As a young child in Glasgow I was desperate to visit the United States, to see its incredible landscapes and its legendary urbanism: the Grand Canyon, the Manhattan skyline. But it wasn’t until visiting much later that I experienced what is truly the iconic American landscape: the strip, that stretch of multi-lane road leading off into the distance, surrounded on either side by fast-food restaurants, islands of retail lost in seas of asphalt.

Strip development, and its cousin the shopping mall, are symbols of America’s gift to urbanism: sprawl. Los Angeles may be the world’s most famously sprawling city but is it the worst culprit? What about Montreal, or Brisbane, both low density cities in countries with no shortage of space and a strong love of the car?

Sprawl, even though we know when we see it, proves extremely difficult to pin down into a functional definition. Urban sprawl is usually huge, mainly low-density, mostly unplanned, and primarily residential development that covers increasing areas of land around city cores. It is suburbia on steroids or, as Robert Kirkman’s The Ethics of Metropolitan Growth puts it: “sprawl suggests the city has collapsed, like a drunkard on a sidewalk, and is now spreading inexorably outwards, oblivious to the surrounding countryside”.

The first and most important common factors is the car. After the second world war, American developers took advantage of cheap oil and the personal mobility made possible by car ownership to create low density residential developments that were not contiguous with places of work, commerce and leisure. These single-use areas often consist of detached houses in the centre of a lawn, scattered over the landscape, often unevenly, with residual gaps of undeveloped land between them and the city.

Although the first use of the term “urban sprawl” appears to be in the context of London in 1955, what we know as sprawl is a quintessentially American invention – it is aspirational, a culture of free individualism, of conspicuous wealth, and requires a seemingly limitless supply of land and resources. To live in sprawl means driving to work, driving to get dinner, driving to meet your friends. It means congestion, as its inhabitants travel ludicrous distances for work or basic services, and isolation due to the lack of the perks of compact city life. Sprawl eats up huge areas in very inefficient ways, destroying arable land and creating monocultures. Furthermore, a sprawling metropolis generates vastly greater amounts of pollution and CO2 than a more compact one.

There is no reliable combined measure of sprawl, and many geographers shy away from using the term because of its negative connotations. Perhaps the most reliable metric comes from looking at population density. To look at the whole world is difficult, as definitions of metropolitan boundaries don’t always match up, but free-market think tank Demographia’s annual World Urban Areas survey attempts just that.

The results seem to tally with what we expect: American areas completely dominate the low densities, and among many smaller conurbations the lowest density large cities are Atlanta, Boston, St Louis, Orlando and San Juan. The only real competition the Americans have is with places like Brisbane, Australia, or Quebec City, Canada, both countries with a lot of land and a love of the car. In Europe, down there in the low densities are a number of French urban areas such as Nantes and Toulon, but little else, while the Middle East sneaks in with Saudi Arabia’s Ad Dammam. According to Demographia’s metric, East Asian and South American cities tend not to sprawl at all.

The problem with an average value though, is that it mistakes particular characteristics. For example, London and Athens have very similar average densities, but the Greek city has a far denser core and a much more sprawling suburbia. Some researchers try to overcome these difficulties through additional metrics. Researchers for Smart Growth America introduced mixes of use, presence of “centres”, and accessibility as other factors. Their research was pretty damning for the southern states, with Atlanta, Nashville and Memphis to the bottom of the list.

Elsewhere, Thomas Laidley of NYU recently created a Sprawl Index using aerial imagery. One remarkable aspect was the discovery that Los Angeles is now the densest urban centre in the US. The city which is usually considered the ultimate in sprawl has reached such a consistent intensity of development that there is very little spare land of any kind that could lower its score. Laidley’s index once again gives the worst scores to cities in the deep south: Columbia in South Carolina and Hickory in North Carolina.

Is sprawl so bad?

If it can’t be carefully defined, is sprawl necessarily bad? The great urbanist Lewis Mumford thought so, arguing hysterically in The City in History in 1961 that the endpoint of sprawling megacities was “total human annihilation”, falling victim to technical rationality and the end of organic human existence. Later, characters as ideologically different as the New Urbanists and Richard Rogers would advocate dense, walkable cities with active street life, an anti-sprawl attitude that is now standard in much planning, at least in Britain.

Even as it stands on the cusp of change, the ultimate sprawling city still has to be Los Angeles. Think of the classic view from the Griffith Observatory, looking down at that vast carpet of concrete with its threads of light from the clogged highways. LA is decentred, potentially limitless, and stands for everything terrible about what happens when cities are developed without planning: swathes of low density housing, completely severed by roads, the whole terrain plagued by filthy smog.

Architecture critic Christopher Hawthorne talks of a ‘first’ Los Angeles at the beginning of the 20th century, when it grew much like a European city, before the ‘second’ LA happened post-war: mass-transit was eliminated, the freeway took over, suburbia raced to the edge of the mountains. This created the LA that everyone loves to hate, the “72 suburbs in search of a city” in the words of Dorothy Parker. The ‘third’ LA, which is developing now, is in Hawthorne’s words “having to re-learn the art of sharing the city,” by intensifying its civic culture within its own boundaries, becoming a haven for artists, constructing more collective housing, and with new transport infrastructure that includes walking and cycling.

Others have defended sprawl on its own terms. Architectural historian Reyner Banham wrote the seminal Los Angeles: City of the Four Ecologies in 1972. Excited by its unapologetic modernity, Banham argued that the sprawling quality of LA offered “radical alternatives” to, not deviations from, what was accepted wisdom in urbanism. This refrain was taken up by Jonathan Meades in his 2012 series On France, where he noted sprawl’s prevalence, and mounted a muted defence: “It is not vilified. It is nothing to be ashamed of. Nor is pride taken in it as pride is taken in conventionally elegant towns and orthodoxly pretty villages. Sprawl is accepted, it’s there.”

At the end of the day, sprawl is probably best described as the antithesis of an ideal city type – the compact, efficient pedestrian settlement with a rich civic culture – and thus is an essentially negative term. Advocates for free markets may well argue that sprawl is simply the free and natural result of people’s desire for their own space but in the future, with massive urban growth still to come, is it materially or ecologically possible to sustain 20th century petrol-soaked American sprawl across the world?