After leaving the EU, the UK government has proposed a new points-based immigration system, prioritising migrants with higher salaries or PhDs. The theory is that this would prevent "unskilled" workers coming to the UK, but still let highly qualified immigrants in. One assumption behind this policy is that immigration is only good for the "host" economy if the migrants are highly skilled. It is not obvious that this is true.

This article was inspired by the chapter "From the Mouth of the Shark", from Good Economics for Hard Times, by Esther Duflo and Abhijit Banerjee, which collects and summarises much of the evidence on this topic.

The Lump of Labour Fallacy

Migrants take jobs that might otherwise be held by native workers, and sometimes they are ready to accept a lower wage than natives would accept. There is a popular belief that this causes unemployment, or decreases wages for everyone, or both. This idea is based on the premise that there is a fixed number of jobs in an economy, so if the immigrants take jobs, then there are less for the natives, and if they accept lower wages, everyone gets paid less. This premise is false. (It is sometimes called the "Lump of Labour Fallacy".)

The Theory

When immigrants settle and get jobs, they become both workers and consumers. They buy food, go to the cinema and use transport. And they pay for these goods and services. This increases demand for food, cinema services, and transport. Over time, this creates jobs in those sectors. So although migrants do take some jobs, they also create others. It is worth noting that getting a job that already exists will occur in the short term, while increasing demand leading to new job creation is a longer term effect, so the "balancing-out" may not be immediate, depending on the flexibility of the labour market.

The Data

In fact, research has shown that a flow of migrants seems to increase, in equal measure, both the supply and demand for labour so average wages are unaffected. To see this, we cannot just look at the correlation between migration and wages. Over time, migration flows are positively associated with wage growth, but this does not tell us about the effect of migration on wages, since it might in fact be telling us about the effect of wages on migration. So, we need to look at examples where migration was caused by a clear factor other than wages in the "host" economy. Examples (taken from Duflo and Banerjee, 2019) include:

• the "Mariel boatlift", where 125,000 mostly unskilled Cubans arrived in Miami in 1980 as a result of Fidel Castro's unexpected authorisation for them to leave if they so wished (Card, 1990)
• the repatriation of French Algerians after Algerian independence in 1962 (Hunt, 1992)
• immigration from the Soviet Union to Israel after the lifting of the emigration restriction in 1990 (Friedberg, 2001)

Across these studies, researchers found little negative impact on the local population as a result of immigration (Duflo and Banerjee, 2019).

Low Skill Migrants

The UK's proposed immigration policy would favour highly skilled migrant workers. This policy assumes that while high skill migrants benefit the host economy, low skill migrants do not. But there is little evidence to suggest any particular harm coming from low-skilled workers: they do not seem to compete directly with low skilled native workers.

One reason for this is that the arrival of low skill migrants creates new roles for natives: the migrants specialise in more manual jobs, allowing natives to upgrade to more rewarding jobs. This is evidenced in Denmark, which saw a large influx of refugees between 1994 and 1998, who were randomly placed in cities with available housing and administrative capacity. Comparing those cities that received more migrants with those that received fewer allows us to see the effect of increased numbers of migrants on native-low-skilled workers. In this study, Danish low-skilled workers actually benefitted from migration, as firms reorganised production, allowing the Danish workers to upgrade to non-manual jobs that required technical expertise and communication skills (Foged and Peri, 2016).

Low-skilled migrants do not compete with low-skilled natives, as sometimes they seem to compete with machines. A reliable source of cheap and effective labour reduces firms' incentives to switch to machines. Mexican immigrant farm labourers, known as the braceros, were expelled California in December 1964 because it was believed that they were depressing wages for natives. Neither wages nor employment increased as a result. Farms that had employed braceros switched to machines, and stopped producing crops that could not be picked mechanically like asparagus and strawberries (Clemens, Lewis and Postel, 2018).

Low-skilled migrants often cannot replace natives due to some particularities of the labour market which give the incumbent native an advantage: efficiency wages, the morale effects of pay inequality, and the advantage of being known and trusted.

"Efficiency Wages"

The biggest incentive firms have to prevent workers from being unproductive is the threat of being fired. But for this to be an effective incentive, the pay has to be enough for being fired to really sting. In economics, we call this the efficiency wage. This means that firms will not pay newcomers much less than existing workers, since they don't want to risk paying newcomers too little (Shapiro and Stiglitz, 1984).

Morale effects of unequal pay

Firms try to keep differences between workers' pay low, due to the negative effects on morale of this kind of inequality. Worker un-happiness leads to lower productivity, so firms are incentivised to keep wages fairly equal. Alongside minimum wage regulations, this means that there is little scope for reductions in the pay of the lowest-skilled, just because of the free availability of migrant labour. (Breza, Kaur and Shamdasani, 2018).

Trust

Finally, low-skilled workers do not replace natives because firms prefer workers they know. Even in the UK, which has fairly relaxed labour laws compared to the rest of Europe, firing someone is not easy, and definitely not pleasant. As a result, firms are worried about hiring "bad workers". While in skilled jobs, firms often invest in tests and interviews to prevent hiring bad candidates, this is harder to do in unskilled jobs. Thus, the known incumbent worker has a huge advantage over the unknown migrant worker.

Empirical Evidence

Evidence supports the view that firms rarely replace incumbent natives with migrants. At one point, Czech workers were allowed to work just across the border in Germany, but went home to spend their money, so did not cause an increase in labour demand (unlike the other studies cited above). The border regions of Germany where the labour supply was swelled with Czech workers saw lower job gains for natives compared to towns where Czechs did not go. Incumbent German workers were not replaced with Czechs, but new hires were sometimes Czech and not German (Dustman, Schönberg and Stuhler, 2017).

In summary...

Despite widely held views, most evidence suggests that an influx of migrant labour rarely causes either job losses, or wage reductions for the native population. Low-skilled migrants in particular do not seem to pose a threat to native low skill workers.

If you enjoyed the read...

Then why not sign up to my mailing list to get notified when I post anything new?

Back to All Articles