costs—reduces the net benefit of the urban exodus. The journey to work is, in academic circles, widely regarded as the most protracted and expensive trip of the day (e.g., in United States [2]). Commuters themselves, however, typically underestimate the true economic costs of owning and operating a car [3,4,5]. While fuel expenses are generally estimated with a reasonable level of accuracy, this does not apply to depreciation, repairs, tax and insurance, which are commonly undervalued (based on a survey of 6233 car owners in Germany [6]). Their findings suggest that the real cost is nearly twice the amount the participants believe they are paying. Awareness of the true costs, the authors argue, could decrease the number of cars on the road in Germany by 37% [6]. With higher mortgage sales being observed in regions with longer commutes and no public transport, it seems like suburban living might undermine the financial viability of homeownership for some [7]. While American suburbs were initially characterised by a significant working-class community, nowadays, suburbs are home to low and low-moderate income households [8]. Also, the growth of impoverished households is most pronounced in a suburban setting [8].

Beyond the physical journey, individuals also spend time at work to finance their daily commute—often an invisible time commitment for the drivers themselves [9]. Once both temporal components are combined, the perceived speed advantage of cars is notably diminished. This applies, in particular, to the less affluent as illustrated in Meira et al. [4], Ivan Illich [10], Kifer [11], Tranter [9,12,13], Litmann [14,15] and Schnieder [16].

Apart from the temporal and monetary burden for the driver, lengthy commutes are also linked to adverse effects on health and wellbeing [17] including stress and tiredness [18] as well as increased strain, impaired mental welfare and decreasing life satisfaction [19]. Based on a sample of 26,000 employees in England, Clark et al. [19] determined that a 10-min increase in commuting time reduces job satisfaction equivalent to a 19% decrease in gross income. The adverse effect of commuting is not only reflected in self-reported data, but also in physiological indicators such as elevated stress biomarkers (e.g., cortisol and blood pressure) [18], lower physical activity and cardiorespiratory fitness (CRF), as well as increased BMI, waist circumference and blood pressure [20]. Commuting by bicycle, on the other hand, has shown to promote favourable physiological and psychological health benefits [17] including higher mental health scores and reduced absence for sickness [21].

Commutes are not only affecting the drivers themselves. Traffic congestion imposes a substantial cost to the British economy [22], accounting to £7.7 billion in 2024, according to INRIX [23]. Harmful pollutants [24], and significant externalities (i.e., external costs) [25] are also generated. Commuting also plays a role in widening social inequality [26].

While the evidence clearly favours active commuting over lengthy car journeys, the high urban housing costs limit who can ‘afford’ to live close enough to cycle to work [27]. Traditionally, measures of housing affordability don’t include transport affordability (i.e., the cost and availability of transport options in a specific location) [28]. Considering one without the other is problematic given that both represent comparable portions of household expenditure [29]. Both are also interrelated, with higher housing costs typically coincide with lower commuting expenses [29]. A case study of Melbourne, Australia by Saberi et al. [7] confirmed a negative correlation between the two; prompting calls for combined affordability assessments [30].

In summary, although housing in the urban core is becoming unaffordable for many individuals [27], motorists tend to underestimate the financial implications of their lengthy suburban commute [6] and are not held accountable for the external costs their journeys impose on society [31]. This prompts the question of whether living in a city and cycling to work is truly as unaffordable as it may appear once the full financial implication of suburban car commuting is accounted for. To address this question, the paper extends the concept of the effective speed to include housing costs; producing a single indicator that captures both housing and transport affordability. In view of the large disparities in commuting distances, the proposed measure only reports the time expenditure (i.e., travel and earning the income to cover both transport and housing costs). A digital model of England and Wales has been utilised to illustrate the proposed measure; showing where individuals would reside—seeking to minimise their expenditure and time effort. In short, the study addresses the following research objectives:

RO 1: Development of a measure, based on the effective speed concept, that accounts for both the time required (i) to travel and (ii) to earn the income necessary to cover housing and commuting expenses.

RO 2: Visualisation of the trade-offs between (i) a low-cost suburban house combined with long commutes versus (ii) living closer to work and cycling.

By fulfilling these research objectives, the study introduces a new perspective on housing and transport affordability indicators. The insights from this study may challenge the initial assumption about the affordability of low-cost housing outside the city. The study further highlights that residing and working within a 15-min city may be more economically viable than it first appears, while not disputing the presence of an affordable housing shortage in city centres. Instead, the study emphasises the need for greater access to central urban living. Further, the study highlights that the much-needed strategies to reduce emissions of the transport sector do not inherently increase costs. Instead, environmental sustainability and economic efficiency can be achieved simultaneously.

2. Background

2.1. Effective Speed Concept

The concept that is now known as the effective speed is often traced back to Henry David Thoreau’s reflections in the book ‘Walden’ from 1854 [9]. He argued that for travellers of modest means, railway travel had limited benefits. The time spent working on the field to pay for the fare (i.e., 1 day) negates any temporal advantage given that walking to the nearby town would take roughly the same duration, or be even faster [12].

Drawing inspiration from Thoreau, Ivan Illich [10] translated this concept into the context of the 20th century automobile dependence. He estimated that the average American allocates approximately four hours per day to travel by car (i.e., driving or working to finance it). This equates to over 1600 h annually to cover 7500 miles—an effective speed of below five miles per hour.

The German Sociologist D. Seifried is another widely referenced contributor to this discourse. He coined the term ‘social speed’, which incorporates external costs (i.e., paid by society) alongside private costs (i.e., paid by the user) [5,9,12,32]. This framework let him conclude that bicycles, at 14 km/h, possess a similar social speed to cars at 18 km/h [33].

Tranter’s extensive body of work on the effective speed features Australian case studies [9,12], as well as cities around the world [13]. His work encompasses both commuting to work and ‘school run’ trips (i.e., dropping off children [32]). Across most of these studies, he concluded that automobiles typically do not provide the time efficiency once assumed.

Subsequent studies, conducted by Litmann [14], Meira et al. [4] and Schnieder [16], reached similar conclusions.

2.2. Effective Accessibility

Vale et al. [34] introduced the concept ‘effective accessibility’, which incorporates both travel duration and the time required to earn the financial resources necessary to pay for the commute. Their Lisbon-based case study highlighted that low-income car drivers experience an effective accessibility of zero due to the need to invest more than 20% of their income/working hours to finance their journeys. They further argued that the ‘effective accessibility’ is a superior indicator to the ‘time-based accessibility’. The latter produced an up to threefold overestimation.

2.3. Housing and Transport Affordability as Well as Transport Accessibility Indicators

Transport accessibility, defined as the ease of reaching destinations [35], is often measured as a function of time or distance [34]. While this may be justifiable due to the importance of once’s time [34], the failure to account for the cost of transport overestimates accessibility [34]. Others criticise supply-oriented indicators that measure access to transport infrastructure instead of access to locations [36].

Transport affordability is commonly defined as (i) the degree to which transport is a financial burden, (ii) whether affording to travel requires sacrifices, or (iii) whether people can travel as they need to [37]. These quantifications are often compared to a specific threshold (e.g., 20% of the household income) [37]. Apart from any such threshold being of an arbitrary nature [38], a common criticism of such definitions is that the focus on transport expenditure may capture only a partial picture of a household’s welfare [38].

Housing affordability is broadly defined as the ability to fund appropriate housing without undue economic hardship [7]. The traditional definition (i.e., the balance between household income and the costs of their housing [7,28,39]) has come under scrutiny [40]. Various authors, including Mattingly et al. [39] and Saberi et al. [7], criticised these ‘one-dimensional’ indicators for their omission of transport costs. The Center for Neighborhood Technology in the USA has been credited as the originator of an indicator that combines housing and transport affordability, known as the H + T index [7]. Various authors, including Mattingly et al. [39] and Saberi et al. [7] have subsequently developed or applied similar indicators.

Others opposed the sole focus on costs that are currently being paid, arguing that a location can have non-monetary effects on someone’s life or cause opportunity costs. Mulliner et al. [40] outlined various recommendations and critiques in their literature review including: (i) inclusion of factors beyond monetary ones, (ii) consideration of the housing condition/quality and neighbourhood characteristics, (iii) over-crowdedness, insecure tenures, unsafe or inaccessible locations, (iv) including opportunity costs due to housing locations (e.g., safety of the area, job opportunities, schools), (v) access to services and facilities as well as (vi) energy costs. Therefore, Mulliner et al. [40] advocated for a more holistic definition of housing affordability going beyond financial considerations. In their survey of 337 housing and planning professionals in the UK, economic criteria (e.g., house/rental costs in relation to income, interest rates and mortgage availability) as well as the availability of rented accommodation were ranked amongst the most important factors. Other higher-ranked indicators include ‘quality of housing’, ‘access to employment’, ‘energy efficiency of housing’, ‘availability of low-cost home ownership products’, ‘access to good quality schools’, ‘access to public transport’, and ‘access to health services’.

Other authors argued that the combined housing and transport indicator is flawed due to the overlooking of the travel time [41].

2.4. Summary

In short, this paper expands and combines various indicators to provide a holistic assessment of transport and housing affordability. While some may still quantify housing affordability as the fraction of income spent on housing, numerous scholars have advocated the use of a combined measure: the fraction of income spent on housing and transport combined (i.e., H + T index).

Job accessibility, which is usually defined as a function of time or distance [34], has already been revised by Vale et al. [34] to account for the time taken to earn the money to pay for the commute. This idea is based on the effective speed concept, which updates the average speed to include the time required to earn the money to pay for the commute [12].

Considering the importance of one’s time [34], this paper first converts the H + T index into the time taken to earn the money to pay for both housing and transport. Then the travel duration is added. The resulting indicator can be specified as follows:

$$D= {\displaystyle \frac{H+T}{W}}+C$$

(1)

where,

$D$ Time duration devoted to commuting and housing (per month)

$H$ Cost of housing (per month)

$T$ Cost of transport (per month)

$W$ Hourly wage

$C$ Time spent commuting (per month)

3. Materials and Methods

3.1. Home and Place of Work

To construct a dataset of workplaces and homes, this study follows the same approach as described in Schnieder [42]—which itself builds on the work of Kelly et al. [43] and Schnieder [44]. The latter evaluated various parcel delivery innovations, while the other two focused on remote working hubs and co-working spaces. The 2011 dataset ‘location of usual residence and place of work by method of travel to work’ was utilised in this study. It specifies for each combination of Middle Layer Super Output Areas (MSOA) the number of people who commute between both. The dataset was provided by the UK Data Service (https://statistics.ukdataservice.ac.uk/dataset/wu03ew-2011-msoamsoa-location-usual-residence-and-place-work-method-travel-work, accessed on 16 December 2023). An MSOA is a medium-sized geographical unit. It combines four or five Lower Layer Super Output Areas (LSOAs). Between 2000 and 6000 households reside in each of the over 7000 MSOAs in England and Wales [45]. Anyone who (i) doesn’t commute by car or van, (ii) works at/from home or at an offshore installation, (iv) doesn’t have a fixed place of work, or (v) works outside of the UK, was removed.

Every individual in the previously mentioned dataset has been allocated one office location within their work MSOA and two residential locations—one within their current home MSOA, and one within their work MSOA. The latter is the alternative location they could reside at in order to avoid commuting by car. The residential locations within the MSOAs were determined using the shapefile of the MSOA boundaries (https://statistics.ukdataservice.ac.uk/dataset/2011-census-geography-boundaries-middle-layer-super-output-areas-and-intermediate-zones, accessed on 16 December 2023), and the population density data (aggregated to arc-second blocks) from the Humanitarian Data Exchange (https://data.humdata.org/dataset/united-kingdom-high-resolution-population-density-maps-demographic-estimates, accessed on 16 December 2023).

The coordinates for everyone’s residence—current and alternative—were determined by randomly selecting an arc-second block within their home/work MSOA using the population density within each MSOA as a weight. Duplicate selection of the same arc-second block was allowed if the population density warrants this. A single location within each MSOA was chosen as the place of work for everyone working within the MSOA using the method previously described. A variety of libraries in Python 3.13 were used including Pandas [46], GeoPandas [47], matplotlib [48], seaborn [49], NumPy [50], Shaply [51] and SciPy [52].

3.2. Routing

A locally hosted Open Source Routing Machine [53] with the free-flow street network from OpenStreetMap [54] was utilised to calculate the shortest round-trip route between work and home locations, using the car and bike profiles as required.

3.3. Cost of Housing

The cost of housing was estimated based on the dataset ‘House Price per Square Metre in England and Wales’ (https://data.london.gov.uk/dataset/house-price-per-square-metre-in-england-and-wales/, accessed: 15 October 2025). This dataset matches the Land Registry’s Price Paid Data (LR-PPD) with the property size information from Domestic Energy Performance Certificates (EPC). The dataset comprises of over 22 million transactions (i.e., sales of homes) in England and Wales, collected from 1995 to the end of 2024. Only those taking place since 2018 were considered in this study to minimise temporal effects like inflation or world events. To account for property value gains over time, the per square metre prices were adjusted to 2024 levels, based on the average annual house price data from England and Wales, provided by Statista [55]. The sales were then aggregated according to postcodes. The house price for each arc second block (i.e., the location where individuals live) were defined by averaging the prices per square metre of the three geographically nearest postcodes. Since the value for a house is not solely determined by its location, this step mitigates potential outliers through data aggregation. Note: In the UK, there are on average only 15 houses per postcode [56].

The price per square metre was converted into monthly mortgage payments assuming a 40.9 square metre floorspace per person and a 25-year mortgage duration with a 4% interest rate—the current interest rate set by the Bank of England [57]. 40.9 square metres is the average usable floor space of all dwellings in England (i.e., 96 m²) [58] divided by the average household size in the UK (i.e., 2.35 residents per household [59]). This study assumed that households acquire property as opposed to renting, reflecting the 2023 owner-occupation trends in England, where most households (65%) live in owner-occupied dwellings [58].

3.4. Selection of Towns and Cities

The dataset titled ‘towns and cities, characteristics of built-up areas, England and Wales: Census 2021’ was used to select suitable cities and towns for this study (https://www.ons.gov.uk/peoplepopulationandcommunity/housing/datasets/townsandcitiescharacteristicsofbuiltupareasenglandandwalescensus2021, accessed: 1 October 2025). Cities and towns classed as ‘large’ or ‘major’ in England and Wales have been included in the list of potential case studies, alongside two further cities from Wales. London has been removed due to the large scale of commutes.

Only those employees who currently live on the periphery of a city’s limits and then commute to its core have been included. Due to the significant variation in the size of cities, no universal distance-based threshold could be used to differentiate between those who lives within or outside the city. Hence, each city’s diameter has been calculated using the ‘Built Up Areas (December 2022) Boundaries GB BGG’ GeoJSON file (https://open-geography-portalx-ons.hub.arcgis.com/datasets/ons::built-up-areas-december-2022-boundaries-gb-bgg/explore?location=53.410935%2C-1.394874%2C11.84, accessed: 15 October 2025) as well as Python libraries including GeoPandas [47]. For selected cities, that are part of metropolitan areas, the calculated diameter was manually checked using QGIS 3.40 and OSM data.

This study focuses on individuals who work in the city centre but live on the periphery with lower housing costs. For each city, those employees who live within 50 km from the centre, but further than 3 km plus a third of the cities’ diameter from the core, have been included. Obviously, cities are not perfectly shaped in a circular pattern, and the house prices may not always decrease proportionally to the distance from the city centre. Other cities in proximity might influence the house prices and certain neighbourhoods might be attractive for other reasons (e.g., school catchment area). Employees, who live in a more expensive neighbourhood than the city’s core, don’t live there due to the cheaper housing costs and were therefore removed. Various cities needed to be withdrawn after this step due to the low number of individuals who fulfil all conditions (i) currently living on the periphery, (ii) commute to the city centre, (iii) live in an area where the housing costs are cheaper than in the city centre. This was especially true for cities near London, as many people who live in the southeast work within London instead of their nearest city centre. Also, when cities are in close proximity to one another it can be observed that individuals may live in one and work in another.

The alternative location, that individuals could relocate to, if they want to avoid commuting by car, was defined as a circle around each city centre with a radius of 3 km, or a third of the city’s diameter—whichever was smaller. The maximum radius ensured that the travel distance is reasonably short to allow most people to cycle comfortably.

3.5. Scenarios and Cost of the Commute

Scenario one visualises the situation where a car is required regardless of the commute and therefore only the fuel costs were considered. In the second scenario, the full cost of car ownership was considered—split into a fixed and a variable part. The third scenario includes the external costs of the commute.

The fuel costs were estimated based on the ‘weekly road fuel price’ statistic, which provides the average UK retail (‘pump’) prices. The 2024 aggregated fuel price was 1.41/L for petrol and 1.48/L for diesel [60]. Combined with the average fuel consumption for petrol and diesel cars for the year 2020 [61] a fuel cost of £0.075/km was assumed in this study—to reflect a cost at the lower end.

An approximate value of £8.87 per day plus £0.16 per km was determined for the total cost of ownership, guided by (i) the dataset ‘household expenditure on motoring for households owning a car, UK: financial year ending 2021’ [62], (ii) the average yearly mileage [63] and (ii) the advisory fuel rates in the UK [64]. UK households owning a car spend on average £88.40 per week on motoring—ranging from £42 to £156 per week for the lowest and highest decile income groups, respectively [62].

The costs to ride a bicycle reported in the academic literature range from 1€ per day [65] to 475 € per year [34]. £40 per month was used as an approximation.

The external costs for driving a car, compiled by the European Commission, were adopted in this evaluation [66]. No external costs for bicycles were included given that in the academic literature, some authors report external costs (e.g., 0.0672€/km [67]), while others report external benefits (e.g., €0.18 per kilometre [68]) and others report both (USD −1.00~1.95/km [69]).

3.6. Time Devoted to Housing and Commuting (Effective Speed)

The time devoted to housing and commuting was calculated by converting the expenditure on housing and commuting into the time required to earn the money to fund both. This time commitment is then added to the travel duration as illustrated in equation (1) in Section 2.4. Various hourly wages were considered to represent different income levels.

3.7. Limitations

While the 2011 data may appear outdated, it was chosen because working from home, or hybrid working, wasn’t common at that time (10.3% in 2011 [70] vs. over 38% in April 2025 [71]). In the 2011 data, the home and work location correspond to where people choose to live if they work in an office. Using commuting data from the subsequent census conducted 10 years later, during the COVID-19 pandemic, appears even less constructive or relevant. The outdated nature of the commuting dataset, due to the increased proliferation of working from home, does not affect the results of this study. This is due to the study only including those who commute to the city centre from a less expensive peripheral neighbourhood. The study is relevant for the contemporary world given that all cost assumptions are up to date and the majority of residential homes included in this study likely still exist. The growing prevalence of working from home doesn’t change the conclusions drawn in this study. Only the population for whom the results are relevant is being reduced.

If housing costs are compared across wider regions, differences in the interest rates and housing-related costs such as energy price/consumption or property tax may affect the financial affordability of homes. However, this factor was deemed irrelevant due to the proximity of individual’s current and alternative housing options.

Furthermore, to compare housing affordability beyond the specific focus of this study, it is important to subtract the acquired equity from the housing costs. Households may treat their home as an investment and therefore might opt for elevated mortgage payments instead of directing their capital towards other investment opportunities.

This study doesn’t include utilities or similar costs as these are assumed to be the same given the proximity and similarity of the two housing options compared. In any other circumstances, the cost of basic utilities should be considered [7].

4. Results

4.1. All Cities

4.1.1. Scenario 1: Fuel Costs

When only considering the fuel costs and the monthly mortgage payments, it becomes apparent that many individuals live in locations that are equally cost and time effective for their situation. Assuming a £10 hourly wage, continuing to live on the periphery is cost and time effective for 75% to 94% of the individuals in cities like Bath, Cambridge and Oxford. Such a high share is not unexpected: since fuel expenses are highly visible to the driver at the fuel station, it may be—consciously or subconsciously—factored into residential decisions. Conversely, in cities such as Aylesbury, Basingstoke, Blackpool, Derby, Leicester, Lincoln, Peterborough and Preston fewer than 10% should continue to live outside the city—if only the fuel and mortgage costs are considered. This suggests that the decision to live on the outskirts may be motivated by factors beyond any financial considerations.

Generally, as income increases, fewer individuals would be expected to have any time advantage by living outside the city or town. They need to spend less time on earning financial resources while the commute duration remains unchanged by their increased income (Figure 1).

Note: This study says nothing about the share of people who should live within or outside of the city. The study only focuses on those who currently live on the periphery in less expensive accommodation than found in the city or town centre.

4.1.2. Scenario 2: Total Cost of Car Ownership

Once the total cost of car ownership is considered, there are only Cambridge and Oxford left, where there is any relevant share of people who should continue to live on the outskirts, instead of moving into the city centre (i.e., 13% and 7%, respectively, for a £10 per hour income). This share has diminished to less than 4% in Guildford and Bath for the same income group. Although the cities with the highest share of people who should continue to live on the outskirts are well known for their high house prices [72], it is not the only influencing factor. Even if the total cost of ownership is considered, commuting is—for a very small few residing near specific cities—the most time and cost-effective option. For all other cities, it’s either nil, or only a few percent, of people who save money by residing on the periphery. In short, while this study does identify some individuals who are currently living in a less expensive area and should continue to commute, the majority would benefit from moving to the city centre. However, any increase in demand for city housing is likely to exert upward pressure on house prices there. This, in turn, will shift the equilibrium in favour of suburban living.

As shown in Figure 2, the potential decreases in hours spent commuting and working are undoubtedly considerable in many cities.

4.1.3. Scenario 3: External Costs

Although individuals may not altogether appreciate the external costs their commute imposes on society, policymakers ought to be aware. As before, only Oxford and Cambridge have any notable share (i.e., 4% and 7%, income: £10 per hour) of individuals left that save money by maintaining a home on the city outskirts. This share accounts for less than 2% in Guildford and Bath. For all other cities, virtually no-one should stay on the cities’ periphery and commute by car if the external costs are factored in alongside all other expenses—conditional on current prices (Figure 3).

4.2. Geographic Distribution

4.2.1. Scenario 1: Fuel Costs

Figure 4 illustrates those who currently live in lower-cost residential housing in the suburbs and commute to the city centre. Marked in blue are those who should move to the centre. Those for whom it is more cost and time effective to remain on the outskirts are marked in red. Even with the lower housing costs, this figure clearly highlights that extra-long car commutes from towns such as Hitchin, Bishops Stortford and Stevenage are not worthwhile from a cost and time perspective. The fact that the individuals marked in blue (i.e., should move) are more predominant in the south of the urban area can easily be explained by the proximity to London. The cost of housing isn’t low enough to make the commute worthwhile compared to cities further north like Peterborough and Huntington. The same also explains the concentration of blue dots in the southeast corner of the map of Oxford.

Note: Despite being populated, some areas are not part of this study as their housing costs are more expensive than the city centre, or due to their lack of commuters to the urban core.

Figure 5 indicates that once the income is increased, extra-long commutes aren’t worthwhile for many workers as they need to spend less time to earn the money to afford living in the more expensive city centre. However, remaining on fringe areas—with a brief commute by car to the city centre—is still the most cost and time effective option for many of those currently living there.

4.2.2. Scenario 2: Total Cost of Car Ownership

As indicated before, Cambridge and Oxford are the two cities where—even with the costs of owning and operating a car—a noticeable share of individuals can still save money and time by not moving into the city (Figure 6). These are predominantly areas that are either relatively close to the city itself or in very reasonably priced neighbourhoods.

4.2.3. Scenario 3: External Costs

When the external costs of car commutes are considered, the maps look similar to Figure 6, just with the red areas (i.e., should not move) even fewer.

4.3. Sensitivity Analysis

The sensitivity analysis in Schnieder [42,44] has already highlighted that the randomness involved in the creation of the population data has little effect on the results. To assess the robustness of the model outcomes to variations in the cost estimates, the following sensitivity analysis was conducted:

This analysis assumes an income of £12.21 per hour. When the fuel costs are increased or reduced by 20%, the share of people living in the most time and cost-effective place changes by between 0pp to 5pp depending on the city. When the total cost of car ownership is reduced by 20%, this share reduces by 19pp and 12pp in Cambridge and Oxford and less than 5pp for all other cities.

The next part applies to an income of £12.21 per hour and for scenario one (i.e., only fuel and mortgage expenses). Changing the costs of riding a bicycle by 20% only changes the share by up to 4pp. When the mortgage costs in the suburbs are changed, the share reduces for most cities by around 22pp or increases by around 11pp. However, reductions of up to 47pp are possible. Changing the mortgage payments for city homes has a similar but opposite effect. Hybrid working (e.g., 3 days per week in the office) reduces the share of people who should move to the city centre by between 6pp to 38pp depending on the city. Adding some rush-hour traffic by reducing the average speed of the commute by 20% increases the share of those who should move by between 1pp and 15pp.

Although the shares (i.e., proportion who should move) of specific cities exhibited a moderate sensitivity to certain cost assumptions—notably the mortgage costs—the overall patterns and conclusions remain stable. However, it is important to stress that this situation reflects a snapshot of the current housing market. Any increase in demand for housing within the city centre is likely to induce a rise in the property value. This effect could be compounded if the housing supply is inelastic in these cities [73]. The economics will then shift towards commuting from the suburbs. However, declining suburban property values could incentivise remote workers to relocate outside the urban core, potentially alleviating city centre housing pressure.

4.4. Environmental Considerations

Although the analysis emphasises the cost and time efficiency in residential choice, the sustainability implications derived from these findings should not be eclipsed. Substituting long car commutes from the suburbs with brief bicycle trips in the city centre reduces harmful emissions and noise pollution, while simultaneously enhancing public health [17], and lowering the costs for society [22]. If individuals subsequently decide to forgo car ownership, further benefits may arise including lower ground-space requirements [74] and avoidance of emissions produced during vehicle manufacture and disposal [75].

4.5. Reasons for Residing Outside the City Core

Even for those who would benefit financially from moving to the city centre, significant barriers persist. The economic constraints limiting access to homeownership in a city range from higher downpayments to individual’s borrowing capacity. By accounting for the lower transport costs in walkable or transit friendly neighbourhoods, ‘location efficient mortgages’ enable borrowers to access larger loans. Although this flexibility in underwriting guidelines may initially appear beneficial, empirical evidence indicates unfavourable consequences in parts due to a sometimes overestimation of the reductions in transport costs [8]. Further impediments to city centre homeownership include geographic limits, housing supply constraints or zoning and permit rules [76], as well as an increased demand for living in housing inelastic cities [73].

Several reasons beyond costs may affect people’s location choices including school quality, crime rates, lifestyle [8], the desire for a garden or aversion to traffic noise [77]. Larger plot sizes, or neighbourhood amenities, might increase the attractiveness of a certain location [78]. Higher rates of job turnover may lead to individuals residing farther from their place of work [78]. The challenges experienced by dual-career couples to find a job for both in close proximity, could result in longer commutes, engagement in commuter partnerships [79], or foregoing one partner’s career advancements [80].

5. Discussion

To demonstrate (i) the economic efficiency of sustainable modes of transport, and (ii) the benefits of merging housing and transport affordability into a unified indicator, this paper presents a digital model of selected regions in England and Wales. As previously emphasised by multiple academic scholars, isolating housing and transport affordability from one another is flawed, since both expenses are interrelated. Higher cost of housing often corresponds to reduced commuting expenses [29], with some studies even identifying a negative correlation between both [7]. Consequently, numerous scholars have advocated for a consideration of both housing and commuting expenses when assessing the affordability of a location [30]. The present study adds further evidence to reinforce the necessity of combining these important measures.

For a growing number of individuals, the city-centre has become prohibitively expensive [27], compelling them to seek housing in suburban and rural regions [1]. Essentially, they trade lower housing costs for a longer commute. At first glance, the lower housing expenses might be appealing; however, the substantial time spent driving and earning the funds to cover the commute cannot be disregarded [16]. This tension between housing and commuting is visualised in this paper by extending the concept of the effective speed to incorporate the time required to earn the income for a chosen housing option.

The analysis in this study demonstrates that residing in peripheral areas around urban centres like Cambridge and Oxford can be considered a rational choice in terms of both time and costs—if only fuel costs are taken into consideration alongside the housing expenses. This observation aligns with the tendency of drivers to underestimate the financial burden associated with car ownership and lengthy commutes [3,4,5]. Unsurprisingly, once the total costs of car ownership are factored in alongside the housing expenses, only a limited number of people, if any, would find it financially beneficial to continue to live outside the numerous cities included in this study. Admittedly, due to the expensive property market in Cambridge, Oxford, Bath, and Guildford, there are still some individuals (3–13%) who benefit from continuing to live further afield. However, for many other cities and towns, the inclusion of the commuting expenses renders the savings on accommodation insufficient to justify the longer commute.

The assertion that high urban house prices limit who can ‘afford’ to live within cycling distance of their place of work, as suggested by Goodman et al. [27], holds some truth for Cambridge and similar cities (i.e., the focus of the study presented in Goodman et al. [27]). However, this pattern does not apply to many other cities explored in this study. Living in the city centre and riding a bicycle can provide significant financial benefits for many (not necessarily for all)—compared to commuting by car from the outskirts. However, this paper does not intend to downplay the existence of an affordable housing crisis in city centres. The choices of families are effectively constrained, when they are unable to obtain a mortgage sufficient to purchase a home within the city, yet can afford one on the periphery. Also, the sometimes-limited availability of city centre housing options should not be disregarded.

However, the psychological dimensions and personal priorities should not be overlooked, as individuals may hold preferences regarding how they allocate their time. Despite extensive evidence regarding the adverse effects of car commuting on personal health and wellbeing [17], as well as the societal costs such as road maintenance, traffic congestion [22], external effects [25] and social inequalities [26], some individuals may prefer to spend time driving rather than working additional hours to afford housing in the city. Although this is a matter of personal choice, the fact remains that mortgage payments generate equity, whereas money spent on commuting does not. In other words, even if the total time invested in housing and transport combined is equivalent, the individual investing in a home accrues assets, while a commuter acquires less due to opting for lower-priced housing and having to pay for their car.

6. Future Work and Policy Recommendations

In the case of Cambridge and Oxford, most individuals generally would not benefit financially by moving from their current suburban home to the city centre when only fuel costs are considered alongside housing. The same is true for most individuals in cities like Bath and Guildford. However, when the full cost of car ownership is considered, this pattern starts to disappear. Considering that drivers are known to underestimate their car costs [6], it raises the question of how an awareness of the full costs of car ownership might influence residential choice. Further, any factor that might discourage individuals from cycling or living in the city centre needs to be addressed. Whether it is a desire for a garden, to avoid traffic noise, or concerns over safety [77], the reasons for avoiding city-centre dwelling are numerous and warrant attention. Beyond personal inclinations, the inability to secure sufficient funds for city-centre housing may force individuals to live in more reasonably priced locations, necessitating a life of commuting. Future studies should address (i) the urban housing affordability as well as (ii) their supply and attractiveness (iii) alongside strategies to promote cycling.

Since similar datasets to those used in this study may also be available in other geographic regions in the world, this study could be repeated to draw attention to the issues raised in this paper to local policymakers.

7. Conclusions

As city-centre living becomes increasingly unaffordable for many [27], individuals are seeking alternative housing options in suburban or peripheral areas [1]. Although a longer commute in exchange for low-cost housing may appear beneficial at first glance, when the commuting time and expenses are factored in, this trade-off becomes less appealing. To illustrate the economic dilemma, the study expands the concept of the effective speed to account not only for the duration spent commuting and the time required to earn the money to pay for it, but also for the time needed to pay for their housing. This approach integrates housing and transport affordability into a single metric.

The analysis in this paper shows that moving into the city centre and cycling to work can often—but not always—provide more time and cost savings than continuing to commute by car from outside of the city.

The pursuit of identifying cost and time effective residences, which dominates this paper, should not obscure the sustainability benefits of cycling as a substitute for the long car commute. In fact, this study demonstrates—rather clearly—that strategies to cut transport emissions can also produce economic gains.