How You Define a Street Vendor May Decide the Policy You Get

WHEN police approach a tea table in Dhaka, some male street vendors have been documented ducking underneath it while their wives step in to keep selling, because officers there are reluctant to arrest women. Who counts as informal in a scene like that? A systematic scoping review of 327 studies from 64 countries found that fewer than a quarter of the studies it analyzed said what they meant by informal food vending, and its authors argue that the definition researchers pick (by location, by stall, by license) steers the policy answer before any data arrive. They call vendors like these a linchpin of food security and livelihoods across low-income and middle-income countries, which makes the muddle more than academic.

Food systems employ an estimated 1.23 billion people worldwide, and a majority of those jobs fall outside licensing and registration. Yet when a team spanning the Alliance of Bioversity International and the International Center for Tropical Agriculture (CIAT), Cornell University and other institutions went looking for what researchers actually know about those workers, they found a literature that was big, scattered and oddly vague about its own subject.

Nobody Agrees What Informal Means

The team began with 23,250 records, drawn from academic databases and repositories of theses and reports, and whittled them down to 1,889 for full-text reading and finally 327 studies of food vendors in low-income and middle-income countries, published in English between 2000 and May 2023. Seven of the 327 were later dropped as very poor quality. That leaves 320, the pool the synthesis really draws on, even though the release and the abstract headline the bigger number. Most of the 320 held up reasonably well on the review’s quality checklist, with 277 rated good, 36 moderate and 7 low.

Fewer than 80 of those 320 spelled out a definition at all, and the ones that did sorted into five camps: the International Labor Organization version (unregistered, unlicensed, no social protection), a fixed or missing stall, the type of product sold, the street or settlement where it is sold, and a handful of researcher-built definitions drawn from the data. The International Labor Organization version was the most common, location-based definitions came next, and the categories overlap.

Measured against the three ingredients of that definition, the literature leans on one. Some 83% of studies mentioned goods made for the market, 23 percent mentioned private enterprises, and only 10% touched on whether vendors have any social protection. So the component that asks whether a vendor is protected at all is the one the literature asks about least.

Mark Lundy, a market and community development specialist at the Alliance of Bioversity International and CIAT and a co-author, puts the stakes bluntly: “The use of any of these definitions induces certain policy conclusions before the data comes in.” A location definition produces place-based policy, he says, an infrastructure one produces market upgrades, and a licensing one produces formalization drives, which earlier work the authors cite says have largely failed. That’s partly because registration and compliance costs are rarely designed around how informal work actually runs.

Where Selling Happens, and Who Sells

In the studies reviewed, most informal food vendors were women, and gender dynamics was the most studied sociopolitical factor (74 studies), well ahead of governance and regulation (43) and access to public space (29). One study of Tanzanian fish markets, as the review describes it, found smoked and dried fish traders, nearly all men, working in the preferred covered areas, while the women selling deep-fried fish worked from small movable cabinets in a separate part of the market. Harassment, gender-based violence and the seizure of goods were reported to shape what women and men sold, and when.

Swetha Manohar, a public health nutritionist at the International Food Policy Research Institute and a co-author, sums up the pattern this way: “Inequality here is about where selling happens as much as who sells.” Bribes, personal safety, migration and childcare, all topics the authors flag as ripe for more work, each turned up in fewer than 20 studies.

On the food side, 200 of the studies touched on nutrition, mostly by describing what vendors sell (140 studies) or how customers choose (93). The studies describe a paradox: a place of traditional foods and biodiversity on one hand, and of easy access to nutrient-poor, energy-dense ultra-processed foods (and, in places, illegal bushmeat) on the other. Prepared foods dominated the research, at 109 studies, with raw vegetables next at 81.

What the literature mostly can’t say is whether any of this shows up on a plate, since the studies tend to track what people choose rather than diet quality or nutrition outcomes, and very few looked at the health or diets of the vendors themselves. Ramya Ambikapathi, a nutrition epidemiologist at the Alliance of Bioversity International and CIAT and the paper’s first author, puts it plainly: “So much of what we eat is shaped by vendors; a closer look into their livelihoods, their motivations, their supply, their diets, and wellbeing is an important gap.”

The evidence is also thin on time. Of the 320 studies, 83% (267) were cross-sectional, one look at one moment, and 84% were set in cities. Nilupa Gunaratna, a public health nutritionist and statistician at Purdue University, says: “Most research has been cross-sectional, with data collected at only one point in time.” The review has its own limits, too: English-language papers only, a search that stopped in May 2023 and was never rerun, one screener for the full-text stage (with 10% rechecked, at 88% agreement), and food-safety studies set aside on purpose.

Beyond the Binary, Start With Water

Of the 214 studies that made policy recommendations, spread across more than 40 countries, the biggest cluster (more than 65 studies) asked for basic infrastructure: water, electricity, designated vending zones, cold storage. A further 21% argued for moving beyond the formal/informal binary altogether, treating informality as a continuum, with options such as legal recognition or collaborative governance. The authors treat that as the recommendation underlying all the others; as a share of studies, it is smaller than the infrastructure ask. Gender-responsive measures such as women’s vendor associations, skills training and accessible microfinance turned up in 16%.

Whether vending is a stopgap or a career, the literature mostly cannot say, because it so rarely follows anyone for long enough to find out. Until researchers settle on who they are counting, or at least say it out loud, studies will keep being hard to compare.

FAQ