Re-Examining Urbanisation as a Key Driver of Future Air Taxi Demand

In assembling research for our forthcoming report, “The Future of Advanced Air Mobility”, it became clear that there is a consistent narrative underpinning the public communications of passenger AAM developers like Joby, Archer, Eve, Supernal, and Vertical Aerospace. These companies frequently cite a long-term UN projection that around 68% of the global population will live in urban areas by 2050, up from roughly 57% today, and argue that rising urban density will place unsustainable pressure on ground transport networks – thereby creating the need for widespread Urban Air Mobility (UAM). The implication of this is twofold: that eVTOL aircraft will become a form of essential infrastructure, and that annual production volumes must reach several thousand units per year to satisfy the expected demand. 

Figure 1: Stated Production Capacity Targets by OEM 

Source: Valour Consultancy 

While urbanisation is a real and measurable trend, the conclusion that it will translate into large-scale demand for passenger eVTOL services in major cities is far less certain. A number of structural factors complicate the narrative: 

1. First, many cities are already pursuing both major extensions to existing metro systems and the construction of entirely new high-capacity transit networks – often at orders of magnitude lower cost per passenger-mile than air taxi operations are likely to achieve. This includes large-scale network expansions such as the Grand Paris Express, New York’s Second Avenue Subway, Dubai’s Blue Line, and substantial additions across Beijing, Mumbai and Sydney. At the same time, several cities have recently introduced entirely new systems, including the Riyadh Metro and the Lagos Blue Line. In practical terms, these developments represent the primary scalable mechanism for accommodating growth in urban travel demand, while eVTOL services are likely to play a more targeted, supplemental role, focussed on specific corridors, premium use cases and time-sensitive journeys, particularly in the near to medium term. 
2. Second, several cities frequently highlighted by OEMs as early AAM markets such as New York, San Francisco, Los Angeles, Tokyo, and several major Chinese urban regions are not where the strongest urban population growth is occurring. Recent data shows that New York City has lost approximately 500,000 residents since 2020, while San Francisco’s population remains ~7% below 2020 levels and Los Angeles has experienced consecutive annual net outflows. In Tokyo, overall population peaked in 2020 and has since edged into decline, and both Shanghai and Beijing have recorded year-on-year population decreases since 2021. Most relevant to the air taxi market, this trend is also pronounced among the commuting-age population (roughly 25 to 54), reflecting ageing population profiles and lower fertility rates rather than temporary shifts. By contrast, the fastest-growing urban centres over the coming decades are concentrated in regions with significantly lower average incomes, such as Lagos, Kinshasa, Dar es Salaam, Dhaka and Karachi. Thus, while the urban population will certainly grow,   demand will be constrained by limited ability to pay for premium, discretionary air mobility services. 
3. Third, travel behaviour itself is changing. Public transport ridership in many major cities remains significantly below pre-COVID levels, reflecting a sustained shift to hybrid and remote work patterns. For example, New York City Subway ridership continues to average around 70–75% of 2019 levels, while that of San Francisco’s BART system remains closer to 50–60%. Even in high-utilisation systems such as London Underground, weekday ridership has stabilised at approximately 80–85% of pre-pandemic volumes, with peak-hour intensity materially reduced. At the same time, the commuter base itself is likely to shrink. The International Monetary Fund (IMF) estimates that around 40% of jobs in advanced economies are exposed to AI-driven automation, particularly in office-based roles. This suggests that fewer people may need to travel to fixed workplaces each day and, as a result, the intensity of peak-hour congestion that eVTOL networks are intended to address may be considerably lower than OEMs assume. And while urbanisation in developing economies will expand service-sector employment over time, much of this growth is expected to be informal or low-income, and therefore unlikely to translate into large office-commuting surges. 
4. Fourth, advances in road vehicle automation are likely to improve the efficiency of surface transport over time. Congestion in dense urban environments is not solely a function of vehicle volume, but also of human driving behaviour, including reaction times, inconsistent lane changes, braking patterns and incident-driven slowdowns. Autonomous and highly automated vehicle systems are designed to reduce these inefficiencies through coordinated acceleration, smoother car-following and network-level traffic flow optimisation. Waymo’s San Francisco robotaxi fleet, for example, has demonstrated more consistent headways and substantially less variability in acceleration and braking than human drivers, which are the key behavioural mechanisms that contribute to stop-and-go congestion. Several others, including Tesla with fully autonomous electric vehicle, Cybercab, are pursuing high-volume, on-demand ride services. As these systems scale, the baseline level of road congestion may fall, reducing the conditions under which short-range aerial transport offers a meaningful time advantage. Moreover, many cities are simultaneously pursuing policies to reduce private car use, through measures such as the ULEZ expansion in London, reduced parking provision and cycling-first street design. 
5. Fifth, a meaningful share of urban road traffic consists of short, last-mile delivery trips, including food, grocery and small-parcel movements. These trips contribute disproportionately to congestion due to frequent stopping, parking friction and kerbside loading activity. If uncrewed delivery aircraft scale in line with ongoing trials and early commercial deployments, a portion of these road-based movements may be displaced. Pilot programmes in different regions already demonstrate the feasibility of routine last-mile drone delivery for medical supplies, food and convenience goods, and several national regulators are moving toward broader approvals for BVLOS operations. Even modest uptake would reduce the number of low-value, high-frequency road journeys in dense areas, easing congestion. This would further lower the severity of the travel inefficiencies that passenger AAM services are often positioned to address. 

In conclusion, rising urbanisation does not, in itself, imply a need for widescale deployment of passenger AAM. The assumption that growing city populations will inevitably lead to severe congestion, and therefore sustained demand for high-frequency aerial mobility, overlooks several structural trends: cities are expanding high-capacity public transit, urban populations in many target markets are stabilising or declining, daily commuting is becoming less concentrated, and road networks are expected to operate more efficiently as automation advances. The congestion problem that underpins many eVTOL demand narratives is therefore not fixed, and will likely not intensify in the way commonly assumed. 

Valour Consultancy’s forthcoming soon to publish report, The Future of Advanced Air Mobility, explores in depth the trends shaping the development of five key AAM use cases: passenger, cargo, specialist, delivery, and inspection/surveillance. The report also includes detailed forecasts on annual deliveries and the in-service fleet out to 2050, underpinned by robust and realistic assumptions. Drawing on extensive primary research and grounded market analysis, it provides a clear-eyed assessment of where genuine opportunities exist, and where expectations may be overstated