are ireland's immigrants integrating into its labor market?

IMR

Volume 42 Number 3 (Autumn 2008):597–619

597

© 2008 by the Center for Migration Studies of New York. All rights reserved.DOI: 10.1111/j.1747-7379.2008.00139.x

Blackwell Publishing LtdOxford, UKIMREInternational Migration Review0197-91831747-7379© 2008 by the Center for Migration Studies of New York. All rights reservedXXXOriginal Articles

⅔ccupational ttainment of

¾reland s ¾mmigrants¾nternational ⅞igration eview

Are Ireland’s Immigrants Integrating into Its Labor Market?

Alan Barrett and David Duffy

1

Economic and Social Research Institute, Dublin, Ireland

Ireland has experienced a remarkable change in its migratory patterns inrecent years and has moved from experiencing large-scale emigration toreceiving significant inflows. In this paper, we use data from a nationallyrepresentative sample of immigrants and natives drawn in 2005 to assessthe occupational attainment of immigrants in Ireland relative to natives.It is found that immigrants, on average, are less likely to be in high-leveloccupations controlling for factors such as age and education. Whenlooked at by year of arrival, it appears as if immigrants who arrived morerecently have lower occupational attainment relative to earlier arrivals,thereby suggesting a process of integration. However, a closer analysisshows that the observation of better occupational attainment for earlierarrivals can be explained by a change in the national origin mix of Ireland’simmigrants, with immigrants from the New Member States of the EuropeanUnion having the lowest occupational attainment. Within national groupsthere is generally no clear evidence of improved occupational attainmentover time.

1. INTRODUCTION

In his influential paper published in 1978, Chiswick appeared to show that thewages of immigrants converged on those of natives as they spent longer in thehost country. Since then, the issue of immigrant integration in the labor marketof host countries has received further attention (

e.g

., Borjas, 1985 and, morerecently, Amuedo-Dorantes and de la Rica, 2006). In this paper, we add to theliterature by exploring the issue in the context of one of the world’s newimmigrant-receiving countries, Ireland. Having undergone a rapid economictransformation in recent years, Ireland has also seen a remarkable change in itsmigration patterns. From being a country of emigrants, Ireland is now arecipient of significant inflows. This provides migration researchers with a newlaboratory to explore issues, including that of integration.

1

We would like to acknowledge the helpful comments of John FitzGerald, David Jaeger, BrianNolan, Philip O’Connell, participants at a seminar at the ESRI, and two anonymous referees.We would also like to acknowledge specific pieces of advice from Denis Conniffe andChristopher T. Whelan. All remaining errors are our own.

598 I

nternational

M

igration

R

eview

While many studies show immigrants experiencing labor marketdisadvantages relative to natives, this is unsurprising in the case of newly arrivedimmigrants. These new arrivals may lack location-specific human capital suchas language and knowledge of the local labor market. Hence, it would beexpected that their earlier period in the host country would see them earningless or holding lower-level jobs. Concerns arise, however, for the immigrantsthemselves and for the host countries if immigrants are unable to overcomethese initial disadvantages. A failure to integrate into the labor market, bywhich we mean that immigrants experience the same levels of labor marketsuccess as comparable natives, can result in immigrants becoming an excludedminority with implications for social cohesion. The persistence of immigrantsin lower-paid and less skill-intensive occupations may also lead to a productivityloss for the host economy.

We look at the issue of labor market integration by drawing on a large-scalenationally representative sample of immigrants and natives in Ireland drawn in2005. As the data include information on the occupations of respondents, weuse this as an indicator of labor market position. As the data also includeinformation identifying immigrants and the year in which they arrived inIreland, we are able to compare labor market outcomes across immigrants andnatives, and across immigrants by year of arrival. We should note that althoughwe have information on whether the individuals are unemployed, we do notanalyze the move between unemployment and employment as another indicatorof immigrant integration, as is done by Amuedo-Dorantes and de la Rica(2006) and by Antecol, Kuhn, and Trejo (2003). The reason for this is thatimmigrants in Ireland show no statistically significant differences in terms ofunemployment propensity relative to natives.

2

However, as will be seen below,patterns of occupational attainment differ across immigrants and natives andacross different categories of immigrants.

While the data allow us to generate insights into labor market outcomesacross recent arrivals and earlier arrivals, we need to be cautious in interpretingany patterns. As we are not using panel data we cannot tell, for example, if anypatterns of improved labor market outcomes are the result of integration,cohort effects, selective out-migration, or changing immigrant (self-) selection.

2

This finding of there being no difference in the probability of being unemployed acrossimmigrants and natives is interesting in itself. It suggests that Ireland is absorbing immigrantsvery successfully in terms of employment, possibly as a result of wage flexibility. For more on thelabor market impacts of immigration in Ireland, including the issue of price and quantityadjustment,

see

Barrett, Bergin, and Duffy (2006).

O

ccupational

A

ttainment of

I

reland

’

s

I

mmigrants

599

Repeated cross-sections can sometimes be used to construct “synthetic cohorts”(

see

,

e.g

., Borjas, 1987) but this is not possible with the data used below.

3

Although the data have been collected on a quarterly basis since 1997, informa-tion on the year of arrival of immigrants is only available to us for the 2005Quarter 2 Survey. In spite of these limitations, the data do allow us to uncovera number of interesting findings.

The paper is structured as follows. In Section 2, we discuss some importantfeatures of Ireland’s recent experience of immigration by way of providing acontext for the subsequent analysis. In Section 3, we briefly set out some of themain papers on the topic of immigrant integration and in particular a paperby Chiswick, Lee, and Miller (2005). Our analysis is based on the approachtaken in that paper and so it is useful to describe it. In Section 4 we describethe data and present the results. We conclude in Section 5.

2. IRELAND’S RECENT IMMIGRATION EXPERIENCE

We will begin this brief review of Ireland’s recent migration experience bydrawing on Figure I, which shows net inflows and outflows to and from Irelandover the last twenty years. A number of points can be distilled from the figure.

3

Even if it was possible to construct synthetic cohorts, Beenstock, Chiswick, and Paltiel (2005) haveshown that this approach can lead to a positive finding with regard to immigrant assimilationwhen such a finding is not found when using panel data. They suggest that the synthetic cohortapproach suffers from survivor bias and hence can produce misleading results on assimilation.

Figure I. Net Migration to and from Ireland 1987 to 2006

600 I

nternational

M

igration

R

eview

As recently as the late 1980s, Ireland was a country of emigration. As can beseen from the negative flows in the figure, Ireland was losing over 40,000 of itspopulation per annum in the late 1980s. As the population at that time wasabout 3.5 million, population losses of over 40,000 amounted to over 1 percentof the population. For the decade 1981 to 1991, the net outflow was over200,000, or almost 6 percent of the population.

Around the mid-1990s, the economy of Ireland experienced a rapidturnaround and began to experience the highest growth rates in Europe.

4

As aconsequence of this, the population outflows were reversed and net inflowsbegan. Initially, the inflows were made up of roughly equal proportions ofnon-Irish immigrants and former Irish emigrants returning to Ireland. Sincethe mid-1990s these inflows have generally increased and the national mix hasshifted from being 50 percent Irish to being about 20 percent Irish. From2004, an acceleration in the inflow has been seen due to the ten New MemberStates (NMS) of the EU being given full access to the Irish labor market. In theyear ended April 2006, 86,000 people arrived in Ireland, with almost 40,000of them being from the NMS. Ireland was one of only three existing membersof the EU to allow full access to its labor market to the EU’s new citizens fromthe date of accession. The other two were Sweden and the UK.

As regards the administrative measures underpinning immigration intoIreland, there are a number of strands. First, all citizens of EU countries (withthe exception of Bulgaria and Romania) are free to work in Ireland withoutrestriction. Second, for non-EU members, a work permit system operates. Inessence, if an employer can show that no EU citizen is available to fill a vacancy,the employer can apply for a work permit and can recruit a non-EU national.

5

Since the opening up of the labor market to all EU citizens in 2004, thenumber of new work permits issued has fallen. For example, in 2003 almost22,000 new permits were issued. In 2006, this had fallen to 6,300 and wasa direct result of the government’s decision to balance the opening up ofthe labor market to the NMS with a more conservative approach to workpermit numbers. Refugees have full entitlements to work but this is not thecase for asylum seekers whose cases are being dealt with. They are generally notallowed to work and so must rely on state transfers.

4

For an analysis of Ireland’s economic turnaround,

see

Honohan and Walsh (2002).

5

At the time of writing, a new system of admission was being legislated that would operate alongthe lines of a “green card” system for certain immigrants, typically the more highly skilled.However, this new system is not relevant for the immigrants analyzed here as the data werecollected in 2005.

O

ccupational

A

ttainment of

I

reland

’

s

I

mmigrants

601

Existing research looking at the labor market characteristics of immigrantsin Ireland and their experiences has produced a number of insights. Ireland’simmigrant population is a highly educated one, with the proportion of degreeholders exceeding that of the domestic population (Minns, 2005; Barrett,Bergin, and Duffy, 2006). However, they do not appear to be accessing jobsthat fully reflect their education levels (Barrett, Bergin, and Duffy, 2006) andappear to be earning significantly less than comparable natives (Barrett andMcCarthy, 2007). The findings with respect to poor labor market outcomescould well be the result of the relatively recent arrival of most immigrants asthey are likely to be lacking location-specific human capital such as languageskills. None of these papers have had access to data on the year of arrival forthe immigrants and so none have been able to shed any light on the issue ofintegration.

We noted in the Introduction that Ireland has become a new laboratoryfor the study of migration. It is useful here, in the light of the discussion of thenature of immigration into Ireland, to set out what makes Ireland of particularinterest for migration researchers. First, as inward migration into Ireland hasoccurred over a period when the economy has been growing at an exceptionallyhigh rate, the economic conditions have been favorable for immigrant successin the labor market. Second, as much of the immigration into Ireland has beenfrom other European countries, many of Ireland’s immigrants will not be subjectto multiple forms of possible discrimination such as those based on religion orskin color. Both of these factors should operate in a way that would seeimmigrants in Ireland doing relatively well when compared to natives.

A third useful feature of immigration into Ireland is the concentration ofEastern Europeans in the latest inflow. This allows us to generate insights intothis new source of population movements, namely, from the new EU to the oldEU. By generating insights into who these new immigrants are and how theyare performing in their host countries, we can draw conclusions on the possibleimpacts of emigration from the NMS.

3. LITERATURE ON INTEGRATION

Chiswick, Lee, and Miller (2005) provide a useful framework in which toanalyze immigrant integration, so we will set out the details. In their model,immigrants experience a “U-shaped” pattern of occupational attainment. Inmoving between their last job in their home country and the first job in thehost country, immigrants are likely to experience downward occupationalmobility. As their skills are unlikely to be directly transferable to the new

602 I

nternational

M

igration

R

eview

setting, they may need to work in lower-level occupations at the point of theirarrival in the host country. As time goes on, the immigrants will acquirelocation-specific human capital and this will allow them to move up theoccupational ladder. In time, they may be able to move into an occupation atthe same level as the one they left. By plotting their occupational attainmentover time, the U-shape is seen.

According to Chiswick, Lee, and Miller (2005), the pattern justdescribed should be particularly pronounced for high-skilled immigrants andfor immigrants from countries where skills are less readily transferable to thehost. In the case of high-skill immigrants, there is more room to drop downthe occupational ladder when migrating and more scope for upward mobilityfollowing migration, relative to less-skilled immigrants. In the case of immigrantsfrom countries with very different labor markets, the move from one countryto another is more likely to involve downward mobility, but, as with high-skilled immigrants, as location-specific human capital is acquired, upwardmobility is possible.

We should note that unlike Chiswick, Lee, and Miller (2005), our datado not allow us to observe immigrants before they migrate and neither do wehave longitudinal data. As such, we cannot test the hypotheses in the mannerundertaken in the 2005 paper. Nonetheless, as we have data on immigrants byyear of arrival, we can think in terms of immigrant outcomes differing accordingto the length of time spent in Ireland and explore how they differ, for example,across skill groups. In so doing, we are mindful not to make strong conclusionsin respect of integration over time based on cross-sectional data, as other forces,such as selective out-migration, could be at work.

Chiswick, Lee, and Miller (2005) find evidence of occupation integrationfor immigrants in Australia. For example, those with higher levels of pre-immigration skills were found to show greater improvements in occupationalstatus, within a time frame of three years and six months. For Spain, Amuedo-Dorantes and de la Rica (2006) show that immigrants from the EU do notdisplay any occupational disadvantage relative to natives, which appearssupportive of Chiswick and colleagues’ (2005) hypotheses on migrationbetween similar countries. They also find evidence of integration among otherEuropean immigrants and Latin Americans but not among African immigrants.Green (1999) finds evidence of occupational mobility among immigrants toCanada, but the precise pattern of mobility was not consistent with a generalpattern on integration. Much of the mobility observed by Green occurred inthe late 1980s but not at other times. In addition, immigrants with differentlengths of time in Canada appeared to experience the same type of mobility in

O

ccupational

A

ttainment of

I

reland

’

s

I

mmigrants

603

the late 1980s. This suggested to Green that what was observed was a differentreaction to economic events in the late 1980s for immigrants relative to natives.Many other studies have looked at the issue of immigrant integration butthrough earnings as opposed to occupation. Among these papers are Baker andBenjamin (1994) and Borjas (1985), both of which found little evidence ofintegration.

4. DATA AND RESULTS

4.1. Data

The data used in the analysis are from the Quarterly National HouseholdSurvey (QNHS), quarter 2 of 2005 (Central Statistics Office, 2005). Thesurvey is undertaken by Ireland’s official statistical agency, the Central StatisticsOffice (CSO). The main purpose of the QNHS is to produce quarterly officialinformation on labor market variables such as employment, unemployment,and participation. Throughout the year 3,000 households are surveyed eachweek; hence each quarterly sample contains 39,000 households.

6

The CSOmake the micro-data available, although some of the collected data is omittedor aggregated. The sample has over 91,000 observations, which amounts toover 2 percent of the population. We only look at labor force participants inour analysis and so the data set on which we conduct the analysis contains justunder 35,000 individuals.

Our next task is to set out precisely how we define “immigrants” and thegroups of “natives” against which we will compare them. We define immigrantsin the following way: people who describe their nationality as being other thanIrish and were not born in Ireland. We only look at immigrants who arrived inIreland in 1995 or later so that our attention is focused on immigrants whoarrived in the context of Ireland’s changed economic fortunes. We compare thisgroup with people who describe themselves as Irish nationals and who say theywere born here. This means we omit Irish nationals who were born outside ofIreland.

7

These selections, along with the omission of individuals for whom wedo not have a complete set of data, produce two subpopulations – 32,536

6

A full description of the sampling techniques can be found in CSO (2005).

7

Because of Ireland’s long history of emigration, there is a small group of people living there whowere born outside of Ireland to Irish parents and who then returned to Ireland. These peopleare not immigrants in the standard sense. However, as they may have spent significant timeoutside of Ireland between their births and when they came to live in Ireland, it would notbe entirely correct to categorize them as natives. For this reason, we exclude them from theanalysis.

604 I

nternational

M

igration

R

eview

natives and 1,634 immigrants. The immigrants are divided into five regionalcategories: UK, USA, EU-13 (

i.e.

, the EU-15 less Ireland and the UK), theNMS of the EU, and “other.” Specific country identifiers are not providedin the data, apart from the UK and the USA.

8

4.2. Results (Descriptive)

We begin our analysis of the occupational attainment of Ireland’s immigrantsby looking at the distribution of immigrants across occupational categories. InTable 1, we show the distribution for all immigrants (who have arrived since1995) collectively and also by period of arrival. The distribution for the nativepopulation is also shown. A number of interesting points emerge.

If we compare the distributions of all immigrants and natives, we see thatimmigrants are more heavily concentrated in the lower end of the distribution.While slightly over 40 percent of natives are in the top three occupationalcategories, the corresponding figure for all immigrants is 35 percent. However,if we look across the immigrants when they are grouped by year of arrival,a different picture emerges. The most recently arrived immigrants show the

8

For people from Northern Ireland, the question on country of birth will show them as beingborn outside of Ireland (because the question refers to the Irish Republic). If they say that theirnationality is Irish, they will be excluded. If they say that their nationality is British, they will beincluded as a UK immigrant.

TABLE 1OCCUPATIONAL DISTRIBUTIONS OF IMMIGRANTS AND NaTIVES

Native%

All Immigrants

%

Arrived 2004–2005

%

Arrived 2002–2003

%

Arrived 2000–2001

%

Arrived 1995–1999

%

Managers/Administrators 19.2 9.5 7.1 9.6 8.1 15.6Professional 12.1 13 7.1 12.4 16.7 19Associate Professional/

Technical 9.5 12.4 8.4 13.7 15.2 13.9Clerical and Secretarial 13.9 10.1 8.8 9.9 11 11.5Craft and Related 15.2 16.8 23.5 15.9 11 14.2Personal and Protective

Service 11.1 17.4 19.3 19 16.7 12.9Sales 9.5 10.8 12.4 10.6 10.5 8.5Plant and Machine

Operatives 9.5 10.1 13.5 8.9 11 4.4100 100 100 100 100 100

% in Top Three Occupations 40.8 34.9 22.5 35.7 40 48.5

N 32,536 1,634 524 395 420 295

O

ccupational

A

ttainment of

I

reland

’

s

I

mmigrants

605

lowest degree of concentration in the top three occupational categories. Thisproportion increases for immigrants who arrived in earlier periods. For thosewho arrived in the period from 1995 to 1999, the proportion in the top threeoccupational groups exceeds that of natives.

The pattern shown in Table 1 is consistent with an integration story,but it is clear that other explanations could underpin the pattern and so it isnecessary to explore this much further. Before moving on to the regressionanalysis below, we can look at other possible explanations for this apparentintegration, namely, educational attainment and national mix.

9

If the mostrecently arrived immigrants have lower levels of education, this would partlyexplain the pattern in Table 1 and what is observed is a compositional effect asopposed to integration. Similarly, if the most recent arrivals are more heavilydrawn from countries whose immigrants generally do less well in Ireland’s labormarket, this too could partly explain the pattern in Table 1.

In Table 2, we show the educational distributions of immigrants andnatives. Before looking at the immigrants by year of arrival, it should be notedthat collectively immigrants in Ireland are a remarkably educated group. Thispoint has emerged from analyses of earlier QNHSs (Barrett and Trace, 1998;Barrett, Bergin, and Duffy, 2006) and also from an analysis of the 2002 Census(Minns, 2005). The most recently arrived immigrants have lower levels of

9

Of course, a changing national mix among immigrants could lead to a change in the educationalattainment of immigrants.

TABLE 2EDUCATIONAL DISTRIBUTIONS OF IMMIGRANTS AND NATIVES

Native%

All Immigrants

%

Arrived 2004–2005

%

Arrived 2002–2003

%

Arrived 2000–2001

%

Arrived 1995–1999

%

No Formal/Primary Education 10.5 4 6.1 2 4.5 2

Lower Secondary 16.7 8.4 5 9.6 8.3 12.9Upper Secondary 29 26.8 36.3 26.6 21 18.6Post Leaving 12 10.4 11.6 10.6 9.5 9.2Third Level 11.8 13.6 13 13.2 13.1 15.9Third Level – Degree

or Above 19.9 36.8 28.1 38 43.6 41.4% Third Level 31.7 50.2 41.1 51.2 56.7 57.3N 32,536 1,634 524 395 420 295

Note: The term “post leaving” refers to educational qualifications which are acquired after second-levelschooling but which are not granted by recognized third-level institutions. The term “third level”refers to qualifications that are awarded by third-level institutions but which are not degrees –diplomas would be an example.

606 I

nternational

M

igration

R

eview

education relative to the earliest arrivals, so this may partly explain their lowerlevel of occupational attainment relative to earlier arrivals. However, there is nodifference between the 2000–2001 group and the 1995–1999 group, and soeducation cannot explain the 8.5 percentage point gap in their proportions inthe higher-level occupations.

In Table 3, we show the nationality distribution of immigrants by yearof arrival and a striking picture emerges. In the period from 1995 to 1999, over50 percent of the immigrants arriving were from the UK, with only a smallproportion (2.4 percent) coming from the countries which now make up theNMS. In the most recent period, these proportions have been reversed, withover half of the new arrivals being from the NMS. Referring back to Chiswick,Lee, and Miller (2005), if the thesis is correct that immigrants from countriesthat are similar to the host country will have better labor market outcomes,then this shift between the UK and the NMS could explain the changingoccupational pattern seen in Table 1.

4.3. Results (Regression)

At this point, we will move beyond the descriptive statistics and will employregression analysis to distil the relative impact of different factors on occupationalattainment. The sets of regressions that are presented below are ordered probits,where the dependent variables are based on the set of occupational categoriesshown in Table 1. Ordered probits are used in situations where the dependentvariable takes the form of a number of discrete categories but where it ispossible to place an ordering on those categories. In the case of occupations,some are lower level and some are higher level when criteria such as wages orskill content are applied, so a ranking is possible.

TABLE 3NATIONALITY DISTRIBUTION OF IMMIGRANTS BY YEAR OF ARRIVAL

All Immigrants

%

Arrived 2004–2005

%

Arrived 2002–2003

%

Arrived 2000–2001

%

Arrived 1995–1999

%

UK 23.1 10.1 20.3 22.4 51.2EU-13 16.4 19.8 14.4 11.4 20NMS 27.7 52.9 25.6 16 2.4USA 1.7 1 2 1.2 3.4Other 31.1 16.2 37.7 49 23.1N 1,634 524 395 420 295

Note: EU-13 refers to the EU-15 less the UK and Ireland.

Occupational Attainment of Ireland’s Immigrants 607

We use a ranking based on a classification devised by CSO. When con-structing social class variables, the CSO maps occupations into social classes bygathering occupations with similar level of skills. In broad terms, and referringback to the occupations in Table 1, the mapping places professional occupationsinto the highest social class. Managers, administrators, associate professionals,and technical workers are placed in the second-highest social class. The fourcategories of clerical, craft, personal/protective, and sales are placed into thenext highest groups of non-manual, skilled, and semi-skilled. Finally, operativeare placed in the lowest social class grouping of unskilled. We use these fourcategories of occupations in our analysis.10

Given this setup, we are modeling the likelihood of being in higher-leveloccupations based on a set of explanatory variables. These explanatory variablesinclude dummy variables indicating age categories and educational categories,and also dummy variables indicating gender and marital status. We also includedummy variables indicating immigrants and natives, with the immigrantsbeing further divided into different groups in later models. This approach issimilar to that taken by Barrett, Bergin, and Duffy (2006), although they useda simpler probit analysis as opposed to the ordered probits used here.

In Table 4, we present the results from an ordered probit regression inwhich all immigrants and all natives are included. For each variable, we presentthe coefficient estimates and also the marginal impacts on the likelihood ofbeing in each of the four occupational groupings.11 Before discussing theimmigrant variable, we will take a brief look at the other variables in order toconfirm that the model produces sensible results. For the age variables, theomitted category is people aged 15 to 19 years, and so the positive coefficientsfor the older age groups make sense. The negative coefficients for the youngergroups are surprising because we would expect these groups to have higheroccupational attainment relative to the youngest group. The result can bepartly explained by the presence among the youngest group of apprenticeswhose occupations will appear skilled, even though they are in training. For theeducation dummy variables the signs and significance are also as expected. The

10For a further description see CSO (2006).11As the ordered probit involves the estimation of a nonlinear function, marginal impacts cannotbe readily inferred from the estimated coefficients in simple linear terms. In order for marginalimpacts to be calculated, it is necessary to do so by calculating the changes in probabilities asimplied by the estimated density function. In the case of dummy variables, the marginal impactsare calculated by comparing the probabilities that result when the dummy takes on a value ofzero and when it takes on a value of one, and when the other variables are held at their samplemeans (see Greene, 1993:672–676).

608 International Migration Review

omitted category is “no formal qualifications.” As can be seen from the coefficientestimates and the marginal impacts, the likelihood of being in a higher-leveloccupation increases with level of education. Time with one’s current employeralso has a positive and significant impact. For the gender/marital status dummyvariables, the omitted category is single female. Men are more likely to be inlower-level occupations, controlling for the other factors.

Turning to the immigrant dummy variable, the coefficient estimatesuggests that immigrants are less likely to be in the higher-level occupationscompared to native employees, controlling for factors such as age and education.In order to get a sense of the magnitude of the disadvantage, it is necessaryto look at the marginal impacts. Immigrants are 2 percent more likely to bein the lowest occupational category and 4 percent more likely to be in the

TABLE 4ORDERED PROBIT MODEL OF OCCUPATIONAL ATTAINMENT FOR ALL IMMIGRANTS AND ALL NATIVES

(DEPENDENT VARIABLE: OCCUPATION LEVEL)

Coef.Std. Err.

Marginal Impacts

Unskilled

Non-Manual, Skilled Manual,

Semi-SkilledManagerial

and Technical Professional

dy/dx Std. Err. dy/dx Std. Err. dy/dx Std. Err. dy/dx Std. Err.

Age 20–24 –0.15 0.04 0.02 0.01 0.04 0.01 –0.04 0.01 –0.02 0.00Age 25–34 –0.09 0.04 0.01 0.00 0.02 0.01 –0.02 0.01 –0.01 0.00Age 35–44 –0.01 0.04 0.00 0.00 0.00 0.01 0.00 0.01 0.00 0.00Age 45–54 0.04 0.04 0.00 0.00 –0.01 0.01 0.01 0.01 0.01 0.01Age 55–59 0.08 0.04 –0.01 0.00 –0.02 0.01 0.02 0.01 0.01 0.01Age 60–64 0.06 0.05 –0.01 0.01 –0.02 0.01 0.01 0.01 0.01 0.01Age 65+ 0.45 0.07 –0.04 0.00 –0.14 0.02 0.10 0.01 0.08 0.02Ed Lower Secondary 0.08 0.03 –0.01 0.00 –0.02 0.01 0.02 0.01 0.01 0.00Ed Upper Secondary 0.41 0.03 –0.04 0.00 –0.11 0.01 0.10 0.01 0.06 0.00Ed Post Leaving 0.53 0.03 –0.04 0.00 –0.16 0.01 0.11 0.01 0.09 0.01Ed Third Level

Non-Degree 1.12 0.03 –0.07 0.00 –0.35 0.01 0.17 0.00 0.25 0.01Ed Third Level

Degree 2.12 0.03 –0.13 0.00 –0.57 0.01 0.16 0.00 0.54 0.01Time with

Current Employ 0.02 0.00 0.00 0.00 –0.01 0.00 0.01 0.00 0.00 0.00Married Man –0.20 0.02 0.02 0.00 0.05 0.00 –0.05 0.01 –0.02 0.00Single Man –0.15 0.02 0.02 0.00 0.04 0.00 –0.04 0.00 –0.02 0.00Married Woman 0.03 0.02 0.00 0.00 –0.01 0.01 0.01 0.01 0.00 0.00Immigrant –0.15 0.03 0.02 0.00 0.04 0.01 –0.04 0.01 –0.02 0.00Number of Obs = 29,925LR chi2(17) = 12,091.68Prob > chi2 = 0.0000Log Likelihood = –28,991.651Pseudo R2 = 0.1726


second-lowest grouping. This mirrors the finding of Barrett, Bergin, and Duffy(2006). Using the 2003 QNHS, the results from a probit model of occupationalattainment where occupations were split into just two groupings suggested thatimmigrants were 7 percent less likely to be in higher-level occupations. Barrett,Bergin, and Duffy labeled this finding the “occupational gap” and pointed outthat it implied a loss of output in the Irish economy as it suggested thatimmigrants were working below their full potential.

As the data that Barrett, Bergin, and Duffy (2006) were working with didnot identify the year of arrival for each immigrant, they were not able to exploreif the “occupational gap” fell with years in Ireland. Our data allow us to examinethis point. In Table 5, we show results from an ordered probit regression thatis similar to that shown in Table 4 but in which the immigrant group is brokenup into four categories depending on their year of arrival. It should be notedthat although the table only shows the coefficients on the immigrant dummyvariables, the variables that are shown in Table 4 were included in the model.

The first point to be taken from Table 5 is that immigrants in the arrivalcohorts 2004–2005, 2002–2003, and 2000–2001 all show lower probabilitiesof being in higher-level occupations relative to natives, controlling for thefactors shown in Table 4. However, the 1995–1999 group are statistically aslikely to be in higher-level occupations relative to natives. Although theestimated coefficients decline across the arrival cohorts (as suggested by thedescriptive statistics shown in Table 1), t-tests show no statistical differencesbetween the three most recent arrivals.

TABLE 5ORDERED PROBIT MODEL OF OCCUPATIONAL ATTAINMENT FOR IMMIGRANTS BY YEAR OF ARRIVAL AND ALL

NATIVES (DEPENDENT VARIABLE: OCCUPATION LEVEL)

Coef.Std. Err.

Marginal Impacts

Unskilled





Arrived 1995–1999 0.08 0.07 –0.01 0.01 –0.02 0.02 0.02 0.02 0.01 0.01Arrived 2000–2001 –0.16 0.05 0.02 0.01 0.04 0.01 –0.04 0.01 –0.02 0.01Arrived 2002–2003 –0.19 0.05 0.02 0.01 0.04 0.01 –0.05 0.01 –0.02 0.01Arrived 2004–2005 –0.23 0.05 0.03 0.01 0.05 0.01 –0.06 0.01 –0.02 0.00Number of Obs = 29,925LR chi2(20) = 12,107.66Prob > chi2 = 0.0000Log Likelihood = –28,983.663Pseudo R2 = 0.1728


The pattern of coefficients in Table 5 is consistent with a process ofimmigrant integration in the labor market, although without statisticalsignificance. However, the changing national mix (as shown in Table 3) couldgenerate the same result if the national groups that make up a larger share ofthe most recent arrivals have lower levels of occupational attainment. For thisreason, we re-ran the occupational attainment ordered probit but this timeincluding both the year of arrival for each immigrant and dummy variablesindicating the immigrants’ nationalities. In order for the estimation to be possible,it was necessary to exclude one of the national groups from the list of nationaldummy variables. If we did not do this, both the year of arrival dummy variablesand the nationality dummy variables would be estimated with respect to thesame base, i.e., the native employees. As this gives rise to perfect collinearity,we excluded the USA from the list of nationality immigrant dummies. Thismeans that the year of arrival dummy variables are interpreted relative to theIrish but that the nationality dummies are interpreted relative to the Irish andthe US immigrants.12

The results are presented in Table 6 and a number of striking resultsemerge. First, the year of arrival dummy variables are no longer statistically

12We chose to exclude the USA because (a) they look most like the Irish in occupational termswhen controlling for age and education and (b) they are the smallest national group.

TABLE 6ORDERED PROBIT MODEL OF OCCUPATIONAL ATTAINMENT fOR IMMIGRANTS BY NATIONALITY AND ALL NATIVES (DEPENDENT VARIABLE: OCCUPATION IS MANAGER/PROFESSIONAL/ASSOCIATE PROFESSIONAL OR OTHERWISE)

Coef.Std. Err.

Marginal Impacts

Unskilled





Arrived 1995–1999 0.14 0.23 –0.01 0.02 –0.04 0.07 0.03 0.05 0.02 0.04Arrived 2000–2001 0.03 0.23 0.00 0.02 –0.01 0.06 0.01 0.06 0.00 0.03Arrived 2002–2003 0.03 0.23 0.00 0.02 –0.01 0.06 0.01 0.06 0.00 0.03Arrived 2004–2005 0.12 0.23 –0.01 0.02 –0.03 0.07 0.03 0.05 0.02 0.03UK 0.05 0.23 –0.01 0.02 –0.01 0.06 0.01 0.06 0.01 0.03EU-13 –0.12 0.23 0.02 0.03 0.03 0.05 –0.03 0.06 –0.01 0.02NMS –0.55 0.23 0.09 0.05 0.09 0.01 –0.13 0.05 –0.05 0.01Other –0.17 0.23 0.02 0.03 0.04 0.05 –0.04 0.06 –0.02 0.02Number of Obs = 29,925LR chi2(24) = 12,166.92Prob > chi2 = 0.0000Log Likelihood = –28,954.033Pseudo R2 = 0.1736


significant. Hence, the impression that was generated from Table 5 of loweroccupational attainment for more recent arrivals was only a part of the story.What stands out from Table 6 is that immigrants from the NMS have thelowest occupational attainment of all immigrants and to a significant degree.As these NMS immigrants are concentrated among the more recent arrivals,their lower occupational attainment led to the somewhat false impression thatemerged from Table 5. In terms of marginal impacts, the immigrants from theNMS are 9 percent more likely to be in the lowest occupational category and13 percent less likely to be in the managerial and technical group. Immigrantsfrom the other regions do not experience an “occupational gap” relative tonatives.

The results in Table 6 raise doubts about whether a process of immigrantintegration is at work in Ireland and suggest that national origin is a moreimportant determinant of occupational outcomes. In order to examine thepossibility of integration more closely, we decided to look within nationalgroups to see if there was any evidence of integration. We do this by estimatingordered probits with the same variables as used in the model in Table 5, i.e.,the set of socioeconomic controls plus year of arrival indicators for immigrants,but comparing each national group of immigrants to natives in separateregressions. As we have five nationality groups, this leads to a set of fiveregressions. We focus on the results for the UK and the NMS as these are thegroups that dominated the inflows in the earliest and latest periods, respec-tively. The results for the other three groups are discussed in brief below.13

In Table 7, we present the results from a regression in which we estimatethe occupational attainment of UK immigrants by years of arrival, relative tonatives. As there is no discernible pattern in the size of the occupational gap (oradvantage) over time, we cannot say that there is any evidence of assimilation.However, this would have been expected based on Chiswick, Lee, and Miller(2005). As UK immigrants are coming from a similar country with similarlabor market institutions, they should be able to find employment that isclosely aligned to their skills.

13We could have conducted this element of the analysis by adding year of arrival and nationalityinteraction terms to the regression reported in Table 6. However, as discussed in the body of thetext, the nationality dummy variables in Table 6 have to be interpreted relative to the Irish plusthe US immigrants due to collinearity constraints. Our main interest is in the outcomes forimmigrants relative to natives. For this reason, we take the approach of comparing each nationalgroup to the Irish separately so that the year of arrival dummy variables can be interpreted asestimating the occupational attainment of immigrants who arrived in year X from country Y,relative to the Irish.


In Table 8, we present the occupational attainment ordered probit for theNMS. It should be recalled from Table 6 that this group experienced the largestoccupational gap. Our interest here is in seeing if it diminishes according to theyear of arrival of the immigrants. The results show that this is not the case andthat the occupational gap is generally quite consistent. The earliest arrivalsappear to have identical occupational attainment relative to natives but therelevant cell size is too small for anything meaningful to be concluded. Hence,we are not finding evidence of integration. It could be that the time period overwhich we are observing the immigrants is simply too short to capture a pattern

TABLE 7ORDERED PROBIT MODEL OF OCCUPATIONAL ATTAINMENT FOR UK IMMIGRANTS BY YEAR OF ARRIVAL AND ALL

NATIVES (DEPENDENT VARIABLE: OCCUPATION LEVEL)

Coef.Std. Err.

Marginal Impacts

Unskilled





Arrived 1995–1999 0.17 0.10 –0.02 0.01 –0.05 0.03 0.04 0.02 0.02 0.02Arrived 2000–2001 –0.06 0.12 0.01 0.01 0.02 0.03 –0.02 0.03 –0.01 0.01Arrived 2002–2003 0.21 0.13 –0.02 0.01 –0.06 0.04 0.05 0.03 0.03 0.02Arrived 2004–2005 0.27 0.15 –0.02 0.01 –0.08 0.05 0.06 0.03 0.04 0.03Number of Obs = 28,355LR chi2(20) = 11,763.59Prob > chi2 = 0.0000Log Likelihood = –27,370.919Pseudo R2 = 0.1769

TABLE 8ORDERED PROBIT MODEL OF OCCUPATIONAL ATTAINMENT FOR NMS IMMIGRANTS BY YEAR OF ARRIVAL AND

ALL NATIVES (DEPENDENT VARIABLE: OCCUPATIONAL LEVEL)

Coef.Std. Err.

Marginal Impacts

Unskilled





Arrived 1995–1999 0.26 0.41 –0.02 0.03 –0.08 0.13 0.06 0.09 0.04 0.07Arrived 2000–2001 –0.55 0.12 0.09 0.03 0.08 0.00 –0.13 0.03 –0.05 0.01Arrived 2002–2003 –0.53 0.10 0.09 0.02 0.08 0.00 –0.13 0.02 –0.04 0.01Arrived 2004–2005 –0.41 0.06 0.06 0.01 0.08 0.01 –0.10 0.01 –0.04 0.00Number of Obs = 28,607LR chi2(20) = 11,838.11Prob > chi2 = 0.0000Log Likelihood = –27,586.182Pseudo R2 = 0.1767


of integration. It could also be that our cell sizes, although allowing us to capturestatistically significant differences between immigrants and natives, are notallowing us to capture within-immigrant differences. For these reasons, weneed to stress that we can only conclude that we are not finding evidence ofintegration (as opposed to saying more definitively that such integration is notoccurring).

Referring back again to Chiswick, Lee, and Miller (2005), we wouldexpect immigrants with higher levels of education to be more likely to experienceupward occupational mobility. Based on this, we reestimated the orderedprobit model for the NMS immigrants but this time restricted the sample tothose immigrants and natives with third-level qualifications. The results arepresented in Table 9 and a mixed picture emerges. The negative coefficient forthe most recent arrivals is the largest and this coefficient is statistically differentto the coefficient for the 2002/2003 arrivals (the difference between thecoefficients has a t-statistic of 2.1). However, the difference between the 2004/2005 coefficient and the 2000/2001 coefficient is not statistically significant.Hence, we are at best finding patchy evidence on integration, with immigrantsfrom the NMS with third-level qualifications who arrived in the period 2000/2001 being 22 percent less likely to be in the highest-level occupations relativeto comparable natives.

We noted in the Introduction that care must be taken in deriving anyconclusions with respect to integration over time based on a cross-section. Oneof the reasons for this is that the nature of the immigrant selection process

TABLE 9ORDERED PROBIT MODEL OF OCCUPATIONAL ATTAINMENT FOR NMS IMMIGRANTS BY YEAR OF ARRIVAL AND

NATIVES WITH THIRD-LEVEL QUALIFICATIONS (DEPENDENT VARIABLE: OCCUPATION IS MANAGER/PROFESSIONAL/ASSOCIATE PROFESSIONAL OR OTHERWISE)

Coef.Std. Err.

Marginal Impacts

Unskilled





Arrived 1995–1999 0.35 0.52 –0.01 0.01 –0.09 0.12 –0.03 0.07 0.13 0.21Arrived 2000–2001 –0.77 0.19 0.08 0.03 0.22 0.04 –0.08 0.04 –0.22 0.04Arrived 2002–2003 –0.67 0.19 0.06 0.03 0.19 0.05 –0.06 0.03 –0.20 0.04Arrived 2004–2005 –1.16 0.13 0.17 0.03 0.27 0.01 –0.16 0.03 –0.28 0.02Number of Obs = 9,385LR chi2(15) = 474.35Prob > chi2 = 0.0000Log Likelihood = –10,811.698Pseudo R2 = 0.0215


could change over time. As the time period we are looking at involves a changein the immigration regime facing people from the NMS, it is important toconsider whether the pattern of results in Tables 8 and 9 could have arisen fromchanging immigrant selection. For example, it could be the case that thepostenlargement arrivals would only migrate when they could be assured ofslotting into better-suiting occupations on arrival because of the free movementof labor. This would lessen the occupational gap for this group, making it closerto the earlier arrivals and so removing an appearance of integration.

This hypothesis may be true but it is difficult to test it. If we look at NMSimmigrants who arrived just before and just after enlargement, we find verylittle difference in terms of observable characteristics such as labor force statusand education. For example, about 27.1 percent of those who arrived in the years2002/2003 have some form of third-level education, with the correspondingfigure for the 2004/2005 group being 27.2 percent. While this tells us that thegroups are similar in some respects, we are not able to say much more aboutthe broader selection issue.

Before leaving the issue of integration, we should note that little furtherevidence of such integration emerged from the analysis of the three othernational groups – USA, EU-13, and “other.” In general, there was no readilyobservable pattern in the occupational gap by year of arrival, and even when therewas, the differences in the coefficient estimates were not statistically significant.One small exception emerged when we ran the model for immigrants from theEU-13 with third-level qualifications. The results are shown in Table 10. Withthe obvious exception of the earliest arrivals, we do see a declining coefficient

TABLE 10ORDERED PROBIT MODEL OF OCCUPATIONAL ATTAINMENT FOR EU-13 IMMIGRANTS BY YEAR OF ARRIVAL AND

NATIVES WITH THIRD-LEVEL QUALIFICATIONS (DEPENDENT VARIABLE: OCCUPATION LEVEL)

Coef.Std. Err.

Marginal Impacts

Unskilled





Arrived 1995–1999 –0.11 0.18 0.01 0.01 0.03 0.05 0.00 0.00 –0.04 0.06Arrived 2000–2001 0.25 0.22 –0.01 0.01 –0.07 0.06 –0.02 0.02 0.09 0.08Arrived 2002–2003 –0.15 0.18 0.01 0.01 0.05 0.05 0.00 0.00 –0.05 0.06Arrived 2004–2005 –0.35 0.14 0.02 0.01 0.11 0.04 –0.01 0.01 –0.12 0.04Number of Obs = 9,411LR chi2(15) = 359.60Prob > chi2 = 0.0000Log Likelihood = –10,821.483 Pseudo R2 = 0.0163


for the earlier years of arrival, with the difference between the 2004/2005 andthe 2000/2001 groups being statistically significant (t-statistic is 2.3). However,this is relatively weak evidence of increased integration over time.

5. CONCLUSION AND DISCUSSION

The occupational attainment of Ireland’s immigrants relative to natives,controlling for age and education, is lower for more recent arrivals. However,this does not appear to be the result of increased labor market integration overtime. Instead, it is related to a changing national mix in the immigrant inflowinto Ireland. Immigrants from the NMS are heavily represented among themost recent arrivals and they suffer the largest occupational gap. In addition,there appears to be no lessening over time in the occupational gap experiencedby immigrants from the NMS. Hence, we are not finding evidence of immigrantlabor market integration over time.

The focus of the analysis was on the occupational gap, controlling for ageand education. In the context of our findings on that point, it is useful to returnto an issue looked at above, namely, the educational distribution of immigrants.In Table 2, we looked at the distribution of educational qualifications by yearof arrival. In Table 11, we again show the distribution of educational qualifica-tions of immigrants but this time by nationality. The point that emerges is thatimmigrants from the NMS have the lowest level of education attainment, asmeasured by the proportion with third level-degrees. This implies that thechanging national mix is contributing to a reduced educational attainmentfor the newest arrivals in Ireland (see Table 2) and also a lower occupationalattainment controlling for education (see Table 5).

With respect to educational qualifications, it should be noted that theproportion of immigrants with third-level qualifications (31.8 percent) is identical

TABLE 11EDUCATIONAL DISTRIBUTIONS OF IMMIGRANTS BY NATIONAL GROUP

UK%

EU-13%

NMS%

Other%

USA%

All Immigrants%

No Formal/Primary Education 2.4 1.1 6.4 4.7 0.0 4.0Lower Secondary 18.3 2.2 9.3 3.5 7.1 8.4Upper Secondary 19.6 22.4 37.8 25.2 17.9 26.8Post Leaving 11.4 8.2 14.6 7.1 10.7 10.4Third Level 15.1 14.2 12.6 13.6 3.6 13.6Third Level – Degree or Above 33.3 51.9 19.2 45.9 60.7 36.8

100 100 100 100 100 100N 378 268 452 508 28 1,634


to the proportion with third-level qualifications in the native Irish population.For this reason, the conclusion should not be drawn that immigration intoIreland is becoming low skilled. Instead, the more modest conclusion can bedrawn that the exceptionally high-skilled nature of Irish immigration, noted inearlier studies, is being weakened as a result of increased immigration from theNMS. This policy-induced switch in the national origin mix of immigrantsand the consequent change in human capital characteristics is reminiscent ofthe effects in the US of the 1965 Amendments to the Immigration andNaturalization Act, as discussed in Borjas (1987). As much of the analysis ofthe economic impacts of immigration to Ireland has been based on the inflowbeing highly skilled (Barrett, FitzGerald, and Nolan, 2002; Barrett, Bergin,and Duffy, 2006), one implication of the findings here is that the estimates ofthe impacts may need to be reestimated.

These findings with respect to education and occupational attainmentmight be of more limited concern if evidence of increased integration over timewas found, but this is generally not the case. As these immigrants had been freeto work in Ireland for a full year at the time the sample was drawn,14 it is some-what surprising that a greater degree of upward mobility was not found. Itcould be that the immigrants who arrived before 2004, i.e., before full rightsto work applied, were working illegally and that a “scarring” effect has arisenwhereby it is difficult to break out of a weak labor market situation. Otherpossible explanations for the lack of integration include language skills and thenonrecognition of qualifications. It could also be that the time frame is simplytoo short for integration to be observed. For this reason, it will be importantto reexamine this issue as more data become available. Whether the lack ofintegration is short-term or long-term, it is not possible to get a clear sense ofwhat might be causing it from the data used here. However, it is important tohighlight the possible existence of barriers to mobility in Ireland so that anysuch barriers can be lowered, thereby allowing Ireland to avoid the difficultiesexperienced by other countries in terms of immigrant integration.

It is useful to consider what these results tell us that is of specific relevanceto the possible impacts of emigration for the NMS. In Table 11, we haveshown that 31 percent of NMS immigrants in Ireland have third-level degrees.We know from Ireland’s Census 2006 that immigrants from Poland are by farthe largest group among the total immigrant group from the NMS (63,000 outof a total of 120,000), so by comparing the 31 percent with the proportion of

14The New Member States entered the EU on 1 May 2004; the data were collected in the springof 2005.


the Polish labor force which has a third-level qualification, we can get a senseof whether emigration from Poland to Ireland is selective of the more skilled.As the corresponding figure for Poland is 16 percent (OECD, 2006), it certainlydoes appear as if Poland is losing a disproportionate share of its high-educatedemployees.15

Ireland itself experienced a “brain drain” of this sort in the 1980s. However,when those who left in the 1980s returned in the 1990s, they appear to havereturned with enhanced human capital, as evidenced by the higher earnings ofreturning migrants relative to nonmigrants (Barrett and O’Connell, 2001). Ifthe highly skilled Polish emigrants return to Poland with similarly enhancedhuman capital, then the concerns surrounding the brain drain phenomenonwill be reduced. However, as the results here appear to show that immigrantsfrom the NMS are not accessing higher-skilled occupations, in which theywould acquire enhanced skills and competencies, this positive effect maynot arise.

The somewhat negative findings here in respect of immigrants from theNMS are very similar to the findings of Drinkwater, Eade, and Garapich(2006), who look at the experience of this group in the UK.16 According totheir results, immigrants from the NMS who arrived postenlargement earnbetween 33 percent and 40 percent less than other European immigrants whoarrived during the same period, controlling for human capital characteristics.This measured earnings gap drops somewhat when occupation is controlledfor, thereby suggesting that a part of the earnings disadvantage is related toNMS immigrants not accessing higher-level occupations. Citing research byEade and Garapich (2006), Drinkwater, Eade, and Garapich (2006) point outthat many Polish immigrants in the UK expect to spend a relatively long timein the UK. For this reason, the issue of long-run occupational attainment forNMS immigrants will be an important one in both Ireland and the UK andmay be worth studying in a comparative context given the apparently similaroutcomes in the two countries.

15The only other NMS covered by the OECD data is Slovakia – the proportion of the labor forcewith third-level education there is 13 percent.16As is the case for Ireland, the majority of NMS immigrants in the UK are from Poland – over60 percent, based on registrations.


REFERENCES

Amuedo-Dorantes, C., and S. de la Rica 2006 “Labor Market Integration of Recent Immigrants in Spain.” IZA Discussion Paper No.

2104. Bonn: IZA. <www.iza.org>.

Antecol, H., P. Kuhn, and S. Trejo 2003 “Integration via Prices or Quantities: Labour Market Institutions and Immigrant

Earnings Growth in Australia, Canada and the United States.” IZA Discussion PaperNo. 802. Bonn: IZA. <www.iza.org>.

Baker, M., and D. Benjamin 1994 “The Performance of Immigrants in the Canadian Labour Market.” Journal of Labor

Economics 12:369–405.

Barrett, A., A. Bergin, and D. Duffy 2006 “The Labour Market Characteristics and Labour Market Impacts of Immigrants in

Ireland.” Economic and Social Review 37:1–26.

———, J. FitzGerald, and B. Nolan 2002 “Earnings Inequality, Returns to Education and Immigration into Ireland.” Labour

Economics 9(5):665–680.

———, and Y. McCarthy 2007 “Immigrants in a Booming Economy: Analysing their Earnings and Welfare Dependence.”

Labour 21(4–5):789–808.

———, and P. J. O’Connell 2001 “Is There a Wage Premium for Returning Irish Migrants?” Economic and Social Review

32(1):1–21.

———, and F. Trace 1998 “Who Is Coming Back? The Educational Profile of Returning Migrants in the 1990s.”

Irish Banking Review Summer:38–52.

Beenstock, M., B. R. Chiswick, and A. Paltiel 2005 “Endogenous Assimilation and Immigrant Adjustment in Longitudinal Data.” IZA

Discussion Paper No. 1840. Bonn: IZA. <www.iza.org>.

Borjas, G. 1987 “Self-Selection and the Earnings of Immigrants.” American Economic Review 77(4):531–

553.

Borjas, G. J. 1985 “Integration, Changes in Cohort Quality and the Earnings of Immigrants.” Journal of

Labor Economics 3:463–489.

Central Statistics Office 2006 Census 2006 Volume 8 – Occupations. Dublin: CSO.

——— 2005 Quarterly National Household Survey Quarter 2 2005. Dublin: CSO.

Chiswick, B. R. 1978 “The Effect of Americanisation on the Earnings of Foreign-born Men.” Journal of Political

Economy 86:897–921.

———, Y. L. Lee, and P. Miller 2005 “Longitudinal Analysis of Immigrant Occupational Mobility: A Test of the Immigrant

Integration Hypothesis.” International Migration Review 39:332–353.

http://www.iza.org

http://www.iza.org

http://www.iza.org


Drinkwater, S., J. Eade, and M. Garapich 2006 “Poles Apart: EU Enlargement and the Labour Market Outcomes of Immigrants in the

UK.” IZA Discussion Paper No. 2410. Bonn: IZA. <www.iza.org>.

Eade, J., and M. Garapich 2006 Researching Class and Ethnicity: London’s New Poles. Mimeo. CRONEM: University of

Surrey.

Green, D. A. 1999 “Immigrant Occupational Attainment: Integration and Mobility over Time.” Journal of

Labour Economics 17:49–79.

Greene, W. H. 1993 Econometric Analysis, 2nd edn. New York, NY: Macmillan.

Honohan, P., and B. Walsh 2002 “Catching Up with the Leaders: The Irish Hare.” Brookings Papers on Economic Activity

2002(1):1–57.

Minns, C. 2005 “How Skilled Are Irish Immigrants? Evidence and Implications.” Paper delivered to the

Statistical and Social Inquiry Society of Ireland, 24 February.

OECD 2006 Education at a Glance. Paris: OECD.

http://www.iza.org

are ireland's immigrants integrating into its labor market?

Documents