Increasing Labour Supply through Immigration - CiteSeerX

DISCUSSION PAPER

Increasing Labour Supply through Immigration Discussion Paper Alan Barrett Economic and Social Research Institute Section 1: Introduction The purpose of this paper is to explore how immigration has contributed to increases in labour supply in Ireland over the last decade. We also consider how immigrant labour has impacted upon variables such as wages and GNP growth. As this paper will be accompanied by another prepared by the Department of Enterprise, Trade and Employment on policy matters, we do not engage in a full policy discussion here. Instead we provide a brief note later in this Introduction on policy on labour migration. The paper is structured as follows. In Section 2, we present data on Ireland’s recent labour market and migratory experiences. In particular, we aim to show what proportion of recent labour force increases have been accounted for by immigration. In Section 3, we go on to look at the characteristics of immigrants. Section 4 contains estimates of the impacts of immigration. In Section 5, we conclude with remarks on policy implications. Before getting into the detail of the paper, we will set out the broad parameters of Ireland’s policy on labour migration. All citizens of the European Union can work freely in Ireland, including those of the Accession Countries. This sets Ireland apart from most EU countries that chose to restrict immediate labour movement from the Accession States. For non-EU citizens, the main route for entry for work purposes is through the work permit system. Under this system, employers are allowed to recruit workers from outside the EU once they have demonstrated that no EU citizen can be readily found to fill the position in question. Permits are issued for one year, although they can be renewed. The permit only allows the worker to work with the company who applied for the permit – transfer to new companies are not allowed unless the new company applies for a permit for the employee. Table 1 shows the number of work permits issued and renewed between 1999 and 2004. It is clear from the table that the number of permits expanded greatly between 1999 and 2003. However, in 2003 a more restrictive approach began to be applied, partly in anticipation of the opening up of the labour market to the Accession States. Ruhs (2005) makes the interesting observation that the average number of new permits issued between May and December 2004 was very similar to the number issued to non-EU 25 citizens in the period just before enlargement (about 500). Hence, there appears to be an on-going demand for labour from outside of the expanded EU. Table 1: Work Permits Issued (New and Renewals), 1999 to 2004 New Renewed Group Total

1999 4,036 1,448 266 5,750

2000 15,335 2,201 297 17,833

2001 29,491 6,919 346 36,756

2002 23,207 16,861 436 40,504

2003 22,050 25,110 547 47,707

2004 10,020 23,246 801 34,067

Source: Ruhs (2005)

23 – 24 June 2005

Peer Review

Increasing Labour Supply through Economic Migration, Ireland

1

DISCUSSION PAPER

Other routes for labour entry include the granting of refugee status (although asylum seekers, while waiting for applications to be processed, are not entitled to work), student visas (with restrictions on the number of hours that can be worked) and marriage to an EU citizen. Section 2: Ireland’s Labour Market and Migration Experience, 1994 to 2004 We will begin by presenting basic statistics in Table 2 on Ireland’s labour market and migration experience between 1994 and 2004. The dramatic trends in employment and migration can be seen. Between 1994 and 2004, employment grew from 1.22 million to 1.84 million, an increase of 50%. This increase in employment was accompanied by a dramatic fall in unemployment (from 14.7% in 1994 to 4.4% in 2004) and by a reversal in Ireland’s migratory pattern. From being a country experiencing net outflows in the 1980s and early 1990s, net inflows emerged in the mid-1990s and then grew. Table 2: Ireland’s Labour and Migratory Experience 1994 to 2004

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Labour Force Employment UE Rate Inflows Outflows Net migration 000s 000s % 000s 000s 000s 1431.6 1220.6 14.7 30.1 34.8 -4.7 1459.2 1281.7 12.2 31.2 33.1 -1.9 1507.5 1328.5 11.9 39.2 31.2 8 1539 1379.9 10.3 44.5 25.3 19.2 1620.4 1494 7.8 46 28.6 17.4 1685.9 1589.1 5.7 48.9 31.5 17.4 1745.9 1671.4 4.4 52.6 26.6 26 1787 1721.9 3.6 59 26.2 32.8 1840.9 1763.9 4.2 66.9 25.6 41.3 1875.5 1793.4 4.4 50.5 20.7 29.8 1920.3 1836.2 4.4 50.1 18.5 31.6

Source: CSO (2004), Ruhs (2005)

As the figures on migration in Table 2 include non-labour force participants, we need to refine the figures somewhat in determining the contribution of immigration to labour supply growth. In Table 3, we show the sources of labour supply growth for the periods 1995 to 2000 and 2000 to 2005, with the latter figures being based partly on projections. Between 1995 and 2000, immigration contributed just over a fifth of labour supply increases (0.75 of 3.38%). For the latter period, this proportion has risen to almost 30%. This proportion is also reported by the Central Statistics Office (2005); they estimate that close to a third of the increase in the labour force between September-November 2003 and 2004 could be accounted for by immigration. This translates into about 22,000 people out of a total labour force increase of 65,000 over that period.

23 – 24 June 2005

Peer Review


2

DISCUSSION PAPER

Table 3: Sources of Labour Force Growth (annual averages, as % of the labour force) Natural Increase Female Participation Male Participation Migration Labour Supply Migration as % of total increase

1995-00 2000-05 1.17 1.09 1.43 0.63 0.03 -0.15 0.75 0.65 3.38 2.22 22% 29%

Source: Bergin et al (2003)

In order to gain a fuller understanding of the link between immigration and labour supply in Ireland we need to look a little deeper into the figures and in particular to consider the number of former Irish emigrants that are included in the inflow figures. Given the potentially different implications of nonnational immigration and return migration, it is important to make this distinction. In 1994, 55.8% of the inflow was made up of Irish nationals. By 2000, this proportion had fallen to 47.1% and by 2004 it was down to 33.7% (see Table 4 below). As we do not have information on the nationalities of those leaving Ireland each year, it is not possible to say what proportion of the net inflow is Irish. We also don’t have published information of the labour force status of the returning Irish so it is difficult to determine precisely the contribution of non-Irish immigrants to the labour force increase. However, taking the figure for recent years of about 30% of the labour force increase being migration-based and 30% of the inflow being Irish, we can estimate that about 20% of the labour force increase relates to non-Irish nationals. At this point, we need to refer to what Ruhs (2003) calls a “health warning”. The information we have just presented is based on CSO data and estimates. However, figures from the Department of Enterprise, Trade and Employment (DETE) on work permits when combined with figures from the Department of Justice, Equality and Law Reform (DJELR) on asylum applications suggest larger inflows of non-EU nationals. Taking 2002 as an example, CSO (2004) estimate an inflow of 24,400 non-EU national. In the same year, there were 23,326 new work permits issued and 11,634 asylum applications, suggesting a total inflow of over 35,000 non-EU nationals. There are a number of possible reasons for this. For example, some work permits may be for very short durations; also, some work permits may be issued but not taken up. Whatever the reason for the discrepancy on inflows, Ruhs (2003) notes that CSO figures on stocks of non-nationals match more closely administrative estimates of stocks. In Section 3, when looking at the characteristics of immigrants, we focus on stocks and so can be more confident of the data. In order to develop further the picture of the inflow into Ireland, we need to look at another dimension of the inflow – EU nationals versus non-EU nationals. EU national (or more precisely, EEA nationals) can work in Ireland without work permits and any other form of authorisation. However, non-EEA nationals need a work permit, unless they are admitted on other grounds (such as family re-unification). The importance of this distinction is that EEA nationals can almost be thought of as being beyond immigration policy considerations, as least as far as admission is concerned. Since 1 May 2004, the right to work for EEA nationals has been extended to nationals of the accession states of the EU and so the policy control has been loosened further.

23 – 24 June 2005

Peer Review


3

DISCUSSION PAPER

In Table 4, we show the proportions of the broad national groups in the inflow from 1999 to 2004. A number of points should be noted. First, and as already noted, the proportion of Irish people in the inflow is declining. Second, the proportion from the “rest of the world” is increasing. Third, even in 2004 when the Irish component was at its lowest level, two-thirds of the inflow was from the EU and so not subject to any policy restrictions. Table 4: Proportions of Nationalities in the Inflow, 1999-2004 Irish UK Rest of EU USA Rest of World Total

1999 55% 17% 14% 5% 9% 100%

2000 47% 16% 16% 5% 16% 100%

2001 45% 15% 11% 6% 23% 100%

2002 40% 11% 12% 4% 32% 100%

2003 35% 14% 14% 3% 35% 100%

2004 34% 12% 21% 4% 30% 100%

Source: CSO (2004); Rest of EU refers to EU-15

Although CSO (2004) does not contain a more detailed breakdown of the “rest of the world” category, we can get some sense of the distribution of countries concerned from Ruhs (2003) analysis of work permit data. Looking at permits issued between 1999 and 2003 (a total of 106,811, including both new permits and renewals) Ruhs shows that the largest proportion (10.6%) went to Latvia, followed by Lithuania on 7.7%, the Philippines on 6.9%, Poland on 6.8% and South Africa on 5.6%. Although we do not yet have published estimates on estimated inflow of migrants for the year up to April 2005, it is important to have some sense of what has been happening over this period given the opening up of the Irish labour market to the accession states on 1 May 2004. The only source of information at the moment is data derived from the issuing of social insurance numbers that are needed when entering the Irish labour market. Table 5 shows the numbers issued by nationality for the period May-December 2004. Table 5: National Breakdown of Social Insurance Numbers Issued, May-December 2004 Poland Lithuania Latvia Slovakia Czech Republic Hungary Estonia Malta Slovenia Cyprus All

25,222 11,410 5,769 4,513 3,061 1,753 1,637 130 64 23 53,582

Source: Department of Social and Family Affairs, presented in Ruhs (2005)

The figures in the table would suggest a very large inflow of accession state nationals since 1 May 2004 and this may well be mirrored in the forthcoming CSO inflow estimates. However, some notes of caution

23 – 24 June 2005

Peer Review


4

DISCUSSION PAPER

should be made. First, it is possible that many of the people involved were working illegally in Ireland before 1 May and that the figures in Table 5 point to people regularising their situation as opposed to new arrivals. Also, social insurance numbers may have been issued but not used. Section 3: Characteristics of the Inflow From Section 2, we have seen how immigration has contributed significantly to Ireland’s labour force growth over the least decade. In this section, we look beyond the numbers and consider the characteristics of the immigrants. In doing so we draw heavily on a recent paper by Barrett et al (2005). The data used by Barrett et al come from the 2003 (Q2) version of the Quarterly National Household Survey. This survey is undertaken by the CSO and its main purpose is to produce quarterly labour force estimates. Throughout the year 3,000 households are surveyed each week; hence each quarterly sample contains 39,000 households. Barrett et al present a range of descriptive statistics on immigrants and natives where immigrants are defined as people who are not Irish nationals, were not born in Ireland and have been in Ireland for less than ten years. Here, we will focus on the following variables: age, gender, labour force status, education and occupation. In Table 6, we show the age distribution of natives (column headed “Irish”) and immigrants, with the immigrant group subdivided into four categories. The picture that emerges is of an immigrant population that is young relative to the native population. While 44.2 percent of the native adult population is aged between 20 and 44, the corresponding figure for the immigrant population is 81.5 percent. Table 6: Age Distribution of the Native and Immigrant Populations (%s)

Age Group

15-19 20-24 25-34 35-44 45-54 55-59 60-64 65+

Total N

Rest of EU-15 American Irish UK 10.7 7.4 4.4 4.8 10.2 7.8 28.2 18.1 16.5 27.3 46.2 25.7 17.5 29.2 15.1 31.4 16.9 12.5 3.6 8.6 7.2 6.3 1.1 2.9 5.6 3.8 0.9 3.8 15.4 5.7 0.5 4.8 100.0 100.0 100.0 100.0 71995 992 563 105

Other 5.7 24.7 45.4 18.4 4.7 0.6 0.4 0.2 100.0 1868

Total immigrant 6.0 20.3 39.9 21.3 6.8 2.3 1.5 1.9 100.0 3528

In Table 7, we look at the gender distribution. There is little difference between natives and immigrants, or across the immigrant categories, with the immigrant population having almost the same proportion of males as the native group.

23 – 24 June 2005

Peer Review


5

DISCUSSION PAPER

Table 7: Gender Distribution of the Native and Immigrant Populations (%s)

Male Female Total N

Irish 49.0 51.0 100 71,995

UK 50.9 49.1 100 992

Rest of EU-15 American 47.1 48.6 52.9 51.4 100 100 563 105

Other 52.7 47.3 100 1,868

Total immigrant 51.2 48.8 100 3,528

Table 8 shows the distribution of immigrants and natives across ILO categories. The figures shown imply a labour force participation rate of 63.2 percent for immigrants as against 58.1 percent for the native population. It is likely that this difference is partly explained by the age differences across the two groups. The unemployment rates in Table 8 are 6.8 percent for immigrants and 4.2 percent for natives and so in relative terms, there is a substantial divergence. This observation of higher unemployment among immigrants is not unusual (see, for example, Frijters et al 2003). Generally, possible explanations would include a lack of local labour market information that reduces job search efficiency and a lack of informal local contacts. However, the higher unemployment rates amongst UK and US immigrants relative to the “rest of the EU-15” do not fit well with these explanations. Many of the immigrants in the “other” group are likely to need work permits to be in Ireland, so the low unemployment rate among them is likely to be driven by that requirement. Table 8: ILO Labour Force Status Distributions of the Native and Immigrant Populations (%s) Rest Total of EU-15 American Other Immigrant 62.9 37.1 47.4 50.4

In Employment full-time In Employment part-time, not underemployed In Employment part-time, underemployed Unemployed, seeking full-time work Unemployed, seeking part-time work Marginally attached Others not economically active Total

Irish 45.8

UK 50.4

9.8 0.1 2.0 0.4 0.4 41.5 100.0

9.0 0.0 4.2 1.2 0.3 34.9 100.0

9.2 0.2 3.2 0.5 0.5 23.4 100.0

3.8 0.0 5.7 0.0 1.9 51.4 100.0

7.9 0.2 3.6 0.2 0.5 40.1 100.0

8.3 0.1 3.8 0.5 0.5 36.3 100.0

Participation rate UE rate N

58.1 4.2 71,995

64.8 8.4 992

76.0 4.9 563

46.7 12.2 105

59.3 6.4 1,868

63.1 6.8 3,528

Narrowing the focus to people who are labour force participants, Barrett et al look at the education levels of immigrants relative to natives. Barrett and Trace (1998) had observed that immigrants in the mid1990s had a very high level of educational attainment relative to Ireland’s own population. However, with an expansion of the work permit in the late 1990s and early 2000s and an over-representation among

23 – 24 June 2005

Peer Review


6

DISCUSSION PAPER

work permits of lower skill occupations, the suspicion existed that the educational attainment of immigrants would have slipped by the early 2000s. Table 9 contains the Barrett et al findings with respect to educational attainment and it can be readily seen that the immigrant population in Ireland is still characterised by high educational levels. This finding replicates that of Minns (2005) which was based on data derived by the OECD from the 2002 Census. Over a half of immigrants (54.2 percent) have third level qualifications, compared with just over a quarter (27.3 percent) of the native population. At the other end of the educational distribution, while 32.9 percent of the native population have only lower secondary qualifications or less, only 15.1 percent of the immigrant population have this low level of attainment. Table 9: Distributions of Educational Attainment for the Native and Immigrant Populations (%s)1

No formal/ primary education Lower secondary Upper secondary Post Leaving Third level – non-degree Third level - degree or above Total N

Irish 13.7 19.2 27.5 12.3 10.6 16.7 100.0 41,612

UK 6.7 19.6 18.8 10.5 15.8 28.4 100.0 626

Rest of EU-15 1.3 2.5 24.9 8.1 14.5 48.6 100.0 393

American 4.3 14.9 6.4 74.5 100.0 47

Other 6.8 5.8 23.3 8.3 12.9 42.8 100.0 878

Total immigrants 5.5 9.6 22.0 8.8 14.0 40.2 100.0 1,944

An alternative, although less direct, look at the skills distribution across the two populations can be taken by looking at the occupations held by both immigrants and natives. This is done in Table 10 where the occupational distributions of the different groups are shown. The figures on education presented in Table 9 would have led to an expectation a wider divergence in occupational distributions than is found in Table 10. While immigrants are more heavily concentrated in professional/associate professional occupations, this is only marginally the case. Also, immigrants are less well represented in the management/administrator category. As the “other or not stated” category captures the low skilled occupations, it appears that immigrants are slightly more concentrated in this group relative to natives2.

1 2

As the sample is now restricted to labour force participants, cell sizes are getting smaller. For this reason, caution is needed in interpreting the figures within immigrant categories, especially the US category. Minns (2005) also points to this education/occupation discrepancy when comparing the education levels of immigrants from particular countries with their occupations, based on data arising from the issuing of work permits.

23 – 24 June 2005

Peer Review


7

DISCUSSION PAPER

Table 10: Occupational Distribution of Natives and Immigrants (%s) Irish Managers and administrators 17.7 Professional 10.5 Associate professional and technical 8.7 Clerical and secretarial 12.2 Craft and related 13.6 Personal and protective service 9.8 Sales 8.2 Plant and machine operatives 9.8 Other (includes not stated) 9.6 Total 100 N 41,831

UK 16.8 14.5 10.9 11.5 13.5 11.8 5.9 7.2 7.9 100 643

Rest of EU-15 9.1 15.2 10.7 16.8 5.6 19.9 9.3 6.1 7.2 100 428

American Other 22.4 6.9 22.4 10.7 18.4 11.6 6.1 6.6 0.0 10.4 12.2 20.2 8.2 6.4 4.1 11.6 6.1 15.6 100 100 49 1108

Total Immigrants 10.5 12.9 11.4 10.0 10.1 17.5 6.9 9.1 11.6 100 2,228

While the education level of the immigrants should work to increase their occupational level, the relative youth of the group (as shown in Table 6) should work in the opposite direction. In order to isolate the effects of these variables and to establish if there is an independent “immigrant” effect on occupational attainment, Barrett et al used regression analysis. They ran a probit model with the dependent variable equal to one if the individual’s occupation is manager, professional or associate professional and zero otherwise. The independent variables include age cohorts, education levels, time with the current employer/self-employed, time since graduation and a dummy variable equal to one if the individual is an immigrant and zero otherwise. The estimated coefficient of the immigrant dummy variable was negative and significant thereby indicating that in Ireland immigrants have lower levels of occupational attainment, controlling for the other variables3. There are a number of possible explanations for this finding. It could be that employers in Ireland discount educational qualifications earned outside of Ireland and so will offer a lower occupational grade to an immigrant relative to a similarly qualified Irish national. Alternatively, it could be the case that recently arrived immigrants lack local labour market knowledge and so accept jobs below those appropriate to their skill levels while they search for better jobs4. Minns (2005) suggests that the work permit system might be responsible if high-skilled people accept work permits for low-skilled jobs with a view to getting a foothold in the Irish labour market. In order to investigate this finding of an immigrant “occupational gap” more closely, Barrett et al re-ran the probit model but distinguished between the four groups of nationalities. One question that can be answered using this approach is whether the work permit system is responsible for the occupational gap. The results showed that the occupational gap applies to both “rest of EU-15” immigrants and to “others”. This would suggest that the work permit system is not solely responsible since “rest of EU-15” immigrants are outside of this system. Also, the estimated gap for “rest of EU-15” immigrants was found to be stronger than that for “others”. The fact that UK and US immigrants suffer no occupational disadvantage prompts a suspicion that the occupation gap may be related to English language skills. It 3 4

A similar observation has been made in the case of Israel which has also experienced high-skilled immigration (Eckstein and Weiss, 2004) It should be noted that most of the immigrants would have arrived in the later part of the ten-year period under investigation rather than in the earlier part.

23 – 24 June 2005

Peer Review


8

DISCUSSION PAPER

could also be that these groups have a greater willingness to accept lower level jobs while searching for better jobs. Section 4: Impacts of the Inflow Again drawing on Barrett et al (2005), we now present estimates of the impacts of labour immigration. These impacts are simulated using a structural model of the Irish labour model that has been estimated using data from 1966 to 2002. The approach used by Barrett et al is to run a simulation whereby labour supply is increased in a way that captures both the number of immigrants and their skill levels for the ten- year period to 20035. Noting that there is a discrepancy between the immigrants’ education levels and their occupations, two simulations were run. In the first simulation, the immigrants are treated as if they are ”fully employed” and are added to the Irish population taking full account of their higher education levels. In the second simulation, their education levels are discounted; in particular, they are assumed to have the same levels of education as the Irish population, reflecting the similarity in occupational distributions. The results of the first simulation are presented in Table 11. This simulation highlights the crucial role of immigration in the determination of wage and employment levels and the positive role that skilled immigration has played in the Irish economy. The increase in the high-skilled labour supply exerts downwards pressure on the high-skilled wage, which falls by around 6 percent in the long run, with the precise estimate depending on which adjustment process in the low-skilled labour market is assumed. This fall in labour costs implies an improvement in competitiveness and thereby leads to an increase in output, employment and productivity, with GNP being between 3.5 and 3.7 percent higher than it would have been in the absence of the skilled immigration. For low skilled native employees, the precise impact again depends on whether market clearing is assumed or not but in either case the impact is positive. A rise in low-skill labour supply will in itself suppress the wage but because low and high-skilled labour are complements in production, the rise in high-skilled immigration increases the demand for low-skilled labour. Where there is market clearing, low-skilled wages increase by 1%; without market clearing, low-skilled unemployment falls by 1.2 percentage point.

5

This approach was used by Barrett, FitzGerald and Nolan (2002) which contains an analysis of the impact of immigration into Ireland in the mid-1990s.

23 – 24 June 2005

Peer Review


9

DISCUSSION PAPER

Table 11: Effects of Immigration into Ireland (1999-2003) with Full Utilisation of Immigrant Skills No Market Clearing in Market Clearing Low-Skilled Labour Low-Skilled Labour Market Market % Change GNP per head GNP per worker GNP Total Employment High-skilled Low-skilled Labour supply High-skilled Low-skilled Average wage High-skilled Low-skilled As % of labour force: Unemployment rate Low-skilled unemployment rate

1.1 1.0 3.7 2.6 2.8 2.4 2.1 2.8 1.0 -4.8 -6.1 0.0

0.9 0.9 3.5 2.5 2.6 2.3 2.5 2.6 2.3 -4.5 -5.9 1.0

-0.5 -1.2

0.0 0.0

Table 12 contains the result of the simulation when the educational levels of the immigrants are discounted. The overall impact of immigration on the economy is now more muted when the occupational gap is taken into account. The increase in GNP falls from 3.5-3.7% to 2.8-3.0%. This is because wages fall by less than in the previous simulation and so the impact on Irish competitiveness is smaller. The smaller output effect means that the demand impact on low-skilled workers is reduced. This leads to low-skilled wages falling in this simulation, when market clearing is assumed (or low-skilled unemployment rising, when market clearing is not assumed)6.

6

A similar result is obtained by Cohen and Hsieh (2000) when they look at the impact of high-skilled immigration in Israel. They argue that the Russian influx of the 1990s did not lower native skill-premia because the Russians experienced the type of occupational gap observed here and so did not change the relative supply of skilled workers in Israel.

23 – 24 June 2005

Peer Review


10

DISCUSSION PAPER

Table 12: Effects of Immigration Adjusting for Occupational Penalty No Market Market Clearing Clearing in Low- Low-Skilled Labour Skilled Labour Market Market % Change GNP per head GNP per worker GNP Total Employment High-skilled Low-skilled Labour supply High-skilled Low-skilled Average wage High-skilled Low-skilled As % of labour force: Unemployment rate Low-skilled unemployment rate

0.5 0.8 3.0 2.1 2.2 1.9 2.2 2.2 2.1 -3.9 -4.9 0.0

0.4 0.8 2.8 2.1 2.1 1.9 2.0 2.1 1.9 -3.7 -4.6 -0.2

0.1 0.2

0.0 0.0

Section 5: Discussion The brief review of Ireland’s experience of immigration in recent years has highlighted the following points.

Immigration has played a key role in increasing labour supply. With returning Irish migrants included, migration contributed about one third of the labour supply increase in the last five years. Non-Irish immigration contributed about 20%. Even before EU enlargement, a large proportion of immigration into Ireland was outside of the direct policy sphere. In the year ended April 2004, only one third of the migrant inflow was non-EU. Since EU enlargement and the extension of freedom of labour movement to the accession states, more immigrants will have arrived in Ireland without being subject to the work permit or asylum process. In spite of the dominance of EU citizens in the inflow (including people from Ireland and the UK), the non-EU component expanded both in relative and absolute terms from the mid- and late-1990s. The work permit system played a key role in facilitating this. Total permits issued (including both new permits and renewals) increased from 5750 in 1999 to 47,707in 2003.

In addition to the number of immigrants who have arrived in Ireland in recent years, the other notable feature has been their high level of educational attainment relative to the domestic population. It cannot be argued that this was the result of policy. Where policy had a role in selecting immigrants, i.e. through work permits, the occupations were typically low-skilled. For example, in 2002 two-thirds of work permits were in respect of occupations categories as “personal and protective services” and “other” (which are

23 – 24 June 2005

Peer Review


11

DISCUSSION PAPER

primarily low-skilled). However, even immigrants from countries requiring work permits had high levels of education. It has been argued that the work permits system may have worked to keep people in occupations below those that their education may have allowed. However, the finding by Barrett et al (2005) of lower occupational attainment relative to educational attainment for other non-English speaking immigrants suggests that the work permit system is not the only problem. While the policy-makers should generally be congratulated for allowing significant inflows into the Irish labour market (partly through expanding the work permit system and then allowing free access to accession state citizens), some limited criticisms can be made. First, while the work permit system may not be the only factor to restrict highly skilled immigrants from accessing high-skill jobs, it almost certainly has not helped. Allowing individuals to hold permits as opposed to firms would likely assist labour mobility. Second, the restrictive nature of the work permit system (in terms of time and employer limits) may have discouraged certain immigrants from coming here. Third, by restricting immigrants on work permits in their capacity to quit jobs, the potential for exploitation on the part of employers was increased. While much of the policy discussion in Ireland has related to issues of admission, the finding by Barrett et al (2005) of an occupational gap for immigrants raises the issue of whether greater efforts in respect of integration are needed. As the finding is not based on a dynamic analysis of immigrant experiences, we do not know if the occupation gap disappears over time as immigrants assimilate into the Irish labour market. Hence, it may be too early to be definitive on this point. However, it is an issue that should be kept under review because the potential loss in employing highly skilled people in low-skilled occupations could be significant. As a final point, it is worthwhile to raise the following point for discussion. The inflow of immigrants into Ireland and the liberal policies in respect of admission have all occurred in a context of rapid economic growth. Ireland has yet to deal with the issue of large-scale immigration in a context of a faltering economy. It remains an open question as to how well Ireland is equipped to deal with the potential issues (both economic and social) that might then arise. References Barrett, A, Bergin, A. and D. Duffy (2005), “The Labour Market Characteristics and Labour Market Consequences of Immigrants in Ireland”, IZA Discussion Paper No. 1553 (Bonn, Germany). Barrett, A., J. FitzGerald and B. Nolan (2002), “Earning Inequality, Returns to Education and Immigration into Ireland”, Labour Economics Vol. 9 No. 5. Barrett, A. and F. Trace (1998), “Who Is Coming Back? The Educational Profile of Returning Migrants in the 1990s”, Irish Banking Review, Summer. Bergin, A., J. Cullen, D. Duffy, J. FitzGerald, I. Kearney, D. McCoy (2003), Medium Term Review 20032010, Dublin ESRI. CSO (2005), Quarterly National Household Survey – Quarter 4 2004, Dublin: CSO CSO (2004), Population and Migration Estimates, Dublin: CSO. Eckstein, Z. and Y. Weiss (2004), “On the Wage Growth of Immigrants: Israel 1991-2000”, Journal of the European Economic Association, Vol. 2 No. 4.

23 – 24 June 2005

Peer Review


12

DISCUSSION PAPER

Frijters, P., M.A. Shields and S. Wheatley-Price (2003), “Immigrant Job Search in the UK: Evidence from Panel Data”, IZA Discussion Paper No. 902. Minns, C. (2005), “How Skilled Are Irish Immigrants? Evidence and Implications”, paper delivered to the Statistical and Social Inquiry Society of Ireland, 24 February. Ruhs, M. (2005), Managing the Immigration and Employment of non-EU Nationals in Ireland, Studies in Public Policy No. 19, The Policy Institute, Trinity College Dublin. Ruhs, M. (2003), “Emerging Trends and Patterns in the Immigration and Employment of Non-EU Nationals in Ireland: What the Data Reveal”, Working Paper, The Policy Institute, Trinity College Dublin.

23 – 24 June 2005

Peer Review


13

TURKEY

23 – 24 June 2005

Peer Review


14