The Accessibility Turn

From Practice to Theory and Back Again

I first encountered the concept of accessibility in a transport planning class during the late 1980s.¹ In Susan Hanson’s (1986) Geography of Urban Transportation² was an equation adapted from Walter Hansen’s foundational 1959 article:³

\(A_i=\sum_jO_j d_{ij}^{−b}\)

In plain English, opportunities (O) at points j that are closer (smaller distance (d) between points i and j) contribute more to accessibility (A) at point i than those farther away, and the parameter b controls how fast that contribution drops off with distance. Even without math, the main idea was clear: Hansen had essentially captured “accessibility” – the potential for interaction – as a quantifiable concept

Hansen demonstrated this empirically in the Washington, DC region, showing that neighbourhoods with higher job accessibility grew faster​. This insight laid the groundwork for understanding the land use-transport feedback cycle: transport improvements boost access, which in turn influences where development happens.

Yet, planning practice in the late 20th century was still dominated by mobility metrics – how fast traffic moved, how many vehicles passed through an intersection, the ubiquitous highway Level of Service grades. Accessibility was acknowledged in academia, but it hadn’t yet become the central lens for everyday transport and land-use decisions. Fast forward to the 2020s, and we can see a profound change.

Geographers often speak of “turns” in research to signal a fundamental shift in perspective​.⁴ The Accessibility Turn marks a shift in transport and land-use planning —from focusing on how fast people can move, to how well they can reach the things they need. My career tracks this shift. What follows is a first-person account of how accessibility became central, highlighting key ideas, studies, and changes in practice along the way, a loosely chronological narrative from the 1990s to the 2020s, tracing how accessibility rose to prominence and became the guiding principle for measuring transport and land-use outcomes.

The 1990s: From Mobility Measures to Multimodalism

In the early 1990s, a young transport planner for the Maryland-National Capital Park and Planning Commission/Montgomery County Planning Department (MNCPCC/MCPD) confronted the tension between mobility and accessibility.

The county’s growth management policy required us to assess whether transport (and school) infrastructure could ‘adequately’ support new development in each area (Levinson 1997). Initially, the Planning Department focused almost entirely on roads, using VMT-weighted volume-to-capacity LOS measures averaged over zones to determine if an area was ‘over capacity.’ Transit was a secondary consideration, used mainly to adjust LOS thresholds—reinforcing auto-mobility as the default. But it was clear we needed metrics that properly reflected all modes, not just cars. In 1993, as part of overhauling our Policy Area Transportation Review process, our team introduced a new metric called Total Transportation Level of Service (TTLOS) (Levinson and Kumar 1994b)​.

TTLOS combined highway and transit performance into a single multimodal index, similar to a GPA for transport. Each mode’s LOS was weighted by its share of trips—for example, transit carrying 20% of trips would get a 0.2 weight. This approach rewarded dense areas with good transit links, and down-weighted token service in low-density zones:

\(TTLOS=(Highway LOS×Highway Mode Share)+(Transit LOS×Transit Mode Share) \)

Areas with good transit accessibility⁵ could tolerate more road congestion without triggering development moratoria, and vice versa. TTLOS incentivised multimodal outcomes, measuring transit not by inputs (like bus frequency), but by outputs (accessibility to opportunities).

TTLOS wasn’t perfect – we debated how to set ‘adequate’ thresholds, and some worried that bundling roads and transit could mask specific problems (a great transit score might hide awful road congestion, or vice versa)​.

Ultimately, the philosophy was clear: transport performance should be mode-agnostic and measured by accessibility outcomes, not merely by vehicle speeds.⁶

After that young transport planner left the agency in 1994 to pursue transport economics at Berkeley, the system for measuring access was not maintained. Staff struggled to replicate the gravity-based transit accessibility calculations, and TTLOS was eventually dropped. Still, for one brief shining moment, accessibility directly shaped policy decisions.

By the late 1990s, academics were also describing this shift; Susan Handy and Debbie Niemeier (1997) provided a framework for bridging the gap between academic measures and practical planning​. The profession was waking up to the idea that accessibility — not just mobility — deserved to be a performance measure in its own right.

Accessibility and the Journey to Work

Amidst these planning and modeling projects, I was also observing broader travel behavior patterns in the county and region. One puzzle that emerged from our data was the stability of commute times. We saw that, over the years, average journey-to-work times were not spiralling upwards as much as one might expect, even as the Washington metro area sprawled outward. People were moving further out, but on average, people weren’t spending any longer on their commutes than before. This hinted at a self-regulating aspect of human behaviour and land use: what we termed the rational locator hypothesis (Levinson and Kumar 1994a). The idea is that people and employers adjust locations in response to congestion so that commute times stay within tolerable bounds.

Were commuters self-selecting into home-work pairs that kept their travel times roughly constant? And how to measure that effect? The key was accessibility. If a person lives in a job-rich area (high job accessibility at home) or works in a labor-rich area (high worker accessibility at the job site), it should be easier for them to find a closer home-job match and thus commute a shorter time. In contrast, if a worker who lives in a remote area with few nearby jobs and also work in a sparse area would be more likely to end up with a long commute by necessity.

This line of inquiry culminated in Accessibility and the Journey to Work (Levinson 1998)⁷ In that study, I quantified the job accessibility around each resident’s home and the worker (labor) accessibility around workplaces, and examined how those related to the commute durations. As expected, there was a clear relationship: when both home and work locations had high accessibility, commutes tended to be shorter, and overall the region’s average commute time stayed stable even as the region grew in population and extent.

In other words, a balance in accessibility at both ends of a trip contributes to keeping commute times in check. This clarified the jobs–housing balance idea (Cervero 1989), showing that what matters is not simply the ratio of jobs to houses in a specific suburb, but the number of jobs and competing workers reachable from that suburb within a reasonable travel time. I theorised that commuters adapt to urban structure, and accessibility metrics helped quantify these trade-offs. Causal testing, however, had to wait for better time series data (Huang et al. 2018).

One outcome from that logic was that improving accessibility can be a win-win: for example, if we located housing near job hubs, (or jobs in bedroom suburbs) we could reduce commute times for residents while also easing demand for travel. That recognition influenced not just my research but also how I thought about land-use policy. In essence, the Journey to Work project was a bridge between my practitioner experience and academic future. It translated the messy reality of metropolitan commuting in the relatively wealthy and well-quantified Montgomery County into a set of generalisable insights about how accessibility shapes travel behaviour.

It planted the seed for many of my later studies on network design, land use, and accessibility. For example, in the following decades a large part of my research group’s output examined how transport networks evolve when guided by accessibility benefits, culminating in a book on Evolving Transportation Networks (Xie and Levinson 2011), and then a set of dissertations at the Universities of Minnesota (Schoner 2017) and Sydney (Lahoorpoor 2022, Rayaprolu 2023). That line of research provided a theoretical foundation for understanding why certain links are built and how networks self-organise to serve accessibility needs.

The 2000s: Academic Foundations and Early Adoptions

By the 2000s, the concept of accessibility was gaining firmer footing in research and slowly filtering into practice. Scholars like Karst Geurs and Bert van Wee (2004) published comprehensive frameworks for accessibility evaluation, categorising measures (infrastructure-based, location-based, person-based, utility-based) and urging that transport projects be assessed in terms of accessibility outcomes​. The idea that we should plan for accessibility (not just mobility) was no longer radical; it was becoming a research mainstream. The literature from this period is rich with accessibility metrics and case studies – from cumulative opportunity measures (e.g. counting how many jobs or schools can be reached within 30 minutes) to gravity-based indexes and utility-based log-sum measures derived from travel models.

Access to Destinations

During this time, I pivoted from planning practice to academia, determined to refine accessibility metrics and make them more usable. In a project at the University of Minnesota called Access to Destinations (A2D) (circa 2004–2011), my colleagues, especially Ahmed El-Geneidy, and I reviewed existing metrics and experimented with new ones. This research program, funded by MnDOT, diverged from previous access work because at their recommendation, it used measured rather than modelled travel times as its core source of data.⁸ Working out these travel times was a considerable element of the work.

We introduced a concept dubbed Place Rank, inspired by Google’s PageRank, to account for the accessibility of surrounding areas in a recursive way​ (El-Geneidy and Levinson 2006, 2011). More practically, we compared gravity-based vs. cumulative-opportunities measures head-to-head and found that threshold-based cumulative measures (e.g. number of jobs reachable within 20, 30, or 45 minutes) were often more intuitive for policymakers. They were easier to explain to the public and decision-makers: a headline like “100,000 jobs are accessible within 30 minutes from this neighbourhood” resonates more than an abstract index. This emphasis on clarity and usability helped pave the way for the wider adoption of cumulative accessibility metrics — such as 30-minute job counts — in planning practice.

Our research in the 2000s also reinforced that accessibility isn’t just an academic construct – it has real economic and behavioural implications. We found that homebuyers pay a premium to live in locations with better job accessibility (Iacono 2011, 2017)​. In other words, the housing market was capitalising the benefit of access: people were willing to spend more to reduce their commute times or have more opportunities nearby. This finding echoed what urban economists had long posited and what Hansen hinted at decades earlier, but now we had empirical validation. It bolstered the argument that improving accessibility (say, by building a new transit line or enabling mixed-use development) can create tangible value. Planners started to see accessibility as a key outcome to maximize, not just an abstract measure.

In terms of policy uptake, the 2000s saw some early adopters of accessibility-based planning. The UK, for example, rolled out a national Accessibility Planning initiative in the mid-2000s, prompted by a 2003 government report on social exclusion that recognized lack of access to jobs and services as a critical problem (Social Exclusion Unit). This Making the Connections report recommended new accessibility metrics and led the Department for Transport to require that local authorities include accessibility strategies in their transport plans​. Targets were set for indicators like the percentage of households within 30 minutes of a hospital by transit, or travel times from low-income neighbourhoods to employment centers. It was a notable shift from the old predict-and-provide approach focused on road capacity to a focus on ensuring people could actually reach essential destinations. Other countries and cities began experimenting with similar approaches on a project or regional basis, integrating accessibility metrics into scenario planning, transit-oriented development projects, and even project appraisal. By the end of the 2000s accessibility was firmly established in the toolkit.

Land Use and Transport Planning as a Discipline

In 2004, Kevin Krizek and I co-edited the 2005 book Access to Destinations (Levinson and Krizek 2005), one of the earliest comprehensive texts focused on accessibility, drawn from submission from the first of two Access to Destinations conferences held in Minnesota. The A2D project helped set the research agenda for accessibility as a multi-modal, multi-dimensional concept — one that captures the interaction between land use and transport.

Our textbook Planning for Place and Plexus argues that land use and transport are fundamentally interconnected systems that must be planned together, not in isolation (Levinson and Krizek 2007). We proposed a framework based on accessibility as the core measure linking transport, social, and other networks (the plexus) and urban form (place). Changes in one system drive changes in the other, creating a continuous feedback loop.

In 2008, with support of the University of Minnesota, we launched Journal of Transport and Land Use (JTLU), for which I was the founding editor, to provide an open-access platform for research at the intersection of transport systems and land use planning.⁹ It emerged from the recognition that these two fields, often treated separately, are deeply interconnected and require integrated study. JTLU publishes peer-reviewed, interdisciplinary work that advances theory, methods, and practice related to accessibility, spatial development, urban form, and policy. By making research freely available and fostering a global community of scholars and practitioners, the journal plays a key role in shaping how cities are understood and designed through the lens of access and co-evolution.

This was followed by the World Symposium on Transport and Land Use Research, held in Whistler, Canada, which begat the World Society for Transport and Land Use Research (WSTLUR) in 2011 to promote the field. Following on our Access to Destinations conference series, WSTLUR coalesced researchers into a more coherent discipline. WSTLUR organises an approximately triennial World Symposium on Transport and Land Use Research, subsequent symposia have taken place in Delft, Brisbane, Portland (and online due to COVID), and Bogata. The society also supports JTLU.

Later, we co-authored The End of Traffic and the Future of Access (Levinson and Krizek 2017), discussed below.

The 2010s: Measuring What Matters

If the 2000s established the groundwork, the 2010s made accessibility mainstream, thanks in part to better data and the need for smarter metrics in an era of constrained resources.

While at Montgomery County I had noodled on a concept that in my head, and later in my Palm Pilot, was called “Accessibility Nation”. The idea would have been to extend the accessibility analyses for metro Washington and run simple travel demand models for every city in the US and estimate accessibility using Census TIGER files to build the network and census data for estimates of employment and population, using EMME/2 for developing the travel time matrices. It was premature.

At the University of Minnesota, we launched the Accessibility Observatory and the annual Access Across America reports to put hard numbers on accessibility in cities. In 2013, our first report ranked the 51 largest U.S. metro areas by job accessibility by car​ (Levinson 2013). The results were enlightening: Los Angeles, San Francisco, and New York topped the list – not because they have free-flowing traffic (far from it), but because they have huge concentrations of jobs. Despite the reports of congestion given heavy airplay by the TTI Urban Mobility Scorecard (e.g. Schrank et al. 2015), the sheer density of opportunities meant a typical worker could reach a lot of destinations in a short time. This finding flipped the script: it showed that speed isn’t everything. A city can have slow traffic yet still offer high accessibility if jobs, services, and homes are close together. Conversely, a region can have fast highways but if everything is spread out, many people still won’t reach much within 30 minutes. This was a vivid demonstration of why accessibility (outcomes) trumps mobility (speeds) as a performance measure​.

We applied the same approach to transit. By 2017, with the availability of tools like OpenTripPlanner (the forerunner of R5), and data sets like the US Census LEHD, along with standardised transit schedules and Open Street Map, under my former student Andrew Owen’s leadership, we were tracking how many jobs were accessible within 30 or 60 minutes by public transit in dozens of cities, and how those numbers changed year-over-year​. The findings showed that accessibility is not static – it responds to policy and service changes. For example, Kansas City achieved a 17% jump in transit job accessibility in one year after a bus network redesign, the highest gain in the country. Even transit-rich San Francisco managed nearly a 9% increase by optimizing routes and schedules. Such changes underscored that cities can boost accessibility relatively quickly with the right interventions. Importantly, our analysis highlighted the factors behind these outcomes: the size and spatial distribution of the job market, the extent and frequency of transit service, and urban form – especially population density and decentralization​. The policy takeaway was straightforward: invest in aligning transit with land use (for instance, higher transit frequency in dense job centres, or encouraging housing development near transit lines) and you will measurably improve accessibility. Several Metropolitan Planning Organizations (MPOs) and transit agencies began incorporating “access to opportunities” indicators into their evaluation frameworks, and explicitly referenced our reports as justification.

The idea of the “30-minute city,” for example, was adopted as a policy goal in places like Sydney, meaning that most residents should be able to reach a major centre (work, shopping, services) within a half-hour trip by walking, cycling or transit (Levinson 2019). That concept is essentially an accessibility metric in plain language, and its traction in public discourse showed how far the Accessibility Turn had come.

In academia, the phrase “from mobility to accessibility” became a shorthand for The Accessibility Turn, even lending its name to a 2019 book by Levine and colleagues documenting how to transform planning practices accordingly (Levine et al. 2019).

Distributions

A key development in the 2010s was the growing focus on how access varies across the population. Researchers like Karel Martens argued that measuring accessibility is not just a technical task but a moral imperative: transport authorities have a responsibility to ensure fair access for all. In Transport Justice (2016), Martens contended that disparities in access to jobs, education, healthcare, and other services lie at the core of transport inequality, and that accessibility should be treated as a basic right in transport policy.

This perspective led to new methods for identifying who benefits from accessibility improvements — for instance, gap analyses by income or race—and for targeting investment in “access-poor” communities. El-Geneidy et al. (2016) showed that when transit fares were included in accessibility calculations, some low-income neighbourhoods that appeared well-served by time-only measures had far lower effective access. The cost burden reduced residents' ability to use available transit, exposing a gap between nominal and real access.

The movement from measuring average access to examining its distribution highlighted that equal travel times do not imply equal opportunity if costs or barriers vary.

Dynamics

Exploiting the data collected for Minnesota in the Access to Destinations studies, one of our group’s key findings was what we described as a “flattening” of intra-metropolitan accessibility over the period from 1995 to 2005 (Levinson et al. 2017). While overall job accessibility increased — due to highway improvements and employment growth — suburban areas experienced greater gains than central areas. This convergence reduced the gap between city and suburb, pointing to a trend of spatial equalization in accessibility. It also raised important distributional implications: investments were diffusing access more evenly across the region.

We explored how changes in accessibility over time influence land use and property values. Building on our earlier cross-sectional finding that locations with higher regional accessibility commanded higher home prices (Iacono and Levinson 2011), we aimed to establish causation over time.

In Iacono and Levinson (2017), we addressed a central question: Do land values follow accessibility changes? Using a difference-in-differences approach, we matched home sale data with measured changes in accessibility resulting from new transport infrastructure. We found that increases in regional accessibility do translate into higher land values, though the effect varies by location and often occurs with a time lag. The elasticity was positive but modest, suggesting that accessibility remains one of several drivers of housing prices.

Importantly, this research moved beyond the static correlations long known since Hansen (1959) (and observed in our 2011 study) by establishing a causal link between transport improvements and land market responses.

Improved accessibility increases the desirability of locations, raising land values. Thus, accessibility serves as a tangible economic metric, bridging transport investments and urban economics.

The practical takeaway for policymakers is clear: accessibility benefits of new infrastructure tend to be capitalised into land values, providing a rationale for value capture mechanisms. If a new transit line substantially boosts access in a corridor, value uplift can be used to help finance that investment — whether through targeted property tax increments or developer contributions. Our work provided empirical support for such strategies by quantifying the dynamic relationship between access and land value over time.

Agglomeration

In collaboration with colleagues from Imperial College London, we examined how job accessibility contributes to agglomeration economies and urban productivity (Melo et al. 2017). Using matched firm-level productivity data across urban areas in the United States, we estimated the elasticity of productivity with respect to accessibility to labor and employment. The results showed that firms located in areas with greater accessibility tended to be more productive, even after controlling for sector, scale, and regional characteristics. Importantly, the effect sizes varied by industry: knowledge-intensive sectors exhibited stronger sensitivity to changes in access. Our work provided evidence that accessibility improvements can induce measurable economic returns at the firm level, reinforcing the link between transport investment, land use coordination, and regional competitiveness. By translating accessibility metrics into productivity gains, this line of research supports the case for incorporating accessibility into broader economic appraisal frameworks.

Externalities

In our pursuit to enhance accessibility evaluation, my doctoral student, and later post-doc, Mengying Cui, and I developed the Full Cost Accessibility (FCA) modelling framework. Traditional accessibility measures focus on travel time, overlooking other significant internal and external costs associated with travel, such as crash risks, environmental emissions, and health impacts. Recognising that, from a societal perspective, accessibility assessments should encompass the full generalised cost of travel — not just time, or even out-of-pocket costs like fares and fuel, so we introduced the FCA framework.

Initially outlined in Cui's 2018 dissertation and subsequently published in (Cui and Levinson 2018, 2019), the FCA framework integrates temporal, monetary, safety, and environmental cost components into accessibility evaluations. Instead of merely asking, "How many jobs can be reached within 30 minutes?", full-cost accessibility prompts us to consider, "How many jobs can be reached for $X of social cost?"

Applying this framework to the Minneapolis–St. Paul metropolitan area, we discovered that incorporating full costs can significantly alter the accessibility landscape. We quantified an average full cost of approximately $0.68 per vehicle-kilometer in the region, with time and direct monetary costs comprising about 85% of this figure, and externalities (safety, emissions) accounting for the remainder.

These findings suggest that while travellers may not drastically alter their routes to mitigate external costs alone (since time predominantly influences individual decisions), it's crucial to identify which network links, or especially modes — as this facilitates intermodal comparisons — impose higher external costs for the access they provide. Incorporating such insights enables planners to prioritise interventions that enhance accessibility both cost-effectively and sustainably—for example, investing in projects that reduce travel times without significantly increasing crash or emission costs, or implementing policies that internalise external costs (like congestion pricing) to encourage more balanced accessibility outcomes.

The practical significance of FCA lies in its potential to reshape project appraisal and network design. Traditional benefit–cost analyses might favour projects that save travel time, but when considering full costs, alternative projects that modestly reduce travel time while substantially improving safety or reducing emissions could offer greater net accessible opportunities per dollar. Our full-cost accessibility framework thus advances the field toward evaluating transport improvements not solely based on mobility gains, but through a holistic cost–benefit lens that reflects society's ability to access destinations. This comprehensive perspective aligns with the growing emphasis on sustainable transport, capturing how accessibility can be enhanced in ways that also mitigate external harms.

Primal and Dual Accessibility

One of our methodological contributions distinguished between primal and dual accessibility measures (Cui and Levinson 2020b). Primal accessibility fixes a travel cost (e.g., 30 minutes) and counts reachable opportunities from an origin (e.g., “How many jobs within 30 minutes?”). In contrast, dual accessibility fixes the number of opportunities and measures the cost needed to reach them (e.g., “How long to reach 10,000 jobs?”); dual access enables many kinds of comparisons that would otherwise be clunky, such as the time required to reach multiple destinations.

Multi-Destination and Multi-Activity Access

Accessibility is often measured for a single destination type, usually jobs, but in reality, people conduct complex activity patterns throughout the day. Recognising this, my doctoral student Mengying Cui and I developed the concept of multi-activity accessibility to better reflect how people access bundles of essential opportunities, not just one type at a time.

Rather than treating access to employment, retail, education, healthcare, and leisure as separate indicators, we proposed a composite measure that integrates these into a unified index. The goal was to answer a more realistic question: How well can a person at location X reach a variety of essential opportunities within a reasonable time? (Cui and Levinson 2020a).

Institutionalisation: The Transport Access Manual

By the close of the 2010s, accessibility metrics and concepts had proliferated. I sought to consolidate the lessons of these decades of research into resources that would aid both scholars and practitioners. In 2020, a committee of colleagues and I released the Transport Access Manual, an open-access guide for quantifying and evaluating access in cities (COTAM 2020). It provides standardised definitions, methods, and use cases, effectively translating academic research into step-by-step guidance.

The Access Sextet

Another aim was to popularise these concepts, and the aim was to use a series of open-access (freely downloadable) books, drawing from and extending my Transportist blog, to do so.

The End of Traffic and the Future of Access (Levinson and Krizek 2017) looked at trends, both historical and current, and reframed future transport scenarios in terms of accessibility rather than vehicle throughput. In it, we argued that the success of new technologies like autonomous vehicles or mobility platforms should be judged by how they improve access to destinations, not just how they move vehicles.

Spontaneous Access (Levinson 2017) explores how people interact with their environments when access is unplanned, informal, or emergent rather than the result of deliberate infrastructure or policy decisions. It focuses on the everyday, often overlooked ways people find and create access—through shortcuts, informal paths, flexible schedules, or ad hoc coordination. The idea challenges traditional planning models that assume access is delivered only through formal networks and structured systems. Instead, it highlights how urban life depends on a baseline of informal, opportunistic access shaped by local knowledge, adaptability, and human improvisation. Recognising spontaneous access reveals gaps in planned systems and underscores the importance of designing cities that support flexible, adaptive use rather than rigid, top-down control.

Elements of Access (Levinson et al. 2017) formalises accessibility as a practical, measurable framework that bridges the long-standing divide between transport engineering and urban planning. Rather than treating accessibility as a vague ideal, the book disaggregates it into distinct components — people, places, plexus, production, and progress — each capturing an essential part of how individuals reach destinations within urban systems. This structure not only clarifies the concept but makes it usable: engineers gain insight into spatial and behavioral dynamics, while planners are introduced to the mechanics of networks and flows. The book advances the field by repositioning accessibility as the central objective of land planning and transport policy, providing a shared vocabulary and analytic foundation for more integrated, outcome-focused urban decision-making.

The Political Economy of Access (Levinson and King 2019) examines how power, resources, and institutional structures shape who gets access to what, and under what conditions. Access is not distributed evenly; it is produced through political choices, market forces, and governance arrangements that privilege some groups while excluding others. This perspective shifts the focus from access as a neutral technical outcome to access as a contested and negotiated result of economic interests, policy decisions, and social hierarchies. It raises questions about fairness, control, and accountability in how transport and land use systems are designed and maintained. By foregrounding the political and economic dimensions, this approach calls for greater attention to the distributional consequences of planning, and for frameworks that prioritise access as a public good rather than a private entitlement.

In The 30-Minute City I illustrate specific interventions to enhance urban accessibility in (Levinson 2019). For instance, the A Line Bus Rapid Transit (BRT) in Minneapolis-St. Paul, presents operational improvements like off-board fare payment and all-door boarding that substantially reduced delays, making transit faster and more reliable. Similarly, traffic signal timing adjustments should prioritise pedestrian movement over vehicles at intersections, thus reducing pedestrian waiting times and improving overall walking accessibility. Targeted, incremental strategies will collectively advance the realisation of accessible, human-centred urban environments.

The final of these books, Applications of Access, is a collected volume of original research that features case studies and methodological advances across the spectrum of accessibility planning (Levinson and Ermagun 2022).

These works distilled our collective knowledge into actionable insights, helping to mainstream accessibility as a planning objective.

The 2020s: Accessibility Ascendant

By the early 2020s, the Accessibility Turn was essentially complete – yet the work continues to refine and extend the concept. Accessibility metrics are now being applied at global scales. Wu et al. (2021) compared 117 cities worldwide on the number of jobs reachable within 30 minutes by automobile, transit, walking, and cycling​. The results highlighted that no single city “has it all” – different urban development patterns involve trade-offs in accessibility by mode​. US cities, for instance, achieve relatively high job accessibility by car for their size (thanks to extensive road networks and dispersed development) but lag significantly in transit and walk accessibility due to lower densities and weaker transit service. In contrast, many European and Asian cities – denser and better served by high-frequency transit – excel in accessibility by transit and active modes. Australian and Canadian metros tend to fall in between. These patterns underscore a key message: different urban models involve different accessibility trade-offs. Such comparisons give planners a realistic way to benchmark and learn from peer cities. A car-centric city might look to an Australian metro of similar size to gauge the impact of investing in transit, while a transit-rich European city might calibrate expectations about driving access without sacrificing mode share goals​. The fact that we can even make these comparisons reflects how far data and methods have come; a decade ago, assembling standardized global accessibility measures would have been extremely difficult, but with open data and open-source tools it’s becoming routine.

Notably, the conversation has become global: a recent special issue of a planning journal was devoted to the “Accessibility Turn” in Latin America and the Global South, examining how accessibility-focused approaches can address urban challenges in rapidly growing cities​ (Villamizar 2025).

Theory

On the theoretical front, researchers are seeking to unify the various threads of accessibility into a coherent framework. My student Hao Wu and I proposed the idea of a generalized accessibility measure, imagining the ideal of measuring access “for all places, all modes, all purposes, at all times”​ (Levinson and Wu 2020). While such an all-encompassing measure isn’t directly computable, it’s a useful ideal. It pushes us to think beyond modal or spatial silos and to seek common structure among metrics. We showed that under certain conditions, more complex person-based (utility) measures and simple location-based (opportunity count) measures can be made equivalent​ (Wu and Levinson 2020). This kind of synthesis provides conceptual clarity as the menu of metrics expands. By the 2020s, one can find a wide array of accessibility indicators in use – primal and dual measures, cumulative and gravity, weighted by population or jobs, time-based, cost-based, etc. The aim of a general theory is to offer a roadmap for choosing the right metric for a given planning question, and to ensure we don’t lose the forest for the trees. Fundamentally, whether you express accessibility in minutes, in dollars, or in abstract utility units, the goal is the same: to quantify how well the land-use and transport system connects people to the things they need and value.

Interestingly, even as we zoom out for global studies, accessibility thinking is also zooming in to the micro scale of design. For example, Bahman Lahoorpoor and I examined how the placement of entrances in a train station affects the walk accessibility of the surrounding area​ (Lahoorpoor and Levinson 2020). By modeling the pedestrian network, we found that adding a second exit on the opposite side of a station can significantly expand its 5, 10, and 15-minute walking catchment – increasing the nearby population with station access enormously in some cases. This highlighted how small infrastructure tweaks (like where we put a station gate) can yield sizeable accessibility gains. It’s a reminder that accessibility planning operates at multiple scales: from global comparisons down to the design of a sidewalk or station entrance, the principles of access apply.

Access-Based Transit Ridership Estimation

By the 2020s, accessibility research had drilled down to fine spatial scales, linking local transit access with ridership in novel ways. For Portland, we conducted a stop-level analysis using linear regression (with ensemble machine-learning checks) to explain bus boardings (Cui et al. 2022). We found an elasticity of roughly 0.2 — each 10% increase in jobs reachable within 30 minutes by transit was associated with approximately a 2% increase in daily boardings per stop. However, we also saw diminishing returns from overlapping service: adding another stop on the same route within about 400~m reduced a stop’s ridership by roughly 12% due to competition.

Building on this, in Sydney we applied the node-place model (Bertolini 1999) at the station level, defining the node as a station’s additional regional accessibility via transit (beyond what is reachable on foot) and place as the jobs and residents within walking distance. We showed that both nodal and place factors significantly predict rail ridership (Wu et al. 2023), with the number of jobs near a station emerging as the single strongest predictor of usage. This finding reinforces the classic node--place paradigm while addressing its two-dimensional limitation: by using accessibility metrics for both node and place, we captured each station’s connectivity and land-use context in commensurate units.

Most recently, my colleagues and I introduced spatial dependency into the analysis of ridership on metro networks. Examining the Xi’an subway with spatial weight matrices and regression models, we found that improving access to employment and housing around one station not only boosts its own boardings and alightings (Cui et al. 2025), but also increases demand at other stations with complementary land-use functions (e.g., residential areas feeding job centres). Collectively, these studies demonstrate that enhancing both local accessibility and regional connectivity -- while accounting for competition between nearby stops and inter-station dependencies -- yields richer insights into transit ridership patterns, strengthening the case for an accessibility-oriented approach to public transport planning.

Access-Based Cost-Benefit Analysis

In evaluating New York City’s Second Avenue Subway, Yadi Wang and I found that an access-based benefit assessment of the project (based on the land value uplift from new accessibility) painted a much richer picture than conventional evaluation. The accessibility improvements from the subway were associated with substantial increases in nearby land values, indicating benefits that far exceeded what time savings alone would suggest (Wang and Levinson 2022). In a related synthesis, we reviewed dozens of transport projects worldwide and observed a consistent positive correlation between accessibility gains and land value appreciation (Wang and Levinson 2023). These findings reinforce the idea that accessibility is the ‘currency’ of travel benefit that is recognized by the market. By capturing benefits that traditional CBA might miss — such as agglomeration or productivity gains reflected in land prices — access-based methods can give a more complete account of a project’s value.

In ongoing projects, we aim to reconceptualise how we appraise transport projects, shifting the emphasis from travel time savings to accessibility gains. Traditional cost–benefit analysis (CBA) in transport has long relied on monetizing minutes saved as its principal benefit. While this approach has been dominant for decades, it has also been widely critiqued for oversimplifying the value of transport investments.

Mann and Levinson (2024) propose an access-based cost–benefit analysis. Rather than summing time savings, we evaluate projects by how much additional accessibility they deliver — for example, how many more jobs become reachable to residents when a new metro line is built. As with the work with Wang, we then monetize these accessibility gains through changes in land value. The underlying logic is grounded in urban economics: improved accessibility should increase the desirability of land, and hence, its price. Using hedonic pricing models, we estimate the value that households implicitly place on improved access by observing how property prices respond to changes in measured accessibility.

We applied this framework to a proposed Metro line in Sydney. We estimated how many more jobs would be accessible to residents due to the project, then used established land value elasticities to compute the corresponding uplift in land value. Under reasonable assumptions — including complementary land-use changes around stations — we found that the accessibility-based land value uplift of the line could be valued at several billion Australian dollars.

This valuation offered a useful complement to conventional time-savings-based analysis, and also pointed toward a practical funding mechanism: if accessibility gains are capitalized into land value, value capture instruments (such as targeted property taxes or developer contributions) could be used to help fund the infrastructure. More broadly, access-based CBA aligns project appraisal with what ultimately matters to people: the opportunities they can reach.

The conceptual shift here is important. By treating accessibility as the currency of benefit, this approach captures not just mobility, but also wider impacts such as agglomeration, land development potential, and even distributional outcomes. Projects that connect underserved communities to major job centers, for example, might show relatively small time savings per traveler but large increases in access—warranting greater priority in evaluation.

Ultimately, this work (along with parallel efforts by others) reinforces a recurring theme in my research: accessibility offers a unifying metric that ties transport performance to economic, social, and spatial outcomes in a transparent and policy-relevant way. By focusing on access, we ensure that transport investments are assessed in terms of what truly matters: the opportunities they create for people and communities.

---

Looking Ahead: Lessons and Recommendations

Lessons from the Accessibility Turn

Better Data, Better Planning
We now have the tools to measure accessibility in detail across whole regions. Planners can set and track performance indicators or targets — like increasing access to shops or clinics for low-income areas — using open data, Open Street Maps (OSM), General Transit Feed Specification (GTFS) feeds, GPS traces, and land-use maps. These tools support both long-term planning and real-time response.

The Mean, the Variance, and the Skew
Accessibility measured for the average person will tell us little about how non-average people experience the world — and the typical person is non-average. That means examining who has access, not just how much access there is. Differences by income, age, ability, and location should be tracked. We also need to include cost, safety, and comfort — since having access on paper doesn’t always mean it’s real or usable.

Keep Metrics Simple
Accessibility measures should be clear and easy to explain. Simple statistics like “percent of people within 15 minutes of a park” are often more useful than complex scores. Clarity helps with public support and faster decision-making, especially when testing many scenarios.

Make Access Part of Policy
Accessibility still isn’t built into most funding or project review systems. It should be. Adding it to cost–benefit tests and planning rules would help focus investments on what really matters—helping people reach the places they need. This also means giving agencies the tools, staff, and support to use accessibility in daily work.

---

What’s Often Assumed But Not Said

Access Can Create Trade-Offs
Improving access often increases well-being, but it is not without costs. Access can also raise rents, cause displacement, or lead to more local traffic. Planners and policy-makers need to think about who gains, who loses, and what can be done to manage those effects.

Institutions Don’t Always Change Easily
Even with good data and strong arguments, change is hard. Agencies may lack the time, staff, or incentive to switch to access-based planning. Progress depends on shifting rules, funding systems, and habits—what we call the Political Economy of Access (Levinson and King 2019).

People Want Different Things
Standard access metrics assume everyone wants the same things: jobs, schools, clinics. But needs vary by age, income, and culture. A good access plan must reflect that — not just average outcomes, but what different groups care about.

Part of an Agenda for Research and Practice

Virtual Access
Many activities — work, shopping, learning — can now happen online. But most metrics still focus on physical trips. We need new measures that also reflect digital access and how it shapes travel. Work in this area is starting (e.g. Chen et al. 2024a, b), but more is needed.

Resilient and Reliable Access
Disasters both natural and manmade can cut off access (Cui and Levinson 2018a,b). Metrics should reflect not just normal days but disrupted ones, tracking how well people can reach what they need during and after shocks. This includes safe paths by low-carbon modes like walking, biking, or transit.

Perceived Access
People don’t always choose the fastest route, and certainly don’t readily assess access well (Aoustin and Levinson 2020). Habits, comfort, knowledge, familiarity, reliability security, and safety all matter. Analyses should reflect what people actually do, not only what they could do. That will make access metrics more useful in real planning.

Conclusion

The shift from speed to access changed the focus from how fast we can move to whether people can reach what they need. That change has led to better tools, better questions, and better outcomes.

But there’s more to do: we must account for differences across people, behaviour, digital access, and climate risks. If access becomes the goal, cities can become more efficient, more just, and more resilient. That’s the lasting value of the Accessibility Turn.

---

References

• Aoustin, L., & Levinson, D. M. (2021). The perception of access in Sydney. Findings.

• Bertolini, L. (1999), ‘Spatial development patterns and public transport: the application of an analytical model in the Netherlands’, Planning Practice and Research 14(2), 199–210.

• Cervero, R. (1989). Jobs-housing balancing and regional mobility. Journal of the American Planning Association, 55(2), 136-150.

• Chen, J., Cui, M., & Levinson, D. (2024a). The cost of working: measuring physical and virtual access to jobs. International Journal of Urban Sciences, 28(2), 318-334.

• Chen, J., Cui, M., & Levinson, D. (2024b). The cost of shopping: measuring virtual and physical access for obtaining goods. TeMA-Journal of Land Use, Mobility and Environment, 17(1), 129-145.

• Cloke, P. and Johnston, R. (2005), ‘Deconstructing human geography’s binaries’, Spaces of Geographical Thought pp. 1–21.

• Committee for the Transport Access Manual (David Levinson, c. a. (2020), Transport access manual: A guide for measuring connection between people and places, Network Design Lab, Sydney, NSW.

• Crozet, Y. (2020), ‘Cars and space consumption: Rethinking the regulation of urban mobility’.

• Cui, B., DeWeese, J., Wu, H., King, D. A., Levinson, D. and El-Geneidy, A. (2022), ‘All ridership is local: Accessibility, competition, and stop-level determinants of daily bus boardings in Portland, Oregon’, Journal of Transport Geography 99, 103294.

• Cui, M., & Levinson, D. (2018a). Accessibility and the Ring of Unreliability. Transportmetrica A: transport science, 14(1-2), 4-21.

• Cui, M., & Levinson, D. (2018b). Accessibility analysis of risk severity. Transportation, 45, 1029-1050.

• Cui, M. (2018), Full Cost Accessibility, PhD thesis, University of Minnesota, Minneapolis, MN. URL: https://conservancy.umn.edu/handle/11299/200220

• Cui, M. and Levinson, D. (2018), ‘Full cost accessibility’, Journal of Transport and Land Use 11(1), 661–679. URL: https://www.jtlu.org/index.php/jtlu/article/view/1042

• Cui, M. and Levinson, D. (2019), ‘Measuring full cost accessibility by auto’, Journal of Transport and Land Use 12(1), 649–672. URL: https://www.jtlu.org/index.php/jtlu/article/view/1495

• Cui, M. and Levinson, D. (2020a), ‘Multi-activity access: How activity choice affects opportunity’, Transportation Research Part D: Transport and Environment 85, 102364. URL: https://doi.org/10.1016/j.trd.2020.102364

• Cui, M. and Levinson, D. M. (2020b), ‘Primal and dual access’, Geographical Analysis 52(3), 452–474.

• Cui, M., Yu, L., Nie, S., Dai, Z., Ge, Ye. and Levinson, D. (2025), ‘How do access and spatial dependency shape metro passenger flows’, Journal of Transport Geography 123, 104069.

• El-Geneidy, A. and Levinson, D. (2006), Access to Destinations: Development of Accessibility Measures, Technical Report MN/RC 2006-16, Minnesota Dept. of Transportation, Minneapolis, MN.

• El-Geneidy, A. and Levinson, D. (2007), ‘Mapping accessibility over time’, Journal of Maps 3(1), 76–87.

• El-Geneidy, A. M. and Levinson, D. (2011), ‘Place rank: valuing spatial interactions’, Networks and Spatial Economics 11, 643–659.

• El-Geneidy, A., Levinson, D., Diab, E., Boisjoly, G., Verbich, D., & Loong, C. (2016). The cost of equity: Assessing transit accessibility and social disparity using total travel cost. Transportation Research Part A: Policy and Practice, 91, 302-316.

• Geurs, K.T., & van Wee, B. (2004). Accessibility Evaluation of Land-Use and Transport Strategies: Review and Research Directions. Journal of Transport Geography, 12(2), 127-140. URL: https://dx.doi.org/10.1016/j.jtrangeo.2003.10.005.

• Handy, S.L., & Niemeier, D.A. (1997). Measuring Accessibility: An Exploration of Issues and Alternatives. Environment and Planning A, 29(7), 1175-1194. URL: https://dx.doi.org/10.1068/a291175.

• Hansen, W. G. (1959), ‘How accessibility shapes land use’, Journal of the American Institute of Planners 25(2), 73–76.

• Hanson, S. (1986), The Geography of Urban Transportation, Guilford.

• Huang, A. (2013) Accessibility and non-work destination choice: A microscopic analysis of GPS travel data. Dissertation, University of Minnesota.

• Huang, J., Levinson, D., Wang, J., Zhou, J., & Wang, Z. J. (2018). Tracking job and housing dynamics with smartcard data. Proceedings of the National Academy of Sciences, 115(50), 12710-12715.

• Iacono, M. and Levinson, D. (2011), ‘Location, regional accessibility, and price effects: Evidence from home sales in Hennepin County, Minnesota’, Transportation Research Record 2245, 87–94.

• Iacono, M. and Levinson, D. (2017), ‘Accessibility dynamics and location premia: Do land values follow accessibility changes?’, Urban Studies 54(2), 364–381. URL: https://doi.org/10.1177/0042098015595012

• Iacono, M., Levinson, D. and El-Geneidy, A. (2008), ‘Models of transportation and land use change: A guide to the territory’, Journal of Planning Literature 22(4), 323–340.

• Lahoorpoor, B (2022) Terraces, Towers, Trams, and Trains : Examining the Growth of Sydney using Empirical Models and Agent-based Simulation. Dissertation, University of Sydney.

• Lahoorpoor, B. and Levinson, D. (2020), ‘Catchment if you can: The effect of station entrance and exit locations on accessibility’, Journal of Transport Geography 82, 102556.

• Levine, J., Grengs, J., & Merlin, L.A. (2019). From Mobility to Accessibility: Transforming Urban Transportation and Land-Use Planning. Ithaca, NY: Cornell University Press.

• Levinson, D. (1997), ‘The limits to growth management: development regulation in Montgomery County, Maryland’, Environment and Planning B: Planning and Design 24(5), 689–707.

• Levinson, D. (2013), Access Across America: Auto 2013, Technical report, University of Minnesota Center for Transportation Studies, Minneapolis, MN. URL: https://www.cts.umn.edu/publications/report/access-across-america-auto-2013

• Levinson, D. (2017). Spontaneous Access: Reflexions on Designing Cities and Transport. Network Design Lab, Sydney.

• Levinson, D. and Ermagun, A., eds (2022), Applications of Access, Network Design Lab, Sydney.

• Levinson, D., & King, D. (2019). A Political Economy of Access: Infrastructure, Networks, Cities, and Institutions. Network Design Lab, Sydney.

• Levinson, D. and Krizek, K., eds (2005), Access to Destinations, Elsevier, Amsterdam.

• Levinson, D., & Krizek, K. (2007). Planning for Place and Plexus: Metropolitan Land use and Transport. Routledge.

• Levinson, D. and Krizek, K. (2017), The End of Traffic and the Future of Access: A Roadmap to the New Transport Landscape, Network Design Lab, Sydney.

• Levinson, D., and Kumar, A. (1994a). The rational locator: why travel times have remained stable. Journal of the American Planning Association, 60(3), 319-332.

• Levinson, D. M., & Kumar, A. (1994b). A multi-modal trip distribution model. Transportation Research Record, 1466, 124-131.

• Levinson, D., Marion, B., Owen, A., & Cui, M. (2017). The City is flatter: Changing patterns of job and labor access. Cities, 60, 124-138.

• Levinson, D., Marshall, W., & Axhausen, K. (2017). Elements of access: Transport planning for engineers, transport engineering for planners. Network Design Lab, Sydney.

• Levinson, D. and Wu, H. (2020), ‘Towards a general theory of access’, Journal of Transport and Land Use 13(1), 129–158.

• Mann, I. and Levinson, D. (2024), ‘Access-based cost-benefit analysis’, Journal of Transport Geography 119, 103952.

• Martens, K. (2016). Transport Justice: Designing Fair Transportation Systems. Routledge. URL: https://dx.doi.org/10.4324/9781315746852.

• Martens, K. (2019), Why accessibility measurement is not merely an option, but an absolute necessity, in ‘Designing Accessibility Instruments’, Routledge, pp. 37–51.

• Melo, P. C., Graham, D. J., Levinson, D. and Aarabi, S. (2017), ‘Agglomeration, accessibility and productivity: Evidence for large metropolitan areas in the US’, Urban Studies 54(1), 179–195.

• Moreno, C., Allam, Z., Chabaud, D., Gall, C., & Pratlong, F. (2021). Introducing the “15-Minute City”: Sustainability, Resilience and Place Identity in Future Post-Pandemic Cities. Smart Cities, 4(1), 93-111. URL: https://dx.doi.org/10.3390/smartcities4010006.

• Owen, A. and Levinson, D. M. (2015), ‘Modeling the commute mode share of transit using continuous accessibility to jobs’, Transportation Research Part A: Policy and Practice 74, 110–122.

• Palmateer, C. (2018) The Distribution of Access. Dissertation, University of Minnesota.

• Rayaprolu, Hema (2023) The Co-evolution of Public Transport Access and Ridership. Dissertation, University of Sydney.

• Schrank, D., Eisele, B., Lomax, T., & Bak, J. (2015). 2015 Urban Mobility Scorecard. Texas A&M Transportation Institute.

• Schoner, J. (2017) Mutually Reinforcing Relationships Between Bicycling Infrastructure and Demand. Dissertation, University of Minnesota.

• Sheller, M. and Urry, J. (2006), ‘The new mobilities paradigm’, Environment and Planning A 38(2), 207–226.

• Social Exclusion Unit (2003). Making the Connections: Final Report on Transport and Social Exclusion. London: UK Cabinet Office.

• Soja, E. W. (1989), Postmodern geographies: The reassertion of space in critical social theory, Verso.

• Villamizar Duarte, N., Ardila Pinto, A. M., López-García, D., & Hernández López, D.J., (2025), ‘The ‘accessibility turn’ in urban research: contributions from Latin America and the global south’, Area Development and Policy pp. 1–13.

• Wachs, M., & Kumagai, T.G. (1973). Physical Accessibility as a Social Indicator. Socio-Economic Planning Sciences, 7(5), 437-456. URL: https://dx.doi.org/10.1016/0038-0120(73)90041-4.

• Wang, Y. (2023) A Benefit/Cost Analysis of Benefit/Cost Analyses. Dissertation, University of Sydney.

• Wang, Y. and Levinson, D. M. (2022), ‘Time savings vs. access-based benefit assessment of New York’s Second Avenue Subway’, Journal of Benefit-Cost Analysis 13(3), 370–396.

• Wang, Y. and Levinson, D. M. (2023), ‘Access-based land value appreciation for assessing project benefits’, Journal of Transport and Land Use 16(1), 373–393.

• Warf, B. and Arias, S. (2008), The Spatial Turn: Interdisciplinary Perspectives, Vol. 26, Routledge.

• Wu, H., Avner, P., Boisjoly, G., Braga, C. K. V., El-Geneidy, A., Huang, J., Kerzhner, T., Murphy, B., Niedzielski, M. A., Pereira, R. H. M., Pritchard, J. P. and Levinson, D. (2021), ‘Urban access across the globe: an international comparison of different transport modes’, npj Urban Sustainability 1, 16. URL: https://www.nature.com/articles/s42949-021-00020-2

• Wu, H., Lee, J. B. and Levinson, D. (2023), ‘The node-place model, accessibility, and station level transit ridership’, Journal of Transport Geography 113, 103739.

• Wu, H. and Levinson, D. (2020), ‘Unifying access’, Transportation Research Part D: Transport and Environment 83, 102355.

• Wu, H., Levinson, D. and Owen, A. (2021), ‘Commute mode share and access to jobs across US metropolitan areas’, Environment and Planning B: Urban Analytics and City Science 48(4), 671–684.

• Xie, F. and Levinson, D. (2011), Evolving Transportation Networks, Springer, New York.

Footnotes

1. I never finished the degree in planning — I am a proud planning school dropout — I was hired as a summer intern and soon working permanently at the Maryland-National Capital Park and Planning Commission - Montgomery County Planning Department, primarily as a transportation modeller, supporting the development and application of the `four-step' urban strategic transport model Travel, and later Travel/2, based on the EMME/2 software platform.

2. Susan Hanson, was, of course, a student of Bill Garrison, who I would meet later.

3. Hansen’s formula says that the accessibility A_i at location i is the sum of all opportunities O_j (say, jobs or services) reachable at other locations j, weighted by a decay function of distance d_{ij}^{-b}

4. See Cultural Turn: Cloke (2005); Mobility Turn: Sheller and Urry (2006), Spatial Turn: Soja (1989), Warf and Arias (2008); Accessibility Turn: Martens (2019), Crozet (2020) Villamizar et al. (2025).

5. Measured using travel times from model outputs and discounted using a decay formulation from the destination choice component of the County’s strategic model Travel/2.

6. If care is not taken, Simpson's Paradox may arise with this measure when Mode Share is allowed to vary as well as LOS. One imagines an improved Transit LOS would increase Transit Mode Shares, and lower Auto Mode Share. If TransitLOS starts off, and remains, worse than AutoLOS, this could lower the overall TTLOS since the better LOS now has a lower weight. Fixing the Mode Shares resolves this in the short run. This is discussed in Levinson and Wu (2020).

7. My most widely cited paper, notably rejected three times before acceptance. The ideas were born in the Montgomery County Planning Department, but published 4 years after I left.

8. Specifically, this was the recommendation of Tim Henkel, eventually Assistant Commissioner at MnDOT.

9. While it was something that was kicking around, I was motivated to execute on the Journal of Transport and Land Use while on sabbatical at Imperial College London. I was sharing an office with Stephane Hess, who was in the process of launching the Journal of Choice Modelling. And if he could do it, I could do it.

FIN