Lecture 4 & 5 Labour Immigration Claude Notes

Immigration is one of the most politically charged topics in contemporary British public life, yet popular debate frequently diverges from the economic evidence. Understanding the scale, composition, and trajectory of migration flows is a prerequisite for any rigorous analysis of labour market or fiscal consequences.

This figure plots long-term international migration to the UK from 1970 to 2014, decomposing the total into gross immigration, gross emigration, and the resulting net migration figure. The chart illustrates a structural transformation in the UK's migration balance: throughout the 1960s and 1970s, net migration was consistently negative, meaning more individuals were leaving Britain than arriving. This reversed modestly in the 1980s and remained at a low positive level into the early 1990s. From 1994 onwards, net migration has been positive in every single year, driven initially by non-EU flows and subsequently amplified by intra-EU mobility following the 2004 and 2007 enlargements. By 2016, net migration stood at approximately 273,000 persons annually, a figure that features prominently in domestic political debate.

This ONS chart disaggregates net migration by nationality group (non-EU, EU, and British nationals) from mid-2012 to mid-2023. Several structural breaks are clearly visible. Prior to the 2016 EU referendum, EU net migration was the dominant driver of the positive total, running at roughly 200,000 per annum. Following the referendum and especially after the end of the EU transition period in late 2021, EU net migration collapsed sharply and turned mildly negative as EU nationals departed or reduced inflows in response to the new immigration system. By contrast, non-EU net migration surged dramatically after 2021, reaching provisionally estimated figures approaching 800,000 by year ending June 2023. This compositional shift is crucial: it shows that Brexit did not reduce overall immigration but rather redirected its source, replacing free-movement EU workers with visa-based non-EU migrants (predominantly from South and East Asia, and from humanitarian routes).

This figure shows total immigration, emigration, and net migration between mid-2018 and mid-2023 with confidence intervals. It makes unmistakably clear that the explosion in net migration post-2021 was driven almost entirely by a surge in immigration rather than a reduction in emigration. Gross immigration exceeded 1.2 million in provisional estimates for year ending June 2023. The sharp divergence between the immigration and emigration series from late 2021 onwards underlines that the UK became a substantially more attractive destination following the pandemic, with students, workers on the Health and Care visa, and humanitarian entrants all contributing.

This ONS stacked bar chart decomposes non-EU immigration and emigration by reason from 2019 to 2023. The composition of immigration flows shifted markedly over this period. Study and work visas accounted for the bulk of the post-2021 rise, with humanitarian routes (including the Ukraine and Afghanistan programmes) also making a significant contribution. Work-related immigration was the primary driver of the non-EU surge by year ending June 2023. Crucially, emigration remained relatively flat throughout, reinforcing the interpretation that higher net migration was driven by increased arrivals rather than reduced departures. For the purposes of labour market analysis, the dominance of work and study visas in the immigration column confirms that the majority of recent non-EU entrants are economically active or likely to become so.

A foundational question in immigration economics concerns who chooses to migrate. If immigrants are positively selected from their source population (i.e., they possess above-average ability, motivation, or education), the labour market and fiscal consequences for the destination country are likely to be more favourable than if selection is negative. The debate has a long history: Benjamin Franklin's 1753 complaint about German immigrants characterised them as disproportionately drawn from the least able. Subsequent economic scholarship has largely inverted this assessment.

Chiswick (1978) provided influential early evidence that immigrants to the United States were more able and highly motivated than comparable natives. Carliner (1980) similarly found that immigrants chose to work longer and harder than non-migrants. These findings are consistent with a simple theoretical prior: migration is a costly investment, requiring not only monetary resources but also the psychological capacity to uproot oneself and take risks, so only individuals with the drive and ability to profit from migration will undertake it.

Borjas (1987) provided the canonical theoretical treatment of immigrant self-selection. The framework models the migration decision as a comparison of returns to skills in the source and destination countries. The two panels of this diagram plot log wages against skills in both the source country and the US, with the crossing point determining the selection threshold (positive) or (negative).

In the left-hand panel (positive selection), the US wage-skills schedule has a steeper slope than that of the source country, implying that the returns to skill are higher in the US. Only workers above the threshold will find it worthwhile to migrate, so the migrant pool is disproportionately drawn from the upper end of the skill distribution. This arises when the source country has relatively low income inequality (a compressed wage distribution), since high-skilled workers gain proportionally more by moving to the more dispersed US wage structure.

In the right-hand panel (negative selection), the source country offers a steeper wage-skills gradient than the US, implying that the source country already rewards skill more generously. In this case, only workers below the threshold gain from migrating, generating a negatively selected pool. This is more likely when income inequality in the source country is high, since the returns to skill are already being captured domestically.

The empirical record is mixed, but leans towards positive selection on educational attainment. Gould and Moav (2016) found that Israeli immigrants to the United States were positively selected. Feliciano (2005) and Grogger and Hanson (2011) document that, on average, immigrants exhibit positive selection on educational attainment from almost every sending country worldwide. Chiquiar and Hanson (2005) provide a nuanced finding for Mexico: Mexican immigrants to the US are drawn from the middle of the skill distribution rather than the bottom, suggesting neither strongly positive nor negative selection, which they interpret as reflecting the fixed costs of migration.

For the UK specifically, the relevant finding is that immigrants are, on average, better educated than native workers, which is consistent with positive selection.

The central analytical question is whether immigration raises or lowers the wages and employment of native workers. The answer depends critically on two structural assumptions: (i) whether immigrants and natives are substitutes or complements in production; and (ii) whether capital is fixed (short run) or mobile (long run).

This diagram depicts the short-run labour market equilibrium for the UK under the assumption that immigrants and natives are perfect substitutes. The horizontal axis measures total employment ( for natives, for natives plus immigrants), and the vertical axis measures the wage . Initially, the market clears at wage with native employment .

When immigrants arrive, the total labour supply curve shifts outward from the domestic supply to the combined (domestic + immigrant) supply curve. In the short run, capital is fixed, so the demand curve does not move. The new equilibrium sees the wage fall from to , total employment rise from to , and native employment actually fall from to as some native workers are priced out of the market at the lower wage. Capital owners benefit because the return to capital rises as labour becomes cheaper.

In the long run, capital is mobile: it can expand or contract in response to profitability signals. The short-run wage depression increases the return to capital, inducing firms to invest and expand their capital stock. As capital accumulates, the labour demand curve shifts rightward, raising wages back towards and restoring native employment to . Total employment expands further to , incorporating both the original immigrant inflow and the additional workers hired as a result of expanded capital.

The long-run implication is striking: with capital mobility, the wage effect of immigration is attenuated towards zero. The economy effectively scales up, absorbing the immigrant supply shock without a permanent deterioration in native wages. This is the mechanism underlying the finding that immigration may have little long-run effect on native wages even when short-run effects are negative.

The substitute model assumes immigrants and natives compete in the same labour market. However, if immigrants and natives are complements, the analysis is fundamentally different. In this case, immigrants make native workers more productive, shifting the demand curve for native labour outward rather than the supply curve. The diagram shows that native wages rise from to and native employment rises from to .

Two canonical channels generate complementarity. First, high-skilled immigrants (such as scientists and researchers) may enable native academics and researchers to further specialise in areas of comparative advantage, expanding the productivity frontier. Second, low-skilled immigrants who fill manual, routine, or physically demanding roles may free native workers to move into more complex, communication-intensive, and higher-paid occupations. In both cases, the immigrants occupy a distinct niche in the production process rather than directly competing with natives.

Identifying the causal effect of immigration on native wages is methodologically complex. The naive approach of correlating wages across cities with different immigration intensities suffers from a severe endogeneity problem: immigrants are not randomly distributed across labour markets. They tend to cluster in prosperous, growing cities where wages and employment conditions are already favourable. A simple cross-city regression would therefore generate a spurious positive correlation between immigration and wages, biasing estimates away from any negative effect. The literature has responded with two principal strategies: instrumental variables and natural experiments.

A natural experiment is an event that generates quasi-random variation in the treatment variable (here, immigrant supply), allowing the researcher to compare outcomes between an affected group and a control group that was not exposed. Unlike a randomised controlled trial, individuals in a natural experiment have not been randomly assigned; they may have self-selected. The key requirement is that the variation in the treatment is plausibly exogenous to the outcome variable of interest.

The canonical examples in the immigration literature are the Mariel boatlift (Card, 1991), the mass arrival of Russian Jews in Israel (Friedberg, 2001), and the repatriation of French settlers from Algeria (Hunt, 1992). Each provides an abrupt, large-scale labour supply shock that is plausibly independent of pre-existing labour market conditions.

In 1980, Fidel Castro temporarily opened the port of Mariel in Cuba, permitting emigration to the United States. Approximately 125,000 Cuban nationals arrived in Miami between May and September 1980, increasing the city's labour force by around 7%. Critically, the Marielitos were substantially less educated than the average US resident, with around 57% lacking a high school diploma. This makes them a strong test case for the substitutability hypothesis, since they were directly competing with the least-skilled segment of the Miami labour market.

Card (1991) found no discernible effect on the unemployment rate or wages of unskilled non-Cuban workers in Miami. Comparing Miami to a control group of cities (Atlanta, Houston, Los Angeles, and Tampa-St. Petersburg) that did not receive the shock, the unemployment rate for unskilled workers actually fell relative to the comparison cities after the boatlift.

Ottaviano and Peri (2012) shift the unit of analysis from local to national labour markets, arguing that local market studies underestimate the wage impact of immigration because native mobility disperses the shock across all markets. Their results, reproduced in this table for 1990-2006, reveal a striking asymmetry: native-born workers experienced small positive wage changes across all education groups (aggregate +0.6%), whilst foreign-born workers experienced large negative wage effects in every educational category (aggregate -6.7%). The largest losses fell on previous immigrants with a high school degree (-12.6%).

This finding has a clear economic interpretation. It suggests that within education categories, immigrants and native workers are imperfect substitutes, presumably because language fluency, networks, and familiarity with domestic institutions make native workers more productive in many roles. New immigrants therefore compete primarily with existing immigrants rather than with natives, explaining why natives fare reasonably well at the national level.

Friedberg (2001) exploits the mass emigration of Russian Jews to Israel following the dissolution of the Soviet Union as a natural experiment. Between 1990 and 1991 alone, 610,000 Russian Jews arrived in Israel, representing approximately 7% of the Israeli population; over the 1990s as a whole, the Israeli population grew by 20% due to this single migration episode.

Russian immigrants were, on average, more highly educated than the native Israeli population. Yet real wages fell by approximately 5% in the short run. Friedberg explains this apparent paradox through occupation downgrading: Russian immigrants entered occupations well below those commensurate with their qualifications, effectively flooding the lower occupational tiers of the labour market. Crucially, the wage effect was attenuated by substantial capital accumulation in Israel during the same period, consistent with the long-run theoretical prediction that capital investment absorbs the wage impact of immigration over time.

Hunt (1992) studies the repatriation of approximately 900,000 French-born settlers (pieds-noirs) from Algeria in 1962 following the end of the Algerian War of Independence. This constituted roughly 1.6% of the French labour force and was largely unanticipated, providing a plausible natural experiment. Unlike the Marielitos, the repatriates were educationally similar to French natives. Hunt finds only very small negative effects on wages, consistent with the view that skill complementarity and capital adjustment rapidly absorbed the supply shock.

The UK evidence, synthesised across Dustmann et al. (2005, 2008), reveals a nuanced distributional pattern. Immigrants to the UK are, on average, better educated than native workers. The wage effect of immigration is not uniform across the wage distribution: there is a small negative effect at the bottom of the wage distribution and a small positive effect at the top, with no average wage effect overall.

Manacorda, Manning, and Wadsworth (2010) extend this analysis and find, consistently with Ottaviano and Peri (2012) for the US, that immigration primarily depresses the earnings of previous immigrants rather than of native-born workers. They document a small rise in the returns to education for natives and a small deterioration for previous immigrants, reinforcing the interpretation that natives and immigrants are imperfect substitutes in the UK labour market.

Beyond labour market effects, a central policy question concerns the net fiscal contribution of immigrants: do they receive more in public services and transfers than they pay in taxes, or vice versa? This question is analytically distinct from the wage effect question and requires careful attention to the definition of the immigrant population studied and the allocation of fixed public goods such as defence and infrastructure.

Dustmann and Frattini (2014) provide the most comprehensive UK analysis, using Labour Force Survey data linked to tax and benefit records. They distinguish between all immigrants who have been present in the UK since 1995 and "recent" immigrants who arrived from 2000 onwards, and further disaggregate by European Economic Area (EEA) and non-EEA origin.

The aggregate fiscal picture for all immigrants resident since 1995 is mixed. EEA immigrants made a positive net contribution to the UK public finances of £8.8 billion between 2001 and 2011, paying approximately 4% more in taxes than they received in transfers and public services. Non-EEA immigrants made a negative contribution of £104 billion over the same period. Natives made the largest negative contribution of £604.5 billion, reflecting the scale of the native population and their heavier reliance on public services across all stages of the life cycle.

This chart plots the ratio of revenues to expenditures for natives, EEA immigrants, and non-EEA immigrants across fiscal years from 1995 to 2011. A ratio above 1.0 indicates net fiscal contribution (taxes exceed benefits received); below 1.0 indicates net fiscal cost. The EEA line (red) fluctuates above and around 1.0 for much of the period, with a notable peak in the early 2000s when large waves of EU-10 workers arrived. The non-EEA line (green) is predominantly below 1.0, reflecting the older age profile and family-based migration of many non-EEA entrants, whilst the native line (blue) is consistently below 1.0, particularly in the post-2008 recessionary period.

When Dustmann and Frattini (2014) restrict the analysis to recent immigrants who arrived from 2000 onwards, the picture becomes markedly more positive. EEA recent immigrants made a positive contribution of £22.1 billion between 2001 and 2011, paying 34% more in taxes than they received. Non-EEA recent immigrants also made a positive contribution of £2.9 billion, paying 2% more than they received. By contrast, natives made a negative contribution of £624.1 billion.

These findings highlight the importance of the time horizon and cohort definition. Recent immigrants, who are more likely to be of working age and to have arrived specifically for employment or study purposes, exhibit a substantially more favourable fiscal profile than the stock of all immigrants including those who arrived decades earlier.

One of the most striking findings from Dustmann and Frattini (2014) concerns the human capital endowment that immigrants bring. Because immigrants' education was financed abroad, the UK labour market receives the productive returns to this investment at zero educational cost to the UK public sector. Dustmann and Frattini estimate that immigrants arriving since the early 2000s were endowed with human capital that would have cost approximately £49 billion had it been produced through the UK education system. Additionally, recent immigrants are approximately 21% less likely to receive state benefits or tax credits than comparable natives of similar age, gender, and education.

Immigration may restrain inflation through its effects on wage growth. For the United States, evidence suggests that a 10% increase in the supply of low-skilled immigrants reduces wages of other low-skilled immigrants by 8.0% and wages of low-skilled native workers by 0.6%. Since low-skilled workers are disproportionately employed in non-traded services (cleaning, catering, construction, personal care), wage compression in these sectors lowers the price of non-traded goods and services, reducing the overall price level. This mechanism provides a potential welfare gain for consumers, particularly those who purchase labour-intensive services.

Immigration also exerts demand-side pressure on local housing markets. Evidence from the United States, drawing on the Mariel boatlift natural experiment, documents an 8% increase in the prices of low-quality apartments in Miami as a result of the 1980 Cuban inflow, with no significant effect on high-quality apartment prices. This distributional asymmetry makes sense: low-income immigrants compete for low-quality housing, bidding up rents at the bottom of the housing market, whilst their presence does not materially affect demand for high-quality housing. From the perspective of low-income native renters, this constitutes a welfare cost that partially offsets labour market gains.

A more speculative but important strand of the literature concerns the effects of immigration on social capital, institutions, and political behaviour. The political economy of immigration has become more salient in recent years. Alesina and Tabellini (2024) examine whether the political backlash against immigration in high-income countries is driven primarily by economic fears (labour market competition, fiscal costs) or by cultural concerns (preferences over social norms, language, and identity). Their findings suggest that cultural factors play a significant and potentially dominant role in shaping political responses to immigration, even when economic effects are modest or positive. This has important implications for policy design: even technically sound immigration programmes may face political resistance rooted in non-economic motivations.

Brell, Dustmann, and Preston (2020) draw a fundamental analytical distinction between economic migrants and refugee migrants. Economic migrants self-select in response to wage differentials and employment opportunities, as the Borjas (1987) framework predicts. Refugees, by contrast, are forced or unexpected migrants who flee persecution, conflict, or humanitarian crisis. Their migration decision is driven not by expected economic returns but by survival and safety considerations.

This difference in the migration motive has profound implications for labour market performance in the destination country:

This figure from Brell, Dustmann, and Preston (2020) tracks employment rates and wage levels for refugees and other immigrants in nine high-income countries over the decade following arrival. The blue lines (refugees) lie consistently below the red lines (other immigrants) in both employment rate and wage level at arrival, confirming the lower initial labour market performance of refugees. However, both series exhibit an upward-sloping trajectory, and the gap between refugees and other immigrants narrows over time in most countries, suggesting a process of slow but real labour market assimilation.

The lower starting point for refugees reflects both human capital deficits (language, credentials, professional networks) and the structural barriers they face (restrictions on working during the asylum determination process, poor mental health, concentration in low-demand geographic areas due to dispersal policies).

The theoretical analysis and empirical evidence suggest several policy levers for improving refugee labour market integration:

The lecture integrates theory, methodology, and international empirical evidence to reach the following principal conclusions:

Alesina, A. and Tabellini, M. (2024) 'The political effects of immigration: culture or economics?', Journal of Economic Literature, 62(1), pp. 5-46.

Borjas, G.J. (1987) 'Self-selection and the earnings of immigrants', American Economic Review, 77(4), pp. 531-553.

Borjas, G.J. (1994) 'The economics of immigration', Journal of Economic Literature, 32(4), pp. 1667-1717.

Brell, C., Dustmann, C. and Preston, I. (2020) 'The labor market integration of refugee migrants in high-income countries', Journal of Economic Perspectives, 34(1), pp. 94-121.

Card, D. (1991) 'The impact of the Mariel boatlift on the Miami labor market', Industrial and Labor Relations Review, 43(2), pp. 245-257.

Carliner, G. (1980) 'Wages, earnings and hours of first, second and third generation American males', Economic Inquiry, 18(1), pp. 87-102.

Chiquiar, D. and Hanson, G.H. (2005) 'International migration, self-selection, and the distribution of wages: evidence from Mexico and the United States', Journal of Political Economy, 113(2), pp. 239-281.

Chiswick, B.R. (1978) 'The effect of Americanization on the earnings of foreign-born men', Journal of Political Economy, 86(5), pp. 897-921.

Dustmann, C. and Frattini, T. (2014) 'The fiscal effects of immigration to the UK', The Economic Journal, 124(580), pp. F593-F643.

Dustmann, C., Frattini, T. and Preston, I. (2008) 'The effect of immigration along the distribution of wages', IZA Discussion Paper, No. 3861.

Dustmann, C., Hatton, T. and Preston, I. (2005) 'The labour market effects of immigration', The Economic Journal, 115(507), pp. F297-F299.

Feliciano, C. (2005) 'Does selective migration matter? Explaining ethnic disparities in educational attainment among immigrants' children', International Migration Review, 39(4), pp. 841-871.

Friedberg, R.M. (2001) 'The impact of mass migration on the Israeli labor market', Quarterly Journal of Economics, 116(4), pp. 1373-1408.

Gould, E.D. and Moav, O. (2016) 'Does high inequality attract high skilled immigrants?', Economic Journal, 126(593), pp. 1055-1091.

Grogger, J. and Hanson, G.H. (2011) 'Income maximization and the selection and sorting of international migrants', Journal of Development Economics, 95(1), pp. 42-57.

Hunt, J. (1992) 'The impact of the 1962 repatriates from Algeria on the French labor market', Industrial and Labor Relations Review, 45(3), pp. 556-572.

Manacorda, M., Manning, A. and Wadsworth, J. (2010) 'The impact of immigration on the structure of wages: theory and evidence from Britain', Journal of the European Economic Association, 10(1), pp. 120-151.

Ottaviano, G.I.P. and Peri, G. (2012) 'Rethinking the effect of immigration on wages', Journal of the European Economic Association, 10(1), pp. 152-197.

Office for National Statistics (2024) Migration Statistics Quarterly Report. Available at: https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/internationalmigration/bulletins/migrationstatisticsquarterlyreport/latest (Accessed: April 2026).

Roy, A.D. (1951) 'Some thoughts on the distribution of earnings', Oxford Economic Papers, 3(2), pp. 135-146.