There is a pleasing conceit, much loved by urbanists, that the city is a laboratory. The phrase has a respectable pedigree: in the early twentieth century, sociologists in Chicago sent their students into dance halls, boarding houses and street markets, convinced that the swelling city around them could be studied as methodically as a cell under a microscope. A century later the metaphor has been revived with fresh enthusiasm. Mayors announce 'pilots', planners speak of 'prototyping' a street, and whole districts are fitted with sensors so that traffic, noise and footfall can be measured by the minute.
The appeal is easy to understand. Cities concentrate problems, but they also concentrate the means of solving them: people, money, ideas and, above all, proximity. A national government that wishes to try a new approach to, say, recycling must legislate for millions at once, and a mistake is correspondingly costly. A city can close one road to cars for a summer, watch what happens and reopen it in September if the trial fails. Failure on this scale is cheap, visible and, crucially, reversible. It is no accident that many ideas now taken for granted, from the bicycle-sharing scheme to the pedestrianised high street, began life as local gambles before being copied elsewhere.
Yet the metaphor deserves closer scrutiny than it usually receives, because laboratories have features that cities conspicuously lack. A laboratory has controls: a second, untreated sample against which the effect of a change can be judged. Cities have no twin. When cycling increases after a new lane is painted, was it the lane, the fine weather, the price of petrol or a popular television programme about fitness? The honest answer is often that nobody can be sure, and the clamour for quick results tempts officials to claim more than the evidence will bear.
A laboratory also has subjects who have agreed to take part. The residents of an experimental street have, for the most part, agreed to nothing. Consider a shopkeeper, call her Mrs Okafor, who has run a hardware shop on the same corner for thirty years. She did not volunteer to become a data point when her road was closed to through traffic, and she may reasonably feel that the costs of the trial, fewer passing customers and a longer route for her deliveries, fall on her, while the benefits are enjoyed by people who merely pass through. To call such people 'participants' is to stretch the word until it snaps.
Defenders of urban experimentation have answers to these objections, and some of them are good. Ethan Marsh, who advised a northern English city on its programme of temporary street changes, argues that the alternative to experiment is not caution but dogma: decisions taken on the strength of a consultant's report or an official's instinct, which are then defended long after they have proved wrong precisely because they were never framed as tests. 'An experiment,' he says, 'is a promise to listen. A permanent scheme announced with fanfare is a promise not to.' On this view, the provisional character of a pilot is not a weakness but a form of humility.
Others accept the logic but question who holds the clipboard. When the sensors on lampposts belong to a private company, the city's experiments may serve the company's ends, such as the refinement of a product to sell to other councils, as much as the residents'. The data gathered can be opaque to the very people it describes. Here the laboratory metaphor reveals something its enthusiasts would rather it did not: in a laboratory, the subjects rarely see the results.
Perhaps the most useful response is not to abandon the metaphor but to take it more seriously. Good science publishes its methods, admits its uncertainties and reports its failures as well as its triumphs. A city that wished to deserve the title of laboratory would do the same: explain in advance what a trial is meant to show and how success will be judged, share the data in a form ordinary people can read, and give those most affected a genuine voice in deciding whether an experiment continues. That is a more demanding standard than the one usually applied, and few cities meet it in full.
There is, finally, an older sense in which cities have always been laboratories, one that owes nothing to sensors or pilots. For thousands of years, people of different origins, trades and beliefs have been thrown together in crowded streets and obliged to find ways of living alongside one another. Every market, every shared courtyard, every argument over noise or rubbish has been a small experiment in coexistence, conducted without a hypothesis and recorded nowhere. Next to that vast, unplanned enquiry, the latest flagship trial looks rather modest, which may be no bad thing. The city's greatest experiment is, and always was, its citizens themselves.