Labour shortages of young workers willing to perform manual, low-paid jobs have become a persistent economic concern across advanced economies. Recent VoxEU columns show such shortages remain elevated post-pandemic and are concentrated in lower-paid, in-person, manual-intensive roles (Soldani et al. 2022, Causa and Soldani 2025). A related column, using evidence from a US visa lottery, finds that restricting low-skilled migrant hiring harms firms without generating offsetting gains for native workers (Lewis and Clemens 2022). Despite this evidence, public sentiment has grown more restrictive in many countries, even as firms continue to depend on temporary migrant labour in agriculture, manufacturing, hospitality, and personal care.

If migrants and natives are close substitutes, restricting immigration could raise native wages and employment, as some politicians claim. But if they perform complementary tasks within firms, abrupt restrictions may instead disrupt operations and reduce productivity. Earlier research on task specialisation found that immigrants tend to take manual-intensive roles, freeing natives to move into communication- and skill-intensive work (Peri and Sparber 2009, Peri 2010).

The pandemic provided a natural experiment

Our recent research (Lee et al. 2026) examines this question using a natural experiment from South Korea, one of the world’s fastest-ageing economies. Its Employment Permit System (EPS), introduced in 2004, is the main channel for importing low-skilled temporary workers from abroad (see Dong et al. 2024 and Kim et al. 2025 for earlier evidence on the impacts of the EPS). The programme allows small and medium-sized firms in labour-intensive sectors to employ migrant workers when they cannot fill positions domestically.

Before the pandemic, roughly 78% of EPS workers were in manufacturing, typically at small and medium-sized firms facing persistent hiring difficulties. These were not career-track positions; they were physically demanding, basic production jobs that Korean workers were increasingly unwilling to take at prevailing wages. As temporary migrants with shorter tenure and less firm-specific capital, EPS workers were naturally assigned to entry-level tasks, while domestic Korean workers occupied more advanced, experienced roles.

During the 2010s, the EPS quota was tightly managed and the guest worker count was stable. That changed abruptly during COVID-19. While Korea’s overall foreign population held steady, EPS workers – concentrated in labour-intensive manufacturing – fell sharply as strict border controls halted new arrivals and some existing workers returned home. The stock declined from about 276,000 in 2019 to 217,000 in 2021, a 22% drop. For firms that depended heavily on these workers, this constituted a sudden negative labour supply shock.

This abrupt interruption provides a rare opportunity to study firm responses to a sudden loss of low-skilled migrant labour. We combine two original surveys of EPS-using firms – conducted before and after the pandemic – with administrative data from Korea’s guest worker programme. The identification strategy exploits an institutional feature: the maximum number of EPS workers a firm could employ depended on firm size, with quotas changing at arbitrary size thresholds.

This rule generated quasi-random variation in migrant dependence across firms of nearly identical size. Some firms were permitted to employ EPS workers at a higher share of their workforce relative to similar firms, purely because of a one- or two-worker size difference. When border restrictions hit, firms with higher pre-pandemic EPS dependence faced a larger labour supply shock that, conditional on firm size, was uncorrelated with other pre-pandemic characteristics.

Firm survival and internal adjustments

Our first finding is that firms more reliant on guest workers before the pandemic were significantly more likely to exit in 2020–21. The overall closure rate in our manufacturing sample was about 4.4%, but higher EPS exposure was associated with substantially elevated exit probability. Effects were most pronounced among lower-wage, lower-productivity firms – those least able to quickly reorganize production after losing routine labour.

Among survivors, greater EPS reliance was associated with larger revenue losses, more production disruptions, and deteriorating business conditions. Many firms reported continued difficulty filling vacancies and unmet demand for migrant workers. Critically, these firms did not respond by hiring more Korean workers.

This directly challenges a common policy narrative. In many countries, restrictions on low-skilled immigration are justified with the claim that firms will simply hire local workers instead. Korea’s experience offers no support for this view. Firms hit harder by the migrant labour shock did not ramp up domestic hiring – and some closed. Migrant positions were not quickly filled by native workers at prevailing wages and conditions.

Instead, firms adjusted internally. Among survivors, higher pre-pandemic EPS reliance was associated with retaining incumbent Korean workers while reallocating them toward simpler tasks previously performed by migrants – at lower wages. Firms maintained production by shifting domestic workers down the internal job ladder.

Before the shock, Korean workers in these firms specialised in machine operation, supervision, or logistics coordination, while EPS workers handled basic, manual-intensive production. When the lower rung of that job ladder disappeared, firms did not rebuild around new domestic hires. They pushed existing workers into those roles and adjusted wages accordingly.

Lessons about low-skilled immigration restrictions

Our results expose the fallacy of ‘protecting’ native workers through immigration restrictions. A standard supply-and-demand model with homogeneous workers predicts that curbing immigrant labour should raise native wages and employment. But Korea’s experience points to a different mechanism. Low-skilled migrant workers filled production tasks that were essential and complementary to native labour. When they disappeared, production became less efficient, some firms closed, and native workers in surviving firms were downgraded in task assignments and pay.

The broader lesson is that sudden immigration restrictions can generate unintended costs for both firms and native workers, especially in ageing societies where the supply of young workers is shrinking. In the longer run, firms can adapt through technology adoption, organizational change, and shifts in specialisation. But in the short run, abrupt restrictions on low-skilled immigration impose real disruptions that ultimately harm the workers policymakers intend to protect.

Korea’s experience is directly relevant to other countries facing similar demographic and labour market pressures. Much of Europe and North America are grappling with shrinking labour forces and persistent shortages in the entry-level manual jobs that sustain manufacturing and service industries. In these settings, migrant workers underpin production systems that also employ domestic workers. Policymakers who argue that cutting low-skilled migration will automatically benefit domestic workers should consider the Korean evidence: restrictions on migrant labour can harm the very firms and workers they aim to protect.

References

Causa, O and E Soldani (2025), “When upskilling is good but not enough: Understanding labor shortages through a job-quality lens”, VoxEU.org, 27 May.

Dong, M, J Lee, and H-S Yang (2024), “How does low-skilled immigration affect native wages? Evidence from Employment Permit System in Korea,” Oxford Economic Papers 76(2): 433–450.

Kim, H, J Lee, and G Peri (2025), “The effect of low-skilled immigration on local productivity and amenities: Learning from the South Korean experience”, Journal of Urban Economics 146: 103738.

Lee, J, G Peri, and H-S Yang (2026), “The Effects of a Sudden Stop in Low-Skilled Immigration: Evidence from Korea’s Guest Worker Program”, NBER Working Paper 34927.

Lewis, E and M Clemens (2022), “Restricting the hiring of low-skill immigrants hurts firms and does not benefit native-born workers: Evidence from a US visa lottery”, VoxEU.org, 22 November.

Peri, G (2010), “Immigration and productive tasks: Can immigrant workers benefit native workers?”, VoxEU.org, 31 October.

Peri, G and C Sparber (2009), “Task specialization, immigration, and wages”, American Economic Journal: Applied Economics 1(3): 135–169.

Soldani, E, O Causa, N Luu, and M Abendschein (2022), “The post-COVID rise in labor shortages across OECD countries”, VoxEU.org, 28 November.