Revisiting immigration’s effect on US wages and employment

Anti-immigration policies often rest on the premise that immigrants worsen native-born employment and wages. This column uses data from 2000 to 2023 to estimate the impact of immigration to the US on national wages and employment, focusing on national rather than local impacts. The authors find that in these two decades, net immigration to the US was heavily college-intensive and raised the wages of less-educated natives. It also increased natives’ employment. Natives’ reallocation into higher-paying, communication-intensive occupations, complementary to those filled by immigrants, is one mechanism behind these gains.

A prominent justification of anti-immigration sentiments is the claim that immigrants compete with native workers in labour markets, crowding them out of employment and depressing their wages. This is a naïve and mostly incorrect view, which does not account for the expansion of the economy’s size, the efficiency gains, and the complementarity of immigrants with natives. In the US, immigration restriction proposals are frequently justified on the grounds of national interest, a determination that explicitly includes “economic considerations” such as “adverse effects on US workers”, as in the announcements terminating Temporary Protected Status for nationals of designated countries. 

Various natural experiments have called that premise into question. In South Korea, a sudden pandemic-era halt to guest-worker arrivals did not push firms toward domestic hiring. A meaningful share of firms closed, and surviving firms reassigned Korean employees to the more basic tasks migrants had handled, at lower pay (Lee et al. 2026). Two historical episodes point in the same direction: the 1882 Chinese Exclusion Act (Long et al. 2024) and the 1920s US quota acts (Feigenbaum et al. 2026) had long-run adverse impacts on natives.

The gap between this persistent belief that immigrants hurt US-born workers and what the data actually show revives the fundamental question about what immigration does to the wages and employment of native workers. The empirical consensus has long pointed to small-to-negligible effects of immigration on average native wages (National Academies of Sciences, Engineering, and Medicine 2017). However, much of that consensus rests on evidence that is more local than national in nature. The so-called local-area approach identifies immigration effects from sudden, geographically concentrated immigration shocks in specific US cities or states, and must then extrapolate from those local, often idiosyncratic shocks to infer what a more predictable, slower, and nationally dispersed inflow would do. 

By contrast, a different strand of the literature, the ‘factor-supply’ approach, models immigration as a change in the national supply of labour within education-experience skill groups and therefore internalises potential mobility responses, providing aggregate effects and accounting for labour market competition and interactions at the national level. This approach, developed in a series of influential contributions in the 2000s and 2010s (Borjas 2003, Ottaviano and Peri 2012, Manacorda et al. 2012), has not, however, been updated with more recent econometric developments, nor has it accounted for the immigration flows of the late 2010s and 2020s, which differed considerably from the previous ones.

Recent US inflows are smaller and skill-intensive

Our recent work (Caiumi and Peri 2026) examines the consequences of immigration inflows for US native workers, focusing on 2000–2023. As shown in Figure 1, over this period, the population of immigrants without a high school degree stopped growing and began to shrink, while college-educated immigrants continued to grow and even accelerated, becoming the largest group in the US foreign-born adult population by 2015. Net immigration to the US became smaller and considerably more concentrated among the college-educated than the late-20th century inflows analysed by the earlier studies (Ottaviano and Peri 2012, Borjas 2003), with potentially different implications for the labour market.

Figure 1 Immigrant population by education group, 1960-2023

For our analysis, we employ the national factor-supply approach, which we update in three ways. First, we better address potential concerns about the endogeneity of immigrant inflows – which respond to skill-specific productivity growth in the US – by adopting modern econometric methodology for estimating the key immigrant-native elasticity of substitution. Specifically, we use a new instrumental variable that modifies the ‘shift-share’ method by constructing a predictor of immigration flows based on the changing pattern of immigrants’ countries of origin over time (push factors) and the persistent skill characteristics of immigrants by origin. Both factors are much less correlated with US skill-specific productivity, and we show that they are not correlated with the past evolution of US wages and employment. 

Second, we extend this approach by explicitly modelling and estimating the impact of immigration on natives’ employment, which was previously assumed to be fixed in each skill group. Instead, we allow each US-born skill type to adjust its labour supply, responding to wages paid to workers with a similar skill level and to the presence of immigrants in its own labour market. This allows us to estimate whether natives are ‘crowded out’ of employment by immigration or, alternatively, attracted into work by the opportunities that immigration creates. 

Finally, we examine occupational upgrading and task specialisation of natives in response to immigration, as channels through which these effects operate at the national level. This builds on earlier research, which found that immigrants tend to take manual-intensive roles, encouraging natives to move into communication-intensive tasks (Peri and Sparber 2009).

By focusing on national rather than local impacts, our framework captures internal mobility responses by natives and accounts for richer skill dynamics than the local approach usually does. For instance, it can better internalise responses to higher demand from immigrants that are likely to spill over outside the local economy and imply reallocation of workers across commuting zones.

Immigration helps the wage growth of less-educated US-born workers

Our estimates show that the assumption of perfect substitutability (high competition) between immigrants and natives is not correct. We find a significant degree of productive complementarity between immigrant and native workers that outweighs the wage-competition effect for most skill groups of natives. Additionally, we estimate a positive employment response among natives, consistent with their being attracted into the labour force by the improved marginal productivity of their labour. We calculate that immigration raised the wages of less-educated native workers between 2000 and 2023 by an estimated 2.6% to 3.4% – a meaningful boost during a period in which this group’s real wage growth was close to zero. 

Relative to the estimated impact of the immigration flows during the 1990s and early 2000s calculated in Ottaviano and Peri (2012), the effects we find here are more favourable to less-educated Americans. For college-educated natives, whose skill group absorbed the bulk of the more recent immigrant inflow, the impact was small and not significant, more benign than a simple canonical model with perfect substitution would predict. This is because immigrant-native complementarity within the college-educated group is especially strong and large enough to offset the direct competition effect, leaving this group with negligible wage effects despite the scale of the inflow.

An important reason for the complementarity and positive employment effects likely lies in how natives respond to immigration, through efficient specialisation. We document that an increase in immigration prompts natives to specialise away from manual-intensive occupations and toward communication-intensive ones, resulting in higher wages and less competition with immigrants. The positive labour-supply effect on natives and their shift toward communication-intensive jobs are consistent with the idea that natives may prefer those occupations also for their amenity attributes at given wages.

Going forward

Immigration is not costless. Short-run adjustment costs, local labour market frictions, and non-labour-market margins such as housing and public services remain legitimate concerns. However, our results show that the claim that immigrants depress wages and reduce employment opportunities for natives, especially less-skilled ones, is not supported by the empirical evidence on the impact of immigration over the last two decades. The complementarity between immigrants and natives, and the large inflow of college-educated immigrants, combined to produce positive wage effects for less-educated US-born workers. These suggest that cutting immigration further would not help low-income Americans; rather, it may hurt the economy as a whole, as well as the job and earnings prospects of that group.

Source : VOXeu