Fears over immigration are driving populist political sentiment in both the EU and the US. At first glance, the migration pressures on the regions appear similar. Each has absorbed recent waves of economic migrants from poor countries, and each also has large populations of skilled legal immigrants and asylees. Looking forward, however, immigration pressures on the EU and the US will diverge sharply. Europe is directly exposed to immigration from the Middle East and Africa, the primary drivers of future global population expansions, while the Americas are soon to experience uniformly slow population growth. Whether the US chooses to “build a wall” to immigration will become increasingly irrelevant to patterns of international migration, while the decisions the EU makes about policing its external borders will become more consequential.
Net immigration from Mexico to the US, after surging for three decades, has been negative for the past eight years (Villarreal 2014, Gonzalez-Barrera et al. 2015). After adjustment to the Great Recession, the US labour market is picking up, so what will happen to immigration to the US in coming years? Patterns of future macroeconomic shocks, civil conflict, or natural disasters are difficult to foresee with any clarity, but there is one critical determinant of migration pressures that is easy to predict: changes in labour supply. These will arise as a result of differential population growth across countries.
This feature is predictable for two reasons:
- Cohorts that will enter the labour force 15 to 20 years from now are already born, and so their numbers require no prediction.
- The rate of change of future population growth is generally smooth, especially over a one- or two-decade horizon, and is therefore among the most predictable of human behaviours. This smoothness allows us to foresee with considerable accuracy the size of cohorts that will be entering the labour market three to four decades from now.
To get a sense of how differential population growth will drive future migration, we can look to past episodes for guidance. In 1970, the US-Mexican border represented the line between a slow-growing, post-baby-boom population to the north, and the burgeoning labour supply of Latin America (Hanson and McIntosh 2010). This dissonance, accompanied by long-existing migration networks and relatively low migration costs, helped move nearly 15% of Mexico’s population to the US over the following 25 years. This was by far the largest modern international migration in absolute terms to an OECD country.
Figure 1 Population growth heat maps, 1970-80
Source: Hanson and McIntosh (2016).
All major destination countries contain a mixture of legal (destination-selected) migrants and undocumented (origin-selected) migrants. In recent decades, the composition of immigration in the US has tilted toward undocumented migration, while in Europe there have generally been smaller and more recent influxes of undocumented entrants. Our research has shown that over the past three decades, the presence of large-scale undocumented immigration made movement to the US strongly responsive to demographic shifts in origin countries (Hanson and McIntosh 2012). European immigration, in contrast, seemed less responsive to such demographic pressures (and driven more by EU integration). It is likely that this situation is about to reverse.
What do we foresee for global international migration? Using forecasts from the United Nations, we can map the growth of young cohorts who will enter the labour force by 2050. Like Australia in an earlier era, within a few decades nearly all of the New World will be a low-population-growth bastion surrounded oceans that greatly complicate immigration for low-income individuals. Europe, also contracting demographically, will be ringed by high-population-growth regions, including North Africa, Iraq, Iran, Pakistan, Central Asia, and particularly sub-Saharan Africa.
Figure 2 Population growth heat maps, 2040-50
Source: Hanson and McIntosh (2016).
Migration effects of population growth changes
In a recent paper, we build a predictive model that studied the interaction between population growth and pre-existing migrant networks (such as colonial and language linkages) to explain migration flows between 2000 and 2010. We then use these parameters to predict country-specific immigration over the coming decades (Hanson and McIntosh 2016). We predict migration destinations that are more exposed to population growth in sub-Saharan Africa – such as the UK, Spain, and Italy – will see large increases in the stock of first-generation immigrants, whereas the US will experience a gradual decline in its newly arrived immigrant population. Inflows to the US will increasingly be from countries that generate legal immigrants, such as India and China.
Figure 3 Predicted counts of first-generation immigrants aged 15-64 (millions)
This has stark policy implications for both origin and destination countries. The US emphasis on buttressing security along the US-Mexican border against surges in undocumented migration seems misplaced, and concerns about the moral hazard implications of granting amnesty to existing undocumented immigrants, seem less acute than in previous decades. The US has managed large-scale undocumented immigration in part because it has a relatively ungenerous welfare state that moderates the fiscal consequences of immigration, and in part because the children of immigrants have US citizenship and the chance to assimilate. The EU has more generous national welfare systems than the US, and therefore large-scale immigration may pose a more fundamental threat to the social contract between citizens and their states. The power vacuums in the buffer states that sit between the region and sub-Saharan Africa further complicate the EU context. If the EU wants to forestall future immigration, it may be in its interest to fund economic development programmes in Africa.