When factories profit more, do workers benefit? A first-of-its-kind experiment in a coercive labor market has an answer

When factories run more efficiently, workers' conditions improve too. That premise has underpinned economics textbooks for nearly a century — the logic being that when companies earn more, some of those gains flow back to workers through higher wages or better working conditions.

A study testing that premise across 246 brick kilns in Bangladesh found that productivity did rise measurably, but workers received none of the gains.

The findings were published in the Proceedings of the National Academy of Sciences (PNAS), Vol. 123, No. 29, by a joint research team led by Grant Miller, a professor at Stanford University, and Aprajit Mahajan, a professor at the University of California, Berkeley.

The research team randomly divided 246 brick kilns across six districts in Bangladesh's Khulna region into three groups. Brick production was observed over the 2022–2023 season, and 1,442 workers were interviewed individually between March and May 2023.

Bangladesh has around 8,000 informal brick kilns — known as zigzag kilns — and is the world's fourth-largest brick producer.

The team shared a simple set of techniques with the kilns: changes to how coal is loaded into the furnace and how unfired bricks are stacked.

One group received only the new techniques. A second group received the techniques along with the additional message that rewarding workers would help the new methods take hold. Specific suggestions included wage increases, bonuses, provision of protective equipment, and school facilities for workers' children. The third group served as a control, with data collected but no intervention.

The new techniques cut fuel costs by 9.6 percent compared with the control group and raised the share of premium first-grade bricks by 8.1 percent — meaning lower costs and higher revenues.

The research team then examined whether any of those increased profits were passed on to workers as compensation.

To answer that question, they first needed to establish the baseline conditions workers were already living under. Between March and May 2023, the team met separately with six workers per kiln — 1,442 in total — in settings where neither the kiln owner nor the sardar was present.

The interviews found that in 49.6 percent of kilns, at least one of the six workers interviewed met the threshold for classification as a trafficking victim. Across all workers surveyed, the figure was about 20 percent.

The research team used a trafficking identification tool developed by the US Department of State, which screens for 39 indicators across seven domains — including recruitment practices, freedom of movement, debt, and violence — and classifies a person as a victim when a set threshold is crossed.

Seventy percent of workers were performing hazardous tasks without protective equipment. Seventy-one percent had no written employment contract, and 42.4 percent said their movement or communication was restricted. Five percent said they had no freedom at all.

Some 11.8 percent reported having wages or benefits withheld, and 6.4 percent said they had been deceived during recruitment. Another 3.7 percent said their personal space was under constant surveillance.

Child labor was even more widespread. Adult workers at more than 70 percent of kilns said they had witnessed children working on site. At 20 percent of kilns, workers reported seeing children under 14; most child workers were between 14 and 17.

The research team said that despite the productivity gains, trafficking indicators did not decline. There was no statistically significant difference across groups.

Wages did not rise. Workers at kilns that did not adopt the new techniques earned 3,811 taka (about $31) per week, while those at kilns that did adopt them earned roughly 30 taka less — a difference the researchers described as negligible.

A separate check of 19 working-condition indicators — covering accommodation, rest areas, access to clean water and toilets, protective equipment, bonuses, and travel allowances — yielded the same result: no meaningful improvement in any category.

The research team pointed to labor brokers known as sardars as the key structural reason.

Sardars recruit workers, supervise them on site, and serve as the link between kiln owners and laborers. They typically come from the same villages as the workers they recruit.

As the team found, the local labor market operates on a system of advance payments: sardars provide upfront cash on the condition that workers commit to a set period of labor. Workers enter the kilns to repay that debt.

Workers arrive at the kiln to find conditions different from what they were told, and their wages go toward repaying the debt. Those who try to leave are met with veiled threats of retaliation — in some cases directed at their families.

The research team concluded that raising productivity may not be sufficient to improve labor conditions in markets where coercive labor practices are entrenched.

However, the team acknowledged a limitation: the study captured conditions only during a single brick production season over a relatively short period.

DOI: 10.1073/pnas.2528388123

G. Miller, D. Biswas, A. Mahajan, K.S. Babiarz, N. Brooks, J. Brunner, S. Ashraf, J. Shane, A. Scarabosio, S. Maithel, S. Ahmed, M. Mahzab, M.R. Uddin, M. Rahman, & S.P. Luby, Productivity gains and work conditions in coercive labor markets: Experimental evidence from the Bangladesh brick sector, Proc. Natl. Acad. Sci. USA. 123 (29) e2528388123.