While much of the debate over minimum wages focuses on how they affect average pay levels or job losses, an equally important question is how they shape the way economic risk is shared within firms. This column uses data from the Italian metalworking sector to show how, in the case of a negative productivity shock, minimum wages provide insurance to low-skill workers against wage volatility while passing more of the wage adjustment burden to higher-paid workers. These dynamics highlight that minimum wage floors redistribute not only pay but also the risks associated with firm-level fluctuations.
Debates over minimum wages have intensified in recent years (Roth et al. 2022) as policymakers seek tools to reduce inequality without harming employment. Much of this debate focuses on how minimum wages affect average pay levels or job losses. Yet an equally important (but less visible) question is how minimum wage institutions shape the way economic risk is shared within firms.
A large literature shows that firm heterogeneity and the pass-through of firm-specific shocks lead to large variations in workers’ labour earnings (Abowd et al. 1999, Sorkin 2018, Song et al. 2019, Kline et al. 2019). What remains largely unexplored is how this pass-through is shaped by minimum wage institutions.
In our recent paper (Adamopoulou et al. forthcoming), we study this leveraging rich employer–employee data from the Italian metalworking sector, covering more than 600,000 worker–firm–year observations. We merge these data with hand-collected, occupation-specific minimum wage floors from sectoral collective agreements, which we assign to each worker. A key advantage of our setting is that we observe the entire workforce of each firm, which allows us to precisely measure how close each worker’s wage is to the relevant occupational minimum and to construct firm-level indicators of minimum wage exposure. This enables us to study how the pass-through of firm-level shocks depends jointly on worker-specific and firm-wide minimum wage incidence. In addition, we link these data to firms’ balance-sheet information, which allows us to construct firm-level productivity measures and to identify negative total factor productivity (TFP) shocks.
Minimum wage floors and within-firm adjustment
The Italian metalworking collective agreements set detailed occupation-specific minimum wage floors, but there is substantial variation in how binding these wage floors are across workers and firms.
Our analysis shows that for workers whose wages are close to the minima, negative firm productivity shocks impact the separation rates rather than the wages. This implies that low-paid workers who remain employed are insulated from wage cuts but face a higher risk of job loss when firm productivity falls. In contrast, wages of workers who are well above the floors adjust significantly to the same shock. Our key finding is that the magnitude of this adjustment increases with the firm’s share of minimum-wage workers. Consider two firms, A and B, where Firm A has a higher concentration of workers near the wage floor. When both firms experience an identical negative TFP shock, high-wage workers in both companies see their pay decrease. However, those in Firm A are likely to experience a more significant reduction than their counterparts in Firm B. Overall, our results indicate that high-paid employees absorb a disproportionate share of within-firm wage cyclicality, and this is particularly prevalent in firms with a high fraction of workers near the legal minima.
A natural question is whether higher job separation rates outweigh wage adjustments, so that low-wage workers ultimately bear the brunt of firm-level shocks. We show that this is not the case. To assess overall earnings losses, we examine changes in workers’ labour earnings that combine both wage changes and employment outcomes, treating workers who are laid off and do not find a new job as experiencing a complete earnings loss. Using this broader measure, we find that the same patterns persist. First, high-wage workers are, on average, more exposed to firm-level shocks than low-paid workers. Second, the labour earnings of high-wage workers are more sensitive to firm-level shocks if they have a larger share of co-workers close to the minimum wage floors.
Why the pass-through is asymmetric: Complementarities in production
This pattern arises naturally when workers of different skill levels are complements in production (Krusell et al. 2000, Caselli and Coleman 2006). A negative productivity shock exacerbates the severity of the minimum-wage constraint for low-skill employees, whose employment becomes rationed. Crucially, their reduced presence in the production of the firm lowers the marginal productivity of high-skill workers, making the wages of the latter more sensitive to the shock. Minimum wages thus provide insurance to low-skill workers against wage volatility while passing more of the wage adjustment burden to higher-paid workers.
We confirm that this insurance and burden-shifting mechanism is not an artefact of any single way of measuring firm conditions. Our baseline shock is a firm-level TFP innovation recovered from balance-sheet data by estimating a value-added production function and applying the widely used control-function approach of De Loecker and Warzynski (2012), which uses intermediate inputs to proxy for unobserved productivity. We then focus on negative realisations of these innovations. Reassuringly, the same qualitative pattern emerges when we replace TFP shocks with other sources of firm-level exogenous shifts, namely, labour productivity shocks (sales per worker) and export demand shocks constructed from customs transactions. Across all cases, wage floors limit wage cuts at the bottom and the adjustment is reallocated toward workers further from the minima, especially in firms where a large share of employees is paid close to the negotiated floors.
Beyond these robustness checks, we provide direct evidence that skill complementarities are quantitatively important in our data. Using employer–employee information on earnings growth and working-time adjustments, we compute a non-parametric measure of complementarities based on the ratio of idiosyncratic earnings volatility to idiosyncratic working-time volatility. Under perfect substitutability, worker-specific shocks that raise earnings should be mirrored by proportional changes in working time, so the ratio equals one. With stronger complementarities and team production constraints, working time cannot adjust one worker at a time, so earnings move more than hours and the ratio rises. In our data, the ratio averages well above one – around 3.4 overall (about 3.6 for blue collars and 3.1 for white collars) – indicating strong complementarities in production. Importantly, the asymmetric pass-through is stronger precisely in firms where this measure is higher, reinforcing the interpretation that minimum wages interact with complementarities to shift firm-level risk toward higher-paid workers.
We build a quantitative model to study the welfare implications of the interactions between the firm-level shocks, minimum wages, and the complementarities across skill groups in production. The model has the minimal structure to provide a realistic yet transparent laboratory to study these interactions. There are firm-level productivity shocks, incomplete markets, and partial rigidity in worker mobility across firms. Accordingly, there is pass-through of the firm-level shocks on workers that are not perfectly insurable. Production features complementarities between units of different skill groups. Negative firm productivity shocks lead to rationing of workers at the minimum wage, reducing the marginal productivity – and hence the wages – of high-wage earners beyond the direct implication of lower firm productivity. This model, calibrated to the actual wage distribution, turnover patterns, and skill-substitution elasticity, reproduces the empirical patterns and uncovers the underlying mechanisms.
We use the model to identify how welfare gains from minimum wages are distributed across workers with different skills, by comparing an economy with wage floors calibrated to the Italian case to a counterfactual economy without wage floors. Results show that blue collars suffer the most from the removal of minimum wages (see Figure 1). Most importantly, there is substantial heterogeneity in welfare gains and losses over the skill distribution. The consumption equivalent welfare losses are above -0.2% for low-skill blue collars, while high-skill blue collars are slightly better off. The augmented pass-through of the firm-level shocks on high-wage workers, arising from the interaction of the minimum wages and production complementarities, make the minimum wages less beneficial for this group.
Figure 1 Welfare gains/losses from removing the minimum wage
Source: Adamopoulou et al. (forthcoming)
Relevance of the main findings and the policy implications
The findings speak directly to broader debates on wage rigidity. While staggered collective contracts are known to slow the downward wage adjustment and worsen unemployment in downturns (Adamopoulou et al. 2024), our research highlights a different dimension: the role of skill complementarity. We show that, after a negative productivity shock, the interaction of minimum wage floors and skill complementarities shields low-wage workers from wage reductions while forcing higher-paid employees to absorb a larger share of the adjustment. These dynamics highlight that minimum wage floors redistribute not only pay but also the risks associated with firm-level fluctuations.
The interaction between minimum wage floors and skill complementarities is central to understanding who absorbs firm-level shocks, and hence to insurance within the firm. This interaction implies that firms provide more wage insurance to low-paid workers and less to high-paid workers. Minimum wages therefore function as an internal risk-sharing device, shaping how firms allocate the volatility generated by negative productivity shocks.
While the empirical base of our study is the metalworking sector, the mechanism we put forward applies to any sector featuring complementarities between workers of different skills and productivity. Importantly, the differential impact of the minimum wages across workers of different levels of wages would be more pronounced in sectors in which such complementarities are stronger. Sectors with stronger complementarities and interactions between different skill-groups of workers would feature a more disproportionate passthrough on high-wage workers.
Source : VOXeu
