The economics of immigration amnesties

Alessandra Casarico, Giovanni Facchini, Tommaso Frattini 28 June 2018



Illegal immigration is widespread in OECD countries. The recent, unprecedented surge in the number of asylum seekers in Europe (Figure 1) and the high likelihood that their claims will be rejected will increase the stock of undocumented foreign nationals in many destination countries (Casarico et al. 2016, Gordon et al. 2009, Hatton 2011, Triandafyllidou 2009). In an environment in which repatriations are difficult to implement, policymakers have few options available to handle a growing problem.  

Figure 1 Applications for asylum in the EU 27 countries, in thousands

Note: Total number of applications for conventional refugee status in the EU 27 countries.
Source: Authors’ calculations based on UNHCR (2018).

Illegal immigration and amnesties

To understand the dimensions of the challenge, Table 1 provides some recent estimates of illegal immigration for a group of important destination countries – both as a share of the total population (column 1) and as a share of the foreign born (column 2).In 2016, irregular migrants made up 3.4% of the total population in the US, or about a quarter of the total foreign population. In Europe, the figures are significantly lower, but exhibit substantial variation across countries. The phenomenon is virtually absent in Denmark or Sweden, very small in countries like Austria and Germany, but sizeable in Greece, Belgium, Portugal, and the UK. The legal status of migrants clearly reflects the policy stance of the destination country, both in terms of the ex-ante controls introduced to discipline the flows and the ex-post measures taken to grant legal status to existing undocumented foreigners.

Used as a tool to give legal status to a broad set of immigrants who satisfy pre-defined conditions, amnesties have been the focus of much attention, and much controversy. In the last column of Table 1 we can see that some countries have never resorted to amnesties (e.g. Germany and the UK), whereas others have made use of them very frequently, especially before the 2008 crisis. More recently, large-scale regularisation programs have become rarer, even if several proposals for a major amnesty have been put forward in the US, both during the Obama and the Trump administrations.

Table 1 Estimates of undocumented immigrants and number of amnesties

Notes: The table reports minimum and maximum estimates of the size of the undocumented immigrant population for each country in 2008 (except when differently indicated), expressed as a share of the total country population or as a share of the total immigrant population. The last column reports the number of immigration amnesties adopted by each country over the period 1980-2008. * denotes low-quality estimates
Source: Authors’ elaboration on Dustmann and Frattini (2013) for undocumented immigra-tion, updated with data from the CLANDESTINO database on irregular migration for Greece, Germany and Spain, and Pew Research Center estimates for the US and on SOPEMI/OECD International migration outlook (various years) for amnesties.

Why is an amnesty introduced?

What trade-offs does a government face in the decision to implement an immigration amnesty? Under what conditions is it more likely that a government will resort to this instrument to deal with the presence of irregular migrants?

In a recent paper (Casarico et al. 2018), we consider a setting in which immigration policy involves a minimum skill requirement, in order to capture the presence of immigration restrictions which favour skilled workers, as it is common among many Western destination countries. We also assume that immigration restrictions cannot be perfectly enforced – because borders cannot be sealed or because domestic enforcement is not fully effective. Illegal immigration will then naturally arise as long as the policy is binding. To determine whether an amnesty is desirable, consider the following cost-benefit calculus. On the one hand, the legalisation will open up new employment opportunities to newly legalised immigrants, who will now be able to leave the informal sector and find jobs better suited to their qualifications in the formal economy. This will positively contribute to welfare in the host country. On the other hand, existing undocumented immigrants will tend to be on average less skilled than both existing legal migrants and the remainder of the native population. As a consequence of the legalisation, they will end up on the receiving end of the welfare state, and through this channel, will represent a burden for the host country.

The straightforward intuition from this discussion is that the incentives to support a legalisation program are stronger, the greater is the improvement in the labour market opportunities available to legalised workers as a result of them gaining access to the formal sector. They are also weaker, the larger is the fiscal leakage from natives to newly legalised migrants. 

Empirical evidence

Are the channels we have just identified empirically relevant? We can answer this question in two ways. First, exploiting a panel dataset covering a large group of OECD countries over the period 1980-2008, we carry out a cross-country analysis. In particular, we match the time of the approval of a legalisation with a wealth of characteristics of the country, which capture the working of the labour market and welfare state channels. We proxy the output gains from granting migrants access to the full set of labour market opportunities using a micro-based measure of the dispersion of educational attainment by occupation within each country. The extent of redistribution carried out by the welfare state is instead captured by public social expenditure.1Figure 2 illustrates our main results, by showing how the probability that a country introduces a regularisation programme changes, everything else equal, when either the labour market mismatch experienced by migrants (left) or the extent of redistribution carried out by the welfare state (right) increases. A larger initial mismatch, which signals larger potential gains in the labour market from a legalisation, is associated with an increase in the probability that an amnesty is implemented. A larger welfare state, instead, goes with a lower likelihood that an amnesty programme is passed.

Figure 2 Effects of the labour market and welfare state channel on probability of countries passing a regularisation programme

Note: The graphs are constructed by grouping the variable on the horizontal axis in 20 equal-sized bins, and then plotting the mean predicted probability of introducing an amnesty for countries with values of the immigrants’ educational mismatch (left panel) or of the welfare state generosity (right panel) in each bin, after accounting for all other control variables. The red line reports the regression line fitted on all the underlying data points. 

The cross-country analysis we have just presented suggests that the mechanisms identified in our previous discussion play an important role. However, it may suffer from data comparability problems and could be challenged by unobserved heterogeneity, which we can only partly address by using country fixed effects. To overcome these issues, we analyse in detail a single country – the US – and a unique policy experience, the passing of the Immigration Reform and Control Act (IRCA H.R. 3810) of 1986. The vote on IRCA by the House of Representatives is an ideal testing ground for our theory for at least three reasons. First, the enactment of this bill resulted in one of the largest legalisation programs ever undertaken in the Western world. 2.8 million individuals – or 1.2% of the total population of the country – became entitled to permanent residency, with long-lasting consequences for the US economy, and for the political debate around immigration reform. Second, IRCA was a highly controversial bill, which passed after years of negotiations with only 58% of the House supporting it. Third, the data at our disposal are unique as we can match the voting behavior of elected representatives to a wealth of constituency-level characteristics. 

In particular, implementing (and adapting) the procedure recently proposed by Borjas (2017), we identify likely undocumented migrants in the 1980 census. Using this information at the individual level, we then construct detailed measures of the labour market mismatch of illegal immigrants before the legalisation took place – based on their degree of over-education in each two-digit occupation –  and of the fiscal leakage to immigrants – based on the district’s fiscal exposure to immigration (Hanson et al. 2007) and on the extent of redistribution across districts. Figure 3 shows clearly how, everything else equal, the probability that representatives vote in favour of IRCA increases with the fraction of likely undocumented immigrants in a district that are mismatched relative to their education (left panel) and decreases with the expected fiscal leakage to immigrants (right panel). 

Figure 3 Effects of the labour market and welfare state channel on probability of US House Representatives voting in favour of IRCA

Notes: The graphs are constructed by grouping the variable on the horizontal axis in 20 equal-sized bins, and then plotting the mean predicted probability of voting in favour of IRCA for representatives of districts with values of the immigrants’ educational mismatch (left panel) or of the fiscal leakage (right panel) in each bin, after accounting for all other control variables. The red line reports the regression line fitted on all the underlying data points. 

Policy implications

European countries have recently experienced an extraordinary inflow of asylum seekers. Many of them will not be granted refugee status but are unlikely to be repatriated. Thus, policy makers will have to deal with large stocks of undocumented foreigners in the near future. 

Our analysis has identified key economic triggers which can explain the usage of a legalisation program to tackle this problem. Decision makers in our set-up are welfare maximisers caring about the well-being of native citizens. Political economy considerations are, however, very likely to be important in this context. Perceptions about the impact that a legalisation can have on future irregular flows or on how other destination countries are participating in sharing the burden of irregular migration are likely to shape preferences on how to best manage the existing stock of irregulars. As populist parties find their way into governments in many European countries, the political viability of introducing migration amnesties needs to be carefully assessed.


Borjas, G J (2017), “The labor supply of undocumented immigrants”, Labour Economics 46: 1-13.

Casarico, A, G Facchini and T Frattini (2012), “What drives immigration amnesties?”, CESifo working paper series 3981. 

Casarico, A, G Facchini and T Frattini (2018), “What drives the legalization of immigrants? Evidence from IRCA”, Regional Science and Urban Economics 70: 258-273, also CEPR Discussion Paper 12790.

Casarico, A, G Facchini and C Testa (2016), “Asylum policies and illegal immigrants: Perspectives and challenges”, CESifoDICE Report 4/2016.

Gordon, I, K Scanlon, T Travers and C Whitehead (2009), Economic impact on the London and UK economy of an earned regularisation of irregular migrants to the UK, London: LSE and Greater London Authority.

Hanson, G H, K Scheve and M Slaughter (2007), “Public finance and individual preferences over globalization strategies”, Economics and Politics 19(1): 1-33.

Hatton, T J (2011), “Seeking asylum: Trends and policies in the OECD”,, 14 July. 

Triandafyllidou, A (2009), “Clandestino project final report”, European Commission. 

UNHCR (2018), Statistical Yearbook 2016, 16th Edition, Geneva: UNHCR.


[1] We also include a set of additional drivers that might influence the introduction of a legalisation program. For more details, see Casarico et al. (2012).



Topics:  Education Labour markets Migration

Tags:  immigration, amnesties, illegality, Labour Markets, asylum, redistribution

Associate Professor of Public Economics, Bocconi University

Professor of Economics, University of Nottingham and University of Milan and CEPR Research Fellow.

Associate Professor, Department of Economics, Management and Quantitative Methods, University of Milan; CEPR Research Affiliate; CReAM Research Fellow.


CEPR Policy Research