What determines how much a manufacturing company pays its workers? In the United States and Europe, that question has a fairly well-studied answer. In the developing world, much less so. Pamela Medina Quispe, an assistant professor of economics at the University of Toronto, set out to fill that gap using Peru as her laboratory. 

What makes Peru particularly instructive is the size of its informal self-employment sector. Between 50 and 70 per cent of the labour force works outside the formal economy — not as tech entrepreneurs, but as small-scale independent producers. "This is not like your Silicon Valley entrepreneur," Medina Quispe says. "These are people who have a sewing machine in their house and are producing clothes." This informal sector, it turns out, plays a far more important role in wage determination than policymakers have recognized.

The main point in her research, co-authored with Francesco Amodio of McGill and Monica Morlacco of the University of Southern California, is labour market power: the gap between the value a worker generates and what they are actually paid. When firms have market power over wages — a situation economists call monopsony — workers receive less than their full marginal product. In Peru, that gap is substantial. They found that workers in manufacturing receive roughly 70 cents for every dollar of value they generate for their employer.

The intuition is straightforward: when workers have a way to make a living on their own — sewing clothes at home, selling goods on the street, running a small informal business — employers cannot lowball them as easily. The threat of walking away is credible. So the more self-employment options workers have, the more pressure employers face to offer decent wages. And indeed, the research confirms this: in areas where several firms compete for the same workers, wages tend to be higher and the gap between productivity and pay tends to be smaller.

The more surprising finding concerns policy. Medina Quispe and her colleagues examined three initiatives that governments frequently use to create better formal jobs: road improvements that helped firms reach new markets, a youth job training program, and reforms that made it cheaper and easier to register a business. By most measures, all three worked. Firms grew, workers gained skills, new businesses opened, and GDP rose. "All three succeeded in their immediate goals," Medina Quispe says. "But they fell short of creating the wage employment and wage growth that policymakers hoped for."

The reason is a kind of unintended boomerang effect. When the policies made factory jobs more attractive, workers left their informal activities to take them — which sounds like progress. But as fewer workers remained in self-employment, employers lost the competitive pressure that had been keeping wages honest. With the outside option gone, firms could offer less, and much of the gain that the policies had created flowed to profits rather than paycheques. "The policies ended up partially undermining themselves," Medina Quispe explains. "They raised GDP and company performance, but the gains for workers in terms of wages and job quality were much smaller than expected, precisely because labour market power absorbed much of the benefit."

The flip side of this is equally important. In bad times, when factory jobs disappear, workers fall back on self-employment — and that buffer also limits how aggressively employers can cut wages, since workers always have somewhere else to go. Informal self-employment, in other words, is not just a sign of an underdeveloped economy. It is also a form of protection.

The lesson travels well beyond Peru. Medina Quispe says the findings can be applied to other economies where there is a strong informal self-employment sector, like Brazil and Colombia. And within countries like the U.S. and Canada, gig work — driving for Uber, delivering food, freelancing online — plays a structurally similar role. It is easy to enter, it offers an alternative to less desirable jobs, and its existence puts a quiet ceiling on how much employers in those sectors can suppress wages. As the gig economy grows, understanding its relationship to wage-setting power becomes increasingly important. 

Left unaddressed, the deeper problem risks producing what Medina Quispe calls “jobless industrialization” — economies where firms keep getting more productive and profitable, but workers see little of it in their wages or job prospects. For policymakers, the takeaway is uncomfortable but important: helping firms grow and helping workers thrive are not the same goal, and treating them as if they were may quietly ensure that neither happens as well as it should.