This is the 10th in this year’s series of posts by PhD students on the job market.
Employment usually follows a simple rule: you put in the work, and you get paid in return. That exchange – work for a paycheck at the end of the month – is what keeps many of us showing up every day. Now imagine doing the work but not getting paid. Not once, but again and again. Courts are slow, and legal fees are unaffordable. What happens then? Do workers protest, walk away, or simply wait, hoping their salaries will be paid eventually?
In many low-income countries, millions of workers face exactly this situation. Stories of unpaid salaries appear regularly in newspapers across Sub-Saharan Africa, from nurses in Ghana to dry-cleaners in Zimbabwe, suggesting that this is a region-wide challenge.
In Nigeria, the setting of our study, we document the scale of the problem through a large-scale survey of 1,279 current and former employees, randomly approached in markets, office complexes, and industrial areas. About 30 percent said they had experienced delayed or unpaid wages, with a median delay of one month. As shown in Figure 1, such incidents occur in firms of all sizes. Even among large firms with more than 50 employees, difficulties receiving full and timely pay remain common. Strikingly, fewer than one percent of affected workers reported attempting to recover their outstanding balances through the legal system, highlighting the weakness of formal enforcement mechanisms in this setting.
Figure 1: Probability of Salary Difficulty by Firm Size
Note: This figure reports the occurrence of wage withholding by firm size, estimated using our survey data.
These facts are not only worrying because they reveal a practice that could conceivably harm worker welfare – but also puzzling. Standard economic theory predicts retaliation: even in the absence of formal enforcement mechanisms, when firms cheat, workers should push back or quit, as documented in other settings. Indeed, one reason firms often lay off workers instead of cutting pay is the belief that lower wages will undermine morale. By that reasoning, employers should have strong incentives to pay on time. Yet, as we document, delayed and unpaid wages are prevalent. So, we asked two questions:
1. How do workers actually respond when pay is delayed?
2. Does widespread wage withholding affect people’s willingness to take up jobs at all?
A Field experiment on wage withholding
To answer these questions, we ran a field experiment in Lagos, Nigeria. We set up a company, hired 600 people across two employment rounds of three and two months to perform simple image-labeling tasks, and varied how certain they could be about getting paid on time. Some knew their wages were guaranteed, others were told there was a chance they might not be paid for a pay cycle; and a third group received no additional information about payment modalities beyond the specified salary.
We recruited workers for the jobs in two different ways. First, we posted physical job advertisements across Lagos to attract active jobseekers (n=638), allowing us to study the behavior of typical workers. Second, we approached people in busy public areas, telling them about the job to reach individuals who weren’t actively seeking wage work (n=1,079). This second group allowed us to study how salary certainty influences people’s willingness to enter wage employment in the first place.
Once hired, employees completed computer-based labeling tasks – think of identifying objects in photos – so we could precisely measure their productivity. Crucially, among those who had been told their pay might be delayed, we randomly delayed wages for some employees during the job itself. This design let us observe how workers actually behaved when faced with delayed pay. We then tracked who accepted the jobs, who showed up, and how their effort changed when wages were delayed.
When pay was delayed, effort increased
When salaries were delayed, workers’ performance improved. The probability of flawlessly completing labeling tasks increased by approximately 0.6 percentage points – a treatment effect of six percent, modest in size but statistically significant. This effect is comparable to offering a bonus of 5 Nigerian Naira (≈ 0.3 US cents) per correctly labelled image. Figure 2 illustrates the corresponding dynamic effects of salary delays on flawless task completion. We found no meaningful change in absenteeism or total hours worked.
Why would effort increase when pay is withheld? In settings where contract enforcement is weak, workers may work harder to keep their jobs and recover owed wages later. The fear of losing employment – or forfeiting back pay – pushes effort up, not down.
Figure 2: Dynamic Effects of Wage Withholding on Employee Effort
Note: This figure plots the dynamic effect of being owed salary payments on the prob- ability of flawless labelling task completion. The blue squares show dynamic event study estimates using the imputation estimator of Borusyak et al. (2024); standard errors are calculated using the conservative variance estimator proposed in Borusyak et al. (2024), initially clustered at the individual level and aggregated using the delta method. The orange circles show estimates from a separate regression testing for pre-trends as suggested by Borusyak et al. (2024) (these are not from the same estimation and coefficients are not relative to a single omitted time period as is often the case in figures like this one).
Reliable pay attracted additional workers
Telling candidates up front that reliable pay was guaranteed made a big difference in willingness to work for those not actively seeking wage work. Willingness to accept a job offer in this group jumped by 11 percentage points – roughly a 25 percent increase and highly significant. Among active jobseekers, acceptance was already above 95 percent, and the reliability signal changed little. In short, salary uncertainty deters some people from taking up employment – but once that uncertainty is reduced, they are willing to work.
The “new” workers looked a lot like the usual ones
Did guaranteed pay attract different workers – more skilled, less skilled, more productive? Not really. On observable traits and productivity, the new hires were indistinguishable from typical employees. They were, however, slightly more risk-averse – which makes sense: they were only willing to work once uncertainty about pay was reduced. In other words, reliable pay expanded the labor pool without changing its overall quality.
The Big asymmetry: Workers lose more than firms
From a firm’s perspective in a slack labor market like the one we study—where many people want more work—the productivity implications of withholding wages may be minimal. Our confidence intervals rule out that wage withholding reduces employee productivity but not that workers who could be newly recruited if wages were reliably paid are slightly more productive. Combining our estimates of worker productivity and workforce composition, even under this most unfavorable interpretation that takes the lower ends of the confidence intervals, suggests that firms’ output would fall by only about 0.2 percent.
For workers, in contrast, the cost is large. Using randomized wage offers, we estimated how much people would trade off in pay just to be sure of being paid on time. The willingness-to-pay for salary certainty was about 25,000 NGN (≈ USD 15) – more than an entire week’s typical wage in our setting. In other words, workers would give up a week’s pay every month just to avoid the uncertainty of not knowing when – or whether – their next payment would come.
What can be done?
Standard economic theory assumes wages are paid reliably. In many settings, that assumption seems to fail. When legal recourse is weak, firms face little risk when paying late or not at all. Workers hesitate to quit or to retaliate – doing so could mean losing wages already owed.
This creates a moral hazard on the firm’s side: withholding wages saves cash or borrowing costs in the short run while barely affecting productivity. The real burden falls on workers.
Tackling this problem will be difficult, yet our findings suggest a role for government intervention. Workers need credible ways to enforce their contracts and meaningful penalties for firms that fail to pay. In Nigeria, a legal initiative to establish harsher penalties for employers who fail to pay had begun, but progress has largely stalled. Our study underscores why such initiatives matter: without effective enforcement, wage withholding remains a low-cost strategy for firms – whether they have the cash and choose not to pay or save on borrowing when they do not – and a persistent threat to worker welfare.
Daniel Sonnenstuhl is a PhD student at the University of Chicago.
This research is funded by the Weiss Fund for Research in Development Economics, the Development Economics Center and the Becker Friedman Institute at the University of Chicago, as well as by the research initiative ‘Structural Transformation and Economic Growth’ (STEG), a programme funded by the Foreign, Commonwealth & Development Office (FCDO). The views expressed are not necessarily those of FCDO.
Join the Conversation