Who May Stand Where: The Governing Of Street Food

AMore people eat their midday meal standing up than the architecture of any city would suggest. Estimates compiled for the International Labour Office in 2019 put the number of people earning a living from street food at somewhere above forty million, most of them women, and almost all of them working outside the tax and inspection systems that govern a restaurant on the same street. Municipal authorities have responded with an instrument of striking uniformity: the licence. A vendor registers, pays an annual fee, accepts a fixed pitch and a set of hygiene conditions, and in exchange acquires the right not to be moved on. Public health officials present this as the conversion of an ungoverned trade into a governed one. Their critics reply that the licence governs the vendor rather than the food, and that what is actually being regulated is who may stand where.

BOlivia Halim, a food-safety epidemiologist at the Corbett Institute of Urban Health, has spent much of her career asking whether the licence does what is claimed for it. Between 2015 and 2021 her group took swab samples from prepared food sold at 3,100 stalls across fourteen cities, half of them licensed and half not, and matched the results against each city's own inspection records. Licensed stalls returned unsafe bacterial counts in 18 per cent of samples, against 26 per cent for unlicensed ones. The gap, however, was produced almost entirely by four cities in which an inspector had actually visited the pitch within the previous year; in the ten cities where licences were issued and then forgotten, the difference shrank to a statistically meaningless two percentage points. Her conclusion pleases neither camp. The licence, she argues, is not a hygiene measure but a filing system, and a filing system improves nothing until somebody reads the file.

CThat conclusion carries a practical warning, and Halim has tested it. Many cities pair licensing with relocation, moving vendors off the pavement and into purpose-built food centres where water and refrigeration can be supplied. In 2022 her team followed 1,900 vendors through such a move in three cities, recording takings before and afterwards. Bacterial counts fell in every centre. Earnings did not behave so simply: where the new site lay within roughly two hundred metres of the vendor's old pitch, weekly takings recovered to their previous level within four months; where the move was longer, takings were still a third below the old figure a year later, and one vendor in five had abandoned the centre for the pavement again. The determining variable, Halim notes, was not the quality of the building but the distance a customer had to walk to reach it.

DClara Sepulveda, an urban economist at the Halberd School of Planning, accepts Halim's measurements and disputes the lesson drawn from them. A licensing regime, he argues, is never fairly judged against an imagined orderly market; it is judged against the arrangement it displaces, and that arrangement was rarely the free trade of description. Unlicensed vendors, in his account, already paid — to police officers, to gangmasters, to whoever controlled the kerb — and paid more, and received nothing in return that could be pointed to. Surveying 63 cities that introduced fee-based licensing between 2010 and 2020, he calculated that the sums vendors reported handing over informally fell from an average of 11 per cent of turnover to 4 per cent, while the licence fee itself averaged 3 per cent. Sepulveda volunteers two limitations. His figures rest on what vendors told interviewers about payments they had every reason not to describe accurately; and he measured the vendors still trading, not those whom the fee had driven out.

EThe dispute has left the seminar room. A statute adopted in the city of Santa Ferra in 2019 requires any authority removing a vendor to offer an alternative pitch in the same district, and to publish each year both the number of removals carried out and the number of alternatives actually taken up. Sepulveda regards the publication requirement as the more valuable half, because a figure that must be printed can afterwards be checked by people the authority did not choose. Halim, whose findings are the more often quoted by campaigners, is here the more doubtful of the two: a duty to offer a pitch, she observes, says nothing about whether the pitch has water, and a city may comply with the statute in full while moving a trade into premises where hygiene is harder to maintain than on the pavement it left. Both object to the habit, common in official reports, of counting licences issued as though the number were itself an outcome.

FWhat is emerging is a division of the question rather than an answer to it. Where the hazard can be named and measured — a broken cold chain, a water supply shared with a drain, oil reused for a fortnight — rules work in the ordinary way: they specify a standard, an inspector observes whether it is met, and the effect shows up in the samples. Where the complaint is congestion, or the appearance of a boulevard, the same instrument is doing something else entirely, and calling it public health obscures the fact that a decision about space is being taken. Whether municipal motive is shifting from the first of these towards the second is disputed, and neither researcher claims to know. What both reject is the framing that has dominated public argument: that a city must choose between safe food and a living for the people who sell it.