🏙️ Bản đọc song ngữ + Câu hỏi tiếng Việt — NODE10 Đô thị hoá & Giao thông

The most direct remedy is deliberately dispersing people and businesses away from an overcrowded capital city. Left unmanaged, rapid population growth in a capital city such as Seoul carries real consequences of its own: housing prices climb sharply, daily commutes lengthen as roads become congested, and healthcare and education systems can struggle to keep pace with an abruptly expanding population. These pressures are not merely abstract: in Dhaka, decades of largely unplanned growth have pushed peak-hour traffic down to a crawl on many arterial roads, while informal settlements have expanded far faster than the water, sanitation and waste systems built to serve them. Quality of life in a swelling capital can also be protected without relocating anyone at all: Curitiba, in Brazil, built its bus rapid transit network and linked it tightly to land-use zoning decades before its population doubled, while raising green space per resident from roughly one square metre to more than fifty, so that congestion and overcrowding there have stayed far milder than in comparable fast-growing cities that added transport and parks only after gridlock had already set in. South Korea offers a genuine, large-scale test of exactly this idea. Beginning in 2005, the government undertook construction of an entirely new administrative city, Sejong, specifically designed to relieve pressure on Seoul. More than 36 government departments have since relocated there, and by 2024 Sejong had grown to nearly 400,000 residents, supporting a multi-billion-dollar local economy. This demonstrates that deliberately relocating government functions and jobs away from an overcrowded capital can genuinely succeed — but typically only when pursued at a large, coordinated, government-led scale, rather than through the more modest approach of simply encouraging a handful of individual businesses to relocate independently. This approach carries a real cost as well, and even Sejong itself illustrates it. Critics describe Sejong as a "ghost city," since many workers simply commute back and forth, leaving their families behind in Seoul rather than actually relocating, and important institutions such as the National Assembly have yet to move there. This points toward a deeper cause underlying overcrowded capitals in the first place: cities expand so rapidly largely because people migrate there from rural areas in pursuit of superior jobs, wages, and schooling — a shift that has already drawn more than half the world's population into cities. Simply constructing a new city does not automatically eliminate that pull unless the new city can genuinely rival the jobs, culture, and opportunity of the old one.

The case for moving businesses outward has strengthened for a reason unrelated to planning policy, however: cost. Housing has become so expensive in the largest cities that high-cost regions have been steadily losing residents to cheaper ones. In 2024, it was small towns and rural counties that gained young adults. A relocating company is therefore not asking its staff to sacrifice anything — the same salary buys a house rather than a room, which is why firms that move frequently report lower staff turnover alongside lower rent. The disadvantage is equally concrete: a small town's schools, clinics and roads are built for the population it already has. Incoming money can then push local house prices beyond the reach of the families who grew up there. Bozeman, in the US state of Montana, illustrates exactly how sharply this can play out: as remote workers and relocating professionals arrived during the early 2020s, the median home price in the town roughly doubled within only a few years, pushing ordinary teachers, nurses and shop workers out of a housing market their own wages could no longer support. Local schools and clinics, sized for a much smaller population, then had to cope with waiting lists that simply had not existed a decade earlier.

A related question asks whether building more roads and highways is really the most effective way to address congestion, compared with putting the same money into rail and public transport instead. On the road-building side, the accumulated evidence points consistently in one direction. A detailed study of 30 counties in California found that every 10 percent increase in road capacity produced roughly 9 percent more traffic within just four years, almost entirely cancelling out the capacity that had been built. Studies across Europe, Japan, and North America identify a strikingly similar pattern: roughly a 1 percent increase in road capacity generates approximately a 1 percent increase in total vehicle miles travelled. Economists and traffic engineers call this "induced demand". Wider roads do not simply accommodate existing traffic more efficiently; they actively generate new car journeys that would not otherwise have taken place, which is precisely why building additional roads rarely resolves congestion for any lasting period.

None of this, however, settles the question the way it first appears to, because "spending on roads" is not the same as "spending on cars". Buses, coaches, cycles, ambulances, and delivery vans all run on the same tarmac, and in most countries the cheapest way to improve public transport quickly is a bus lane painted on a road that already exists. Nor can rail reach everywhere. A railway needs enough passengers per kilometre to justify itself, which dense corridors between large cities can supply and dispersed rural regions generally cannot. Almost every journey, and almost every delivery, still begins or ends on a road.

The cost comparison complicates matters further. Britain's HS2 line was estimated at 35 to 45 billion pounds in 2019; the official range is now 87.7 to 102.7 billion pounds. A single rail project of that size absorbs money that could have maintained the network the whole population uses daily, and that network is deteriorating. The annual ALARM survey puts the local road repair backlog in England and Wales at a record 18.6 billion pounds, roughly twelve years of work at current funding levels. Deferring maintenance is not a saving; it simply moves a larger bill into the future.

The defensible reading is therefore narrower than the slogan. Induced demand is a powerful argument against widening roads to cure congestion, but not an argument against roads themselves, and not a guarantee that any given rail project is the better use of the same money. The real choice is usually between expansion and maintenance, and between corridors dense enough to fill trains and regions that never will be.

A different kind of road safety question asks whether simply raising the minimum legal age for driving would genuinely make roads safer. The evidence in favour looks considerable: sixteen-year-old drivers crash roughly ten times more often, per mile driven, than middle-aged adults, and when one US state raised its licensing age from 16 to 17, fatal crashes among young drivers fell by 13 percent. Across the United States more broadly, the death rate for sixteen-year-old drivers fell by 68 percent between 1996 and 2010, largely attributable to the wider adoption of graduated licensing systems that phase in full driving privileges incrementally. Some researchers contend, however, that the underlying cause is not age itself but inexperience: these graduated systems typically also mandate dozens of hours of supervised practice and restrict night driving and teenage passengers, and the evidence suggests it is these specific practice and supervision requirements, rather than the age threshold by itself, that actually reduce crashes. This implies that simply raising the legal age, absent any accompanying supervised-practice requirement, might merely displace the same high-risk first year of driving to a somewhat later age rather than genuinely eliminating the underlying risk.

A related but considerably bolder claim holds that stricter punishment constitutes the only way to improve road safety. Criminal justice researchers have examined this exact question for decades, and the accumulated evidence points in a rather different direction: what actually deters most people from breaking a rule is not the severity of the punishment but the likelihood of actually being caught. A clear real-world illustration is the speed camera. A substantial review of 28 separate studies found that speed cameras installed at dangerous locations reduced crashes there by up to 50 percent, and reduced serious injury or fatal crashes by up to 44 percent, not through harsher penalties for speeding but simply by making drivers believe they were considerably more likely to be caught if they sped. In New York City, installing speed cameras reduced the number of drivers caught speeding by 72 percent and reduced crashes by 14 percent, without any alteration whatsoever to the severity of the fine itself. This does not imply that punishment is irrelevant, but it does suggest that "much stricter punishment" is highly unlikely to constitute the only genuine answer, when simply increasing the likelihood of apprehension already accomplishes a considerable share of the work unaided. A closely related question applies not merely to drivers but to lawbreakers more broadly: once an offender has been apprehended, does simply committing them to prison accomplish more than requiring supervised, skill-building work instead? Here too, the country-level evidence points in the same direction — altering behaviour matters more than harsh punishment in isolation. In the United States, which relies predominantly on conventional imprisonment, approximately 82 percent of individuals released from state prison are rearrested within 10 years. Norway, which places considerably greater emphasis on cultivating genuine skills while an individual serves their sentence, has an overall reoffending rate of only around 20 percent. A specific American initiative called Vocational Village, which teaches prisoners a skilled trade such as welding rather than merely confining them, found that its graduates were rearrested at a rate of only 12 percent within three years, compared with roughly 23 percent for prisoners generally. This does not imply that prison should be abolished — certain crimes are serious enough that it remains necessary to safeguard the public — but the broader evidence suggests that teaching a genuinely useful skill does more to prevent reoffending than punishment alone. None of this means severity never matters. Researchers normally separate deterrence into three components: the certainty of being caught, the severity of the penalty, and the speed with which it follows. Speed cameras operate on the first. The second still does genuine work wherever the offence is a deliberate choice rather than a lapse. Disqualification for drink-driving both deters and physically prevents the same driver from repeating it, which no amount of road design can achieve. The same division applies to imprisonment. The evidence there is genuinely mixed rather than one-sided: studies in Norway and North Carolina have found that longer sentences can reduce reoffending. And for a small number of dangerous offenders, separation from the public remains the only measure that reliably protects anyone while it lasts. This concentration effect is well documented: a landmark long-term study tracking a cohort of nearly 10,000 boys born in Philadelphia found that a small group of chronic offenders — roughly six percent of the cohort — accounted for approximately half of all the crimes recorded, including the large majority of the most violent ones. For that specific group, removing them from general circulation is not merely punitive; it demonstrably prevents new victims for as long as the sentence lasts, an effect criminologists term incapacitation, distinct from either deterrence or rehabilitation.

Making a city function well is not only about controlling roads and crime, though — it is also about where people actually do their jobs each day.

Working from home has become considerably more widespread, and a genuine question follows closely behind it: does the practice primarily benefit employees, or does the employer actually capture the larger share of the gain? Detailed research by the economist Nicholas Bloom, based on a large-scale study of workers at a major travel company, found substantial benefits accruing on both sides simultaneously. Employees who worked from home two days a week were 13 percent more productive and 50 percent less likely to leave their jobs entirely. Employers benefited considerably as well: broader studies found that hybrid working from home raised average productivity by roughly 5 percent, while cutting staff turnover by approximately a third and reducing the amount of costly office space a company needs to maintain. This is a case where the honest answer is not "one side benefits at the other's expense," but rather that both sides genuinely benefit, through different, independently measurable mechanisms. Where people work is only part of how they move around a city, though — how they get there matters just as much.

Should cities primarily encourage cycling and denser urban living, rather than continuing to build additional space for private cars? Traffic congestion itself has a specific, direct driver: peak-hour road space is scarce but costs nothing to use, so demand for it keeps climbing until congestion becomes the only thing left to ration it. This is not merely theoretical — when American cities added new highway lanes, one influential study found that the total distance driven rose in almost exact proportion to the new capacity, so that congestion returned within only a few years, a pattern researchers call induced demand. One common reason cycling has declined in many cities is practical rather than cultural: many roads still lack protected, dedicated bike lanes, unfavourable weather discourages regular riders, and as urban sprawl has pushed housing further from employment centres, commuting distances have often grown too long for a comfortable ride. Two European cities offer a genuine, direct point of comparison. In Copenhagen, 49 percent of all trips to work or school are made by bicycle, supported by 850 kilometres of dedicated bike lanes — the city even clears snow from bike lanes before clearing roads intended for cars. Amsterdam, despite actually possessing more bicycles per resident than Copenhagen, records a slightly lower share of trips made by bike, around 38 to 40 percent, supported by 400 kilometres of bike paths. Both cities illustrate the same underlying point with unusual clarity: when a city combines high population density with serious, dedicated cycling infrastructure, cycling can genuinely displace a substantial share of car journeys. Density on its own and bike lanes merely painted onto existing roads are not equivalent interventions, however — the gap between the two cities shows how much the finer details of infrastructure matter in practice. Some cities go further still, holding regular car-free days that temporarily ban private vehicles from parts of the city centre and redirect residents toward buses, bicycles, or walking instead. Bogotá's weekly Ciclovía, for instance, closes major roads to cars every Sunday, and although this measurably reduces pollution and congestion for that single day, it does little to resolve the underlying, everyday problem in the way that permanent infrastructure such as Copenhagen's bike lanes does. Two objections are usually raised against all of this, and only one survives contact with the evidence. Retailers regularly fear that closing a street to cars will destroy their trade. Yet a 2016 study of more than 100 cities found that pedestrianised streets raised retail sales by roughly half: customers arriving on foot or by bicycle spend less per visit but return considerably more often. The objection that does survive concerns who gets left out — deliveries, ambulances, tradespeople carrying tools, disabled residents, and anyone living beyond the reach of a bus route still need a road. The same logic explains why simply encouraging people to move into cities rarely works. Where a person lives is anchored by a job, a school and a rent they can afford. Meanwhile the highest-cost cities have been losing residents to cheaper places rather than attracting them — precisely the opposite of what encouragement alone is supposed to achieve.

Bike lanes and car-free days are solutions built for the transport of today, though — a very different kind of promise has been made about the transport of tomorrow.

Self-driving cars have been described as "just a few years away" for well over a decade now, and the real history underlying these predictions rewards closer examination. In 2016, BMW promised a fully self-driving car by 2021; that promise remains unfulfilled to this day. In 2017, Ford's own chief executive promised a fully self-driving commercial vehicle by 2021, and years later Ford publicly acknowledged the company had been considerably too optimistic in its projections. Tesla promised fully self-driving software by the end of 2019, and separately, Lyft predicted the majority of its rides would be self-driving by 2021 — neither prediction materialised. The most recent expert surveys now suggest fully self-driving cars for ordinary private buyers may not become widely available until around 2035, and even that projection remains far from settled fact. The consistent pattern of missed predictions across company after company suggests genuine caution is warranted whenever anyone confidently states exactly when self-driving cars will finally arrive. Beyond the question of WHEN they will arrive, however, lies an entirely separate question: whether they constitute a truly good idea once they do. The strongest argument in favour of self-driving cars is safety: the US government's own traffic safety agency found human error to be a contributing factor in roughly 94 percent of serious crashes, and Waymo's self-driving vehicles have demonstrated up to 85 percent fewer crashes causing serious injury at intersections relative to an average human driver. The technology is not, however, yet unambiguously safer overall — official crash data currently indicates self-driving cars are actually involved in more crashes per mile travelled than conventional vehicles, since the software continues to struggle with certain unusual situations a human driver would navigate with ease. A substantial economic cost accompanies this as well: one industry report estimates that between 2 and 4.4 million truck-driving positions across the US and Europe alone could be eliminated should self-driving trucks become widely adopted. The honest answer, then, is neither simply "yes" nor "no" — today's safety record remains genuinely mixed, and the job losses are real, even as the long-term safety potential appears promising.

Finally, some cities have converted public parks into vegetable gardens, and a genuine historical case demonstrates this can succeed at a meaningfully large scale rather than functioning merely as a small, symbolic gesture. In the early 1990s, Cuba abruptly lost access to cheap fuel and agricultural supplies, and its economy contracted by 36 percent within just three years. In response, ordinary Cubans converted empty lots and open spaces across every city into small-scale urban farms. By 1996, urban farms in the capital, Havana, alone were producing 8,500 tonnes of food, 7.5 million eggs, and 3,650 tonnes of meat annually for city residents. This demonstrates that converting urban green space into food production is not simply an appealing idea in principle — under the right conditions, it can genuinely help feed a real city's population at a scale that matters. Converting parks into gardens is not, however, free of genuine drawbacks. A park typically serves an entire neighbourhood equitably — anyone can walk, play, or simply sit there — whereas a garden plot functions best for residents who actually possess the time, physical capacity, and horticultural knowledge required to keep it productive, meaning the benefit does not automatically reach everyone equally. Converting shared green space also eliminates playgrounds, running paths, and open lawns that families who do not garden nevertheless relied upon. This trade-off is not merely theoretical: in the United Kingdom, where local councils have offered small growing plots for generations, demand for a plot has long outstripped supply, and prospective gardeners in high-demand areas such as inner London now face waiting lists that stretch well beyond a decade before a single plot becomes available. Converting more parkland into gardens therefore does not simply add a new amenity; in practice it can lock ordinary shared land behind a scarce allocation system that only the most patient residents can ever access. The genuine question, therefore, is not simply whether food gardens can succeed at scale — Cuba clearly demonstrates they can — but whether a city retains sufficient shared, open green space for residents who will never cultivate a single vegetable.